Windows XP source code has leaked | Hacker News Reader

Windows XP source code has leaked

gizmodo.com.au

485 pointsa5withtrrs473 comments

Forgeting about legality for a moment, this is great. Windows XP is probably the greatest operating system if you factor usability and how many people it has reached. It was also from a time where people still did not expect proprietary operating systems to be actively working against you. So it's great that this source will not be forever lost.

> proprietary operating systems to be actively working against you

Where is this happening ? Are we expecting this now ? What did I miss

Windows 10 is a resource hog and contains adware.

macOS ... Is special. Some of the most notable changes in the last few years have been to try to restrict what an app can do.

I agree with Spyware but resource hog? Everyday performance is certainly better than on earlier Windows versions.

My experience is that windows 10 spends more time on scanning for viruses and serving ads than useful work. Maybe your experience is different.

It sounds like you got infected with adware. Maybe next time... don't turn off the updates? :)

So you're telling me that macOS is more performant than Windows 10? Have you used both? Do you not feel the delay in every mouse click you do in macOS that is so maddening after coming from Windows?

Both the UI lag and general performance in macOS is just a sad story.

My windows machine runs 24/7 for months without needing a restart. I am forced to restart macOS weekly, usually because of system issues.

I don't share your experiences.

I'm a developer on Windows and OSX, so I have MacBook Pro and a ThinkPad.

I want to run the ThinkPad over with a car and then drop it in the sea.

MacOS is far better. This mouse click delay you're talking about, is in your head.

To be fair, it can happen when the kernel is very busy, for example while paging. Even with a SSD and 16 GB RAM, it happens regularly for me due to bloated webpages, or apps like VS Code.

I used macos, I was forced to restart and update. I used windows, I was forced to restart and update. I use kubuntu now, and I haven't turned my computer off(unless you count sleep state) in a few weeks, sometimes i've gone months. Both operating systems feel slow to me now that I've used linux. I was apprehensive to use it at first, and feel kinda guilty evangelizing it, but if you have these problems, I think this is your solution.

I’ve been on macOS for 16 years. Never has it forced me to restart or update (not counting panics or other bugs).

Exactly. Having it ask you if you want to update is not the same as in Windows where you get forced.

I have zero problems with my Mac and I use it all day long for work. I can put off updates for as long as i like. It doesn't get in the way or spy on me like Windows 10, or attempt to throw up shitty ads. It's also snappy. If yours isn't then there's something wrong with your Mac; some software that you installed or your hardware is buggy.

yes. most of the things I see about windows vs linux involve "well you can just turn all the spyware stuff off, all you have to do is dive into these 10 menus and flip these switches" "Its easy to disable the ads in the start menu" etc etc. Even if you can turn the stuff off, the fact that you have to feels a lot like the OS working against the user.

And the next update may reset some of them.

The Edge update comes to mind that had reset the browser preference presumably to be able to push edge again.

Because that's how installing software works on Windows. You install a new program which can open JPGs, it's not allowed to grab control of all JPGs unilaterally. Instead it registers itself as a handler of JPGs. Next time you try to open one, Windows prompts you "you have a new program which can do X, what do you want to use?" and you get to keep your old one, or choose the new one, once or every time.

Get a new browser installed, prompt for your choice of browser. This is an improvement to "get Edge installed, have Edge take over everything silently" and an improvement on "I installed a thing and it doesn't do anything, where is it", and an improvement on "IE 11 for eternity".

after the edge update, when i started my computer, edge opened itself and forced me through a "first run" mini tutorial thing that i couldnt skip or ignore. it wasn't even a "you have a new program" prompt, it was a "YOU ARE USING EDGE NOW".

one of the updates most certainly was not just a mime handler update. They pushed a new edge (when they swapped to chromium) and rather than leave people with old edge as their browser, they just flat out set new edge as your browser without checking what you had otherwise, and set it to auto run on your next boot.

I agree with you but you could say the same thing about a lot of linux installations.

"Its easy to install arch, all you have to do is run these 10 arcane commands"

With some linux distros is definitely feels like it is hostile to entry-level users. Which is probably by design, but I'm just saying that Windows isn't the only OS where you have to do a bunch of weird hostile setup stuff when you first install it

The amount of user data collected in Windows 10 for example. Next to no user benefits, just value extracted for Microsoft.

Automatic updates on reboot. For example when someone giving a presentation or class has to restart their computer for some reason (quite common in Windows still, or one just ran out of power). Boot up the computer, people waiting impatiently - and then have to wait for 5-30 minutes for the computer to finish updates before being able to use it. No way to cancel/post-pone, no time-estimates for when it will be done.

I go months without needing to reboot windows, and have for over a decade. Typically if I need to reboot it’s due to an update or software that requires a reboot to install.

How? When I last used Windows it forced a reboot to do updates once a month. If I let it pick the time it was often a bad time.

To be fair, that's typically a policy enforced by administrators: "patch X should be force-installed if not present by Date Y" + "reboot should be forced if not happening after Z days from patching" => "reboot will happen at Date Y + Z, regardless of user's wishes".

O&O ShutUP 10. It will let you disable the automatic update nonsense (along with a bunch of other nonsense).

I dont use Windows much, but you can go to services and disable Windows the update service and.. in the Wifi settings check a box that says you have a limited connection.

Unfortunately, this is not possible on their consumer OS branch (unless you hack the system, but I have found that it often makes the OS unstable because its in an unsupported system state). For me, our IT manages the systems. They get to decide when the systems update. I have a test box that regularly sees months of uptime.

https://i.imgur.com/XN2Hjia.png

The bigger issue, I think, I really having to restart just to update minor things i.e. what linux distro needs to restart just to update (the equivalent of) visual studio

Unpopular opinion but: telemetrics help developers prioritize stuff and triage bugs. From a user’s point of view, isn’t that a substantial benefit?

Telemetrics are frequently inaccurate, sometimes to the point they paint a picture are interpreted in a a way that’s the opposite of reality. Windows telemetry in particular is not nice because it’s not really possible to turn off on the lower editions and it also does not seem to translate to benefits–especially as around the time of its inclusion word got out that Microsoft was cutting a large portion of their QA team.

> especially as around the time of its inclusion word got out that Microsoft was cutting a large portion of their QA team.

Or telemetry made it way simpler.

What are you basing this on? Are you talking about Windows or just random telemetry?

The first statement was general, the second is specific to Windows.

Very good software was written before people that think like you were taken seriously. Clearly it is not necessary.

You know you can disagree with me while still respecting my opinion?

Also, beneficial does not imply necessary.

I can respect you as a person while still thinking your opinion is not to be taken seriously. Respecting an opinion which is this far off the mark isn't something I'm willing to do.

Today's software is vastly better in UX than software in the 90s and 00. Certainly I don't attribute all of that to telemetry, but your opposite view is incorrect.

That kind of telemetry was in Windows 95 already with dr Watson. If people want the benefit of having the problems they run into fixed and they believe telemetry helps with that, they can just turn it on. No need to force things.

> No need to force things.

Forced telemetry and opt-out telemetry are two different things.

The former is clearly evil. The latter can be helpful if it’s communicated clearly and not hidden in some EULA.

Crash reports usually yes (but how to prevent receiving personal data from stack dumps?), features selection probably not. The reason is that you don't get data from the people who disable telemetry and they could use different features. You must find them, poll them and merge the results with what telemetry gives you. If you don't do that, telemetry is detrimental to those people.

Yeah, such statistics will always be skewed to some extent.

One could argue that telemetry _is_ the poll. If a user decides to disable feature-usage-tracking telemetry, and then realize they’d prefer to have a voice regarding feature selection, one possible solution would be for them to re-enable it.

If they absolutely won’t opt-in, they’re always free to email developers to tell them about their use case – a pretty powerful way to make yourself heard.

I haven't seen any benefit. Why should microsoft know everything and record everything I do with my computer. It could be opt in I suppose. But as it is, it is -extremely- hard for the regular user to opt out.

Indeed, Windows update delaying a reboot at a wrong time can affect real world events. See this example from 2015:

[German pro basketball team relegated to lower division due to Windows update | Ars Technica](https://arstechnica.com/information-technology/2015/03/germa...)

> "But as both teams warmed up, the computer crashed," he said. "When we booted it again at 7:20pm, it started automatically downloading updates. But we did not initiate anything."

> After all the updates were installed, Paderborn was ready to start the game at 7:55pm.

> But then Chemnitz formally protested, saying that because Paderborn had delayed the start time of the match by 25 minutes (instead of the 15-minute maximum as allowed under the German basketball rules), they should be penalized. As a result, Paderborn lost another point (Google Translate) in the standings, according to a Basketball Budesliga press release, which meant that it would certainly be relegated to the "ProB" league of German pro basketball.

The telemetry is shady but the automatic updates are a... necessary evil. People would delay updates for MONTHS by clicking that postpone button.

Windows is by design meant to be used by not very computer literate people and unfortunately these kinds of users need to be forced to do the right thing sometimes. As they said back in the day: "the user will do everything possible to see the dancing bears, including downloading that toolbar, installing that gizmo, disabling security settings".

People smugly complain about this, in many cases from their Linux box (= technical users on a system with a 1% desktop market share => smaller target and better human protection) or MacOS box (= slightly more technical users than the average Windows user, on a system with a 5%-10% desktop market share).

Da, it is for your own good comrade!

I really wish people who insist on claiming some right to control how users use their own computers would leave IT forever.

And if it were left open ended at the request of people like you...we'd still have compromised systems out there spread amongst normal people as well as within infrastructure and other more critical pieces of hardware.

People cannot be trusted to upgrade. It REALLY IS good to force people to upgrade.

> People cannot be trusted to upgrade. It REALLY IS good to force people to upgrade.

The very idea that users should be at the mercy of their computers instead of the other way around is contrary to the concept of personal computing.

People like you are why I hate IT these days. You seem to think that because you have some specialized knowledge about computers that you're smarter than everyone who doesn't. You're totally willing to sacrifice hours of peoples lives, ruin their projects, and break their systems in the name of "I know better". It doesn't matter what they're doing, or how important it is to their lives, what matters is what you think is important.

It's not about specialized knowledge...it's about regular maintenance.

Hate me or not...but people can't be trusted (broadly) to maintain things...thats just the way life works.

Why should they be forced to, at the expense of things that matter to them, though? That's the problem.

Automatic updates were a default. People turned them off because they were annoying. One solution to this problem was to make them not-annoying, but instead Microsoft opted to become aggressive and force people to update.

I will never hold anyone who thinks this is a good idea in high regard because they are actively making the world a worse place.

Microsoft forces them because it has to, there is no law for it. To drive a car, the car needs to pass a yearly inspection.

It saves them money in support calls.

Based on how ludicrously frequently updates are breaking things, this is a dubious assertion.

They have to be forced to because they often refuse to.

It's really that simple...cry about it all you want. It's a necessary evil...because humans.

I fully understand the frustration after seeing people having their computers reboot in the middle of presentations and other important events. That's why they made multiple products. Consumer OS = you don't want to be the sysadmin of the computer. Enterprise/LTSB OS = do what you want.

I still get forced reboots on my workplace Win10 laptop. I can just postpone the update like 4 weeks. And then stuff brake.

> People like you are why I hate IT these days. You seem to think that because you have some specialized knowledge about computers that you're smarter than everyone who doesn't. You're totally willing to sacrifice hours of peoples lives, ruin their projects, and break their systems in the name of "I know better". It doesn't matter what they're doing, or how important it is to their lives, what matters is what you think is important.

A thousand times this.

>People cannot be trusted to upgrade. It REALLY IS good to force people to upgrade.

This is such a stupid, frustrating argument to hear.

Pro and Enterprise shouldn't be the same in this regard as Home but often are.

Four days ago I had to entertain a courier waiting to pick up a 6-figures piece of equipment for a waiting air-freight cargo plane, for immediate transport to a remote mining site, all because the fucking piece of shit fucking Win 10 Enterprise on a very expensive rugged laptop decided to hang itself in an update loop for an hour.

We dealt with different garbage during the Windows 7 and previous eras, yes, but I'm not sure we're better off on the whole. Windows 10 has given a whole new meaning to the "Blue Screen of Death".

> This is such a stupid, frustrating argument to hear.

Reality is often frustrating, I agree.

To be fair it asks to update or not to update. You can also set how often it does check and when it installs (at night). It's just a google search away.

Windows 10 telemetry, secretly turning on auto updates even after you turn it off and correct me if I am wrong, even if you disable the update service or add the urls to the hosts file it now "mysteriously" undoes that stuff over time.

Windows 10X will probably have some more goodies.

I believe auto updates aren't malicious. Microsoft had a huge problem with worms infecting millions of computers that used outdated Windows version. It made Windows less secure in public opinion, even though these issues had often been fixed a long time.

Forcing people to update was probably the only way to stop that.

I'm also not a fan, I got burned at least once by a Windows 10 computer restarting in the middle of the night when I had programs running on it, but I guess it's one of these things that's good for 99%, the 1% is just overrepresented here.

If you want people to accept forced updates, just send the security fixes. Don't change the UI and don't mess with the users preferences, especially concerning privacy.

Honestly, I've never heard anyone complain apart from tech people. Long reboots are annoying but overall, Windows 10 is pretty well accepted by every non-technical person I know.

Oh, the non-tech people hate it too, but what can they do? They have literally zero choice in the matter.

That's the thing with tech, and that's why I consider all arguments that "people voted with their wallets" to be bullshit in this space. Most users don't understand how computers work. They don't have mental models to recognize how things should work. They have no choice but to accept whatever they're being given.

Devices around them getting less ergonomic, slower, and more flashy UIs? "There are smart people building this magic, they must know what they're doing!". Their computers slowing down to the point of uselessness in the span of a year? "It must be these viruses!". Techies complain because techies understand it's all greed and laziness, and that same technology is capable of being much better than it is.

I mean, ask your parents whether they like their operating system. I'll bet their answer will be a resounding "no", and you may get an earful about their general frustrations with technology.

One notable moment where the dissatisfaction of masses was voiced in unison was the forced Windows 8 to Windows 10 upgrade. You didn't have to have any sophisticated mental model there to recognize you're being made to do something bad for you against your will.

That's because people expect technology to suck[0]. Users can be unhappy, even if they don't realise it.

[0] https://www.kilobitspersecond.com/2020/09/22/people-expect-t...

Because non-technical people probably don't even know they have a choice in the matter. Macs are too expensive and they've probably never heard of Linux.

What are you talking about? Non-tech people complain all the time about it. Updates deciding to run right before you have to give a critical presentation is a meme at this point.

All this "well, users are stupid so it is ok to treat them like garbage" nonsense is part of the reason I've come to fucking hate IT. I wish all of you who think that way would go away.

Out of curiosity, would you mind if there were forced updates for security only?

I'm mostly very strongly opposed to forced updates, but things like WannaCry make me understand the viewpoint for forcing security upgrades.

I'm not trying to be combative, I'm just trying to better work out my own views on the topic.

A friend of mine uses his PC only in the weekends (he uses the company's PC during the week) and hates Windows because every time it takes hours to download and install updates over his slow ADSL (I think there is no fast Internet there.) He has to remember to turn the PC on in the morning to use it in the afternoon. If he forgets, next weekend. Of course he could turn it on Friday night but the user experience is still really bad.

That makes no sense, I never had Windows block waiting to download updates and not let me use the computer.

It swamps the connection.

I work at a repair shop, and like clockwork, we get non tech people complain about windows updates breaking their stuff.

Unfortunately, Microsoft also has a trust problem with updates. Back in Windows XP era, there was no question about whether you wanted to upgrade to SP2 or not -- you'd maybe wait for a few days/weeks to make sure no major bugs were in, but there was no need to force anything.

Today, you might run an upgrade and reboot to find a more secure operating system -- or a full-screen Edge ad, with Edge being immediately auto-started afterwards and re-pinned on the taskbar. (Guess who had to drive across town in the middle of the Covid-19 lockdown because one of their parents had to teach an online class and couldn't open Firefox anymore. Yep.)

It's a hole that Microsoft dug themselves in, and long before Windows 10, and instead of doing anything to dig themselves out of it, they just dragged everyone else down.

It solves the problem -- nobody's claiming it doesn't -- but it's not something that should be defended for technical or usability reasons, it's the worst solution that happens to still work, more or less.

But you experience the same when you login to GMail and it has a new interface or your Phone looks different in the morning because of iOS/Android updates.

Users are so used to be annoyed that updates disrupt UX, Microsoft isn't an especially bad actor here. In the end, users tend to disable updates as a response which is bad for security and makes vendors force them to update.

Maybe not "improving" UX with every release would be a solution but there are very few companies that stick with their UX for many years, redesigns are way too easy to sell as a new feature.

I don't think it is sufficient to justify bad choices because everyone is making them. Especially in the case of an os, which is a lot more critical than the Gmail web interface or something similar.

A long time ago, I had a GEOS installation that used to reliably crash every Sunday at 12 PM.

Should mandatory reboots every Sunday at 12 PM be introduced in a future Windows update? I mean, someone else has done it before. If you do it every second Sunday it's even better, Microsoft wouldn't be an especially bad actor here.

The fact that other companies are even worse doesn't make this any better. For any program that has a bug, I guarantee there are thousand other programs with even worse bugs. Pointing fingers and returning EWONTFIX may fly for some FOSS projects, where maintainers can at least claim that if you want a fix in open-source software that you didn't pay for you should fix it yourself, but it's a pretty ridiculous stance if you have paying customers.

I don't have a problem with forced security fixes. I DO have a problem with forced updates which are completely unrelated to security, install thinly veiled ads disguised as apps, break existing functionality, change my browser/PDF reader preferences, occasionally brick the whole machine.

Also I have a huge problem with so much of this crap needing a reboot. Reengineer the damn thing so it doesn't in 99% of cases.

How many forced Windows 10 updates have bricked peoples computers? Just looking at the dates after googling has dates from '18, '19 and '20. So at least once a year. On an OS that has been out for 5 years now.

Do I control my computer or not? If I do then I should be able to control what happens when.

You control your computer, turn off updates, get malware, then come back and complain how M$ / Windows is crap. I really don't understand why so many people in this thread are actively complaining about not being able to turn off updates. It's like once a month and they give you advance notice and "pick a time"; you have a whole week to choose a good time to restart and apply some updates...

Because the user should be in control, no microsoft.

I'm actually OK with Microsoft being in control of security patching cadence. Too many people are simply unable, unqualified or unwilling to do the job themselves, and when there's millions of them, you get botnets. I don't like the loss of choice, but I readily admit it's a necessary evil.

But they should do much better QA and not brick anyone's PC ever, and absolutely NOT ship anything which isn't strictly security-related in this undeferrable-unskippable security patches channel.

I do not know when there will be a good time for it, how long it is going to take, what it may break, or what settings it may reset. I update my Linux distro whenever I feel like it. I definitely like it this way. I have no "malware" on my system.

From my perspective windows updates are malware. I received one of Microsoft's malware updates on Windows 10. The computer forces updates on reboot except the update never completes. I've waited 2 hours for an update and let the computer stay on overnight. After 20 hours it still didn't complete. The solution? Reboot and turn windows updates off by disabling the windows update service. Except then they decided to override that option and suddenly my system started boot looping again. Why can't they just fix their crappy software? They've had two decades. It's always the exact same failure type on Windows Vista, Windows 7 or Windows 10. Sometimes you you can clear the windows update cache by hand but guess what. That only works if you can actually disable the windows update service which they prevent you from doing now.

Yeah, but when Fall Creators Update cripples pen functionality to the point that I assumed it was a bug and rolled back:

http://forum.tabletpcreview.com/threads/fall-creators-update...

and there's no way for me to use my machine as I purchased it (Samsung Galaxy Book 12 bundled w/ a Staedtler Noris Digital Stylus) and I've had to roll back an un-asked for upgrade since, the bottom line is Microsoft needs to completely decouple the OS layer where the security updates occur, from the UI layer.

If I pay for a system with an active stylus, I don't want it to be dumbed down to an 11th touch input which scrolls and won't select text or interact w/ the system as I've been accustomed to since Windows for Pen Computing.

I do not like the idea of not being able to control my personal computer and wasting my time because of updates that may or may not break crap, that may or may not reset some settings. Ugh.

>I got burned at least once by a Windows 10 computer restarting in the middle of the night when I had programs running on it

That means you have been spared. For me "getting burned" means endless boot loops and reinstalling windows. Windows updates have bricked my system on every version of windows so far. Windows Vista, 7, 10. Their updates all suck.

Yes, I'm one of those "stupid" people who installed updates as soon as possible. No more bullshit. I'm using Linux now.

Nostalgia is, it seems, a hell of a drug.

In early Windows XP days everybody said it was crap compared to Win2k. XP was the first one with product activation as well. And people called the Luna theme childish.

Then Vista came along and people said XP was the good one.

Then Windows 7 came along and nobody said Vista was the good one.

i think the consensus is that windows 95, XP, 7 were the least troublesome versions. I still think 7 > 10

10 has tons of under-the-hood improvements that are great, honestly.

However, it also has a ton of advertisments everywhere, and those annoying auto-installing apps.

I don't use Windows day to day anymore but, my family has Windows 10 and I manage it.

When I sit in the front of the computer, I feel like I'm inside the big space ship in the Wall-e. All the advertisements, forced consumerism and everything.

I had run tr0nscript on the computer I support for family a few years ago, and that made the experience almost reasonable.

I know tr0nscript has fallen out of favor since, don’t know the exact reasons. I did need to fix the hosts file to allow windows to get security updates - but otherwise, that machine still feels as usable as win7 except for the degraded shell (which could be replaced with classic shell, I guess, but I didn’t bother).

> I know tr0nscript has fallen out of favor since, don’t know the exact reasons.

Looked to what it's doing in every step [0]. It's a bit overzealous while doing what it's doing. Also, its default are not the most sensible ones so, if you're not running it after installation it may change the behavior of the system quite a bit.

Removal of metro apps, resetting IE, purging %APP_DATA% storage, removal of some VCS snapshots... It's a thermonuclear option to take. To be able to run in a milder, sensible manner needs a lot of tuning. So it's not practical.

[0]: https://github.com/bmrf/tron/blob/master/README.md#full-tron...

I really don't know what people are talking about ads? My win 10 has zero ads on it. I turned off all telemetry (i think) with a couple of programs dedicated to doing that. Maybe my pi-hole is blocking the ads?

And all forget that 8 even existed.

Although I find 8.1 still better than 10. Some of its preinstalled apps and UI were just weird, but at least it wasn't anti-user by design. Both are a clear downgrade to Windows 7 though.

Personal experience would indicate that a lot of people hated Windows 8, not just because it took something they were familiar with and threw the biggest curveball at it possible, but because they actively made the UI/UX significantly more difficult to use.

For me the icing on the cake was segregating different apps between "tablet mode" and "desktop mode," where the former would take up the entire screen with no reasonable way to multitask between programs. Not only that but how many apps had (and even still have) both a UWP and standard implementation, from my experience that just confused users as to which they were supposed to use. Heck even today we still have the Settings app and the Control Panel (which is increasingly difficult to access I might add), both with various controls and settings only found in one but not the other.

IMO Windows 10 (sans tracking and forced pre/auto-installing apps) is what Windows 8 should have been from the beginning. Many beneficial changes under the hood, and minor UI changes that don't entirely change the workflow of the OS, but instead augment it for a better experience for both regular and power users. Somehow we've eclipsed the 5-year mark since its release, yet Windows 10 still feels incomplete. Not even just that, but they manage to make it feel even less finished with every new update.

I think the biggest change that killed 8 was the lack of the start menu. Having to go to the start screen - pulling you away from your entire desktop just to launch an app - was awful. I know Win 8.1 got it back, but for the time between 8 and 8.1 Classic Shell was almost necessary to use Windows 8. All that for no discernible benefit over using 7.

Trauma is best to forget

8 was great on the intended devices. I used it on an asus transformer tablet/laptop and it worked really well. W10 was a step down for that machine because they dialed back the tablet UI.

Windows 8 is fine, _if_ you actually know how to use Windows. (Protip: Win-D takes you to the desktop from the app grid.) I would go so far as to say it's good, if you're using a touchscreen or tablet; the live tiles were useful.

Always amused me that the OS X Launchpad is also a ugly grid of huge app icons that takes up the entire screen, just like the Windows 8 start screen, yet nobody says a word about that.

Launchpad is simple to make work. I don't use Mac, but I can use Launchpad.

I do use Windows, in as much as I have to for work, and I still can't make Win8's launcher work.

Edit: Wrong version of Windows.

To me it's always been win2k, win 7, win 10, win 98. In tthat order. I've been using win2k rather than windows xp when I need legacy application support, though you might need to jump through some hoops for XP compatibility and drivers.

Though I've been inpressed by how snappy windows 8 feels, especially the boot time which is the fastest of all versions.

8 was the first version that hibernates the kernel on shutdown so it can reload it from snapshot on boot. That’s why all windows versions since 8 boot so fast.

Still Windows 8 boots faster for me. A clean installation is 2 seconds from bios to login (not hibernated, but "fast startup"), and perhaps 5-6 seconds for windows 10 (also with fast startup enabled, which it is on both on a default install).

The humorous but not wrong "consensus" was that you always had to skip a version. In this case 95 > not 98 > 98SE > not Me > XP (this was a bit of a departure because it was a successor to both the 9x and the NT lines) > not Vista > Win 7.

And both XP and Win 7 were criticized at the beginning because they had to compete with very solid predecessors, and because they came with plenty of teething problems even if they had potential. That potential was fully realized some SPs later.

Yet, NT4 -> W2K -> XP.... strange how the consumer degraded but the business was a real progression. I was using NT4 at my first job in 98, through till 2000 became more wide spread. Win2000 was a step up, but the driver support (IIRC) could be a bit spotty. But overall it was great. Then XP came a long and added Crayola UI and Zip folder access. It took me a lot longer to move to XP, and when I did I would always go to Control Panel and turn off the dreadful UI and make it go back to using the classic desktop style. But XP stayed relevant right up till 2013 when I changed jobs and no longer needed to use VMs as much... even then there was still one VM I needed to use that was supplied by head office. So XP for me was still in use when they EOL'd it, though only to run a legacy simulation in a VM.

W2k was the first possibly-not-utter-crap version. In the consumer space W3.1 was just DOS with Windows, and didn't even have full multitasking.

W95 was 60% bugs and 40% OS. There was even a really obvious bug in the calculator app. There were all kinds of very very weird - i.e. bad - design decisions both in the code and in the UI. Trying to find fixes and workarounds literally wasted days of almost everyone's time.

The initial release of XP was workable, but by SP1 it was fairly stable and you could actually use it as a working OS without having to deal with driver issues and constant crashes.

Vista was another dog. W7 was very possibly the best Windows. 8 was another WTAF release. 10 is 8 with some improvements, but some added evil in the form of forced telemetry for most users, and such.

This is because since at least ‘95, MS has alternated focus between base technology and polish/reliability/UX. The ones we think of as good (98SE, 2000/XP, 7) are the ones with the focus on polish. The ones we remember as flaky (ME, Vista, 8) are the ones with substantial new tech in them that hadn’t had the additional years of refinement.

I think Windows Me was just an attempt at making 98 more modern and multimedia-friendly. It was just bad, with no relevant upgrades at any layer of the stack.

Yeah, actually let’s not talk about Windows ME. :P

i think it was the only still dos-based windows to automaticall install usb drivers for mice and usb sticks and things. we had a sony laptop running ME for a long time because it was too slow to run XP, but had a USB port already

What was the new tech in ME ?

For me, Windows 95 OSR2 (OEM service release 2) and 98 were more stable than stock 95. I remember having to obtain it via one of those "100 in 1" pirated CDs, because the OSRs weren't made available to the end users.

> I still think 7 > 10

Me too, for a simple reason: Windows 7 can be made to look and behave almost identically as Windows 95 did in terms of UI - which is, to this day, widely praised: https://news.ycombinator.com/item?id=21878006

Windows 8 and 10? Not at all. It's a wild, wild mixture of the good old UI and... an abomination of stuff clearly "designed" for touch screens. Not to mention all the privacy and advertising issues.

7 was just rebranded vista. It was basically vista sp3. MS knew the vista name had been badly tarnished.

The are several reasons why vista churned so badly at startup was the indexing service and the service that was trying to load everything into memory. Turn those 2 off and vista worked as good as 7, I did this to dozens of computers and the owners were so happy. Several other background services needed about 1.5GB of memory just to get the system up and running. This was when 4GB of machine was considered 'huge'. By the time 7 came out they had fixed both of those services and a 'crappy' machine had 4GB. Dumping it on a box that 512MB of memory and it struggled badly (512-1024 was the sweet spot for XP). 8GB for Vista and up and for proper use of that, you need the 64bit ver.

Windows 7 STILL does this in fact.

I am currently using Windows 7 in my personal machine (and Linux in others).

The indexing service often gets disabled because it makes the machine slow down a lot.

Also I really miss Linux "swapyness", Windows Vista+ (10 included) are really aggressive in swapping, more than needed, often my machine ended trashing because of Swap while still having 20+ GB free, this is in part because of that stupid indexing and cache, Windows 7 has a habit of giving priority to the cache, putting files from the most accessed programs in the RAM, and not caring if your CURRENT program need RAM...

I use the indexer setup to ignore parts of my system. That works pretty good. Especially if you are doing any work with nodejs. Also defragging just the index file helps a lot. That thing can end up with thousands of bits all over the drive. Even with SSD it is not great as random seek on many SSDs are fairly terrible and contig read is amazing. One trick I also do is pre-allocate my swap file (I usually pin it to 4GB). As the default is to grow/shrink.. That can do the same thing as the indexing file.

Something also changed around sp2 with 7 and file writes. Files get fragmented very quickly now even when there is plenty of space not to do so. I usually would only end up with systems badly fragmented in XP if the drive is full (much like ext4) now it just does it as a matter of course and is acting like the old DOS alg of find the next free spot.

You can still turn off the readyboost/superfetch cache service (not sure where it is win10). If you have an SSD you should not notice much difference on or off but it can happen. Turning those off makes your free RAM act more like in linux where it just caches recently read items and gives it up right away if needed.

When Vista came out I bought a new machine with AMD A6 4600+ and 2GB of RAM with Vista 64bit and it ran great. I had no driver issues or performance issue even with default install. You didn't need 4GB of RAM, you just needed new hardware. Limited RAM and drivers for the older hardware were the biggest problems, Microsoft drastically changed the driver DDK and many hardware venders had poor quality or no drivers which led to many of the problems people had. Kinda sad I got rid of this machine, it would be interesting to benchmark it to a Raspberry PI 4 in Linux.

For my use case 2GB was borderline. 4GB was the sweet spot with Vista. 8GB was great but on the pricey side.

I'm convinced if they had set a higher minimum system requirements and hid the pre-loaded memory from the task manager or labeled it more clearly Vista would be a lot more fondly remembered. Most of the people I met with negative experiences were running it on walmart "vista ready" computers that should have never certified in the first place or upgrading their old pentium 4 machine with 1GB of ram and expecting a good time.

And the 5400RPM drives most of them had. There was a lot of clearing out of old stock going on. Also sp1 fixed a lot of issues too.

Whether you call it an SP or not, Windows 7 rolled back the UI mess of Vista.

Funny enough win2k was the first to have a somewhat decent multi-language management, in particular you didn’t need a dedicated CJK version. So that’s where the nostalgia loop would stop for a lot of us (and it was buggy as hell, of course)

That doesn't prove anything. People said it was crap because it was. Then SP1 was released with 1+ year and it became pretty good. By SP3 it was rock solid because that's 7 years of bug fixes keeping other changes to minimum. That's what giving bugs does to software, no surprise here.

Microsoft made a big security pivot during XP. Gates put his foot down and forced the company to reprioritize secure engineering over everything else during the SP(2?) timeframe.

They also made a pretty big push toward formal verification of drivers, which were often the culprit for insecure/unstable machines. Their SLAM model checker [0] was avant-garde (in industry terms, at least).

[0]: https://en.wikipedia.org/wiki/SLAM_project

Yep, the whole Trustworthy Computing initiative: https://en.wikipedia.org/wiki/Trustworthy_computing There was a huge investment in training, tools, and creating security specialists. It took a long time to turn the ship around but security issues started trending down a few years after that.

... I'm sorry about typos

with 1+ year -> with 1+ year of bug fixes

giving bugs -> fixing bugs

It's worth mentioning that WPA2 for WiFi doesn't work in XP unless you install SP3, regardless of the WiFi drivers you use.

As this leak is only XP SP1, it will be missing a lot of the bugfixes and upgrades that XP added by SP3 (and beyond).

Think of how easily it will be though to reverse engineer the patches in the SP and apply them to the actual source.

Don't forget Windows 2003 Server in that list. That, to me, was the ultimate Windows. Had the XP stuff, but with the ugly theme disabled by default. Was super stable even compared to Windows 2000. Was lightweight. Only annoying thing was it came with sound disabled, but a few clicks in the Control Panel after installing, and you were golden.

That theme was wonderful in gray/olive variant if you ask me. It looked polished.

And contrary to consumer versions, free trial was available!

And, you could also get a free license through university (MSDN AA). That's how I got to run mine.

Amen. Windows 2003 Server was perfect.

XP was a bit crap in the early days, but by SP2 it was a lot better and by SP3 it was really solid.

It is quite sad that we might never experience that stability and usuability of commercial big apps/os anymore in the age of penny squeezing and spyware.

I had my Win XP, Photoshop and MS Office with ever lasting licsenses. You need FOSS today for that. As soon as you pay for software nowadays you are a loser, in general.

>As soon as you pay for software nowadays you are a loser, in general.

That is so true! I buy'd a Lifetime License for Insync and Jaikoz, Insync's lifetime was 4 Years (then they changed the license for updates) and Jaikoz "Lifetime" was 6 Years, change the name of the Software and split it in two different product and the 'old' one was not maintained anymore.

Windows XP SP3 was solid as hell. It was easy to change the theme, which I always did along with the brain dead "hide extensions and system files" and ridiculous swap file settings that guaranteed you'd always be resizing incessantly

I personally tried to use Windows XP (for my Windows systems, which is mostly VMs nowadays) for as long as I could.

First, it was the first solid (in relative terms) consumer operating system from MS.

Second, it was lightweight (in relative terms) in terms of UX and, somewhat consequently, engineering. With Windows Vista (but in particular, Windows 8), MS started to add complexity on every layer.

Nowadays, I'm really afraid to update Windows. I don't know what new services and programs are going to be added. I don't know if the tweaks still work. I don't know if I'm imposed some new behavior. I don't know where certain settings (even entire setting domains) are going to be located, and how they're going to be named.

In a way, Windows XP is the Windows equivalent of the Linux XFCE/LXDE desktop environments.

I think a lot of people have sort of backed their chair away from the table like you.

I remember reading Matt Ridley's The Rational Optimist and it talked about trade. Basically it said with trust, trade is unlimited.

But what's been happening with computers and software is that trust been going downhill so quickly that many people don't install anything ever anymore. Privacy, dark patterns, monetization, form over function, qa by customer, and just a general lack of empathy, respect and common sense.

It's a shame.

i think you are onto something, but i don't think the market cares that much about privacy, dark patterns etc

also see how software has stabilised over the years, both as a product and as a marketplace. there are now clearly defined best in class software that will take years to be disrupted.

but one of the major aspects against computer software has come from mobile. lots of development time has gone to browsers and mobile than desktop.

I consider XP to be the first instance of Microsoft tightening the screws, as it was the first version to require activation, with the potential for hardware upgrades to require reactivation. Even then, I thought that it would ultimately lead to locked-down bootloaders and other such tomfoolery, and we're now most of the way there, or, indeed, already there for mobile.

I think you're talking about power users, programmers, and IT. Regular everyday users still don't care and install whatever they like.

and parents.

> First, it was the first solid (in relative terms) consumer operating system from MS.

That was actually Windows 2000, based on the NT 4 code base. It was near perfect, given the era it was made in.

The only reason I upgraded from Windows 2000 was that Microsoft was extremely reluctant to back-port USB- or WiFi-fixes and support, effectively forcing you to upgrade to XP if you wanted to fully use your modern hardware.

For that reason alone, XP always left a sour taste in my mouth.

> That was actually Windows 2000, based on the NT 4 code base. It was near perfect, given the era it was made in.

I also remember Windows 2000 fondly. It was really really stable. Even if it looked like it's frozen like a polar ice cap (they weren't melting that bad back then), a short coffee break would give it enough time to normalize itself and continue like nothing happened.

Should retry it in a VM sometime. :)

I actually ran a series of them in VMs for years because of their quick response and light resource usage, and they still ran all the stuff I needed. Loved that OS

I have a Windows XP VM to run the software for my MD player. You're right, the OS from that era is lightning quick in today's hardware, even in a VM. Also, the USB passthhru is problem-free interestingly.

I stuck with Windows 2000 for many years after XP was released. XP brought absolutely nothing I wanted and a lot of things I didn’t want. All the truly interesting stuff, like WinFS, never made it in.

WinFS wasn’t destined for Windows 2000 nor XP, but rather Vista.

I was an intern at Microsoft in 2004, and personally helped make it clear that WinFS also wasn’t ready anyway. Fwiw, I think the iOS model of “makes users and apps no longer care about directories” has been more successful than the original WinFS plan of “maybe it’s all a SQL-ish database instead”.

>It was really really stable.

Windows 2000 was the last version of the NT branch where Dave Cutler was in charge. He's a legend for a reason.

https://www.youtube.com/watch?v=J2GV_bCfnCw

Computer History Museum long form interviews:

Part 1 https://www.youtube.com/watch?v=29RkHH-psrY Part 2 https://www.youtube.com/watch?v=SVgSLud50ss

Official PDF transcript:

https://www.computerhistory.org/collections/catalog/10271716...

One thing I read when dicking with how USB drivers work is that the Microsoft's first USB driver model was hopelessly broken, so they started over. And that also turned out badly enough that they redid it again. That's what ended up in XP and it mostly worked okay.

You can see why they didn't want to/couldn't backport USB drivers.

Can you provide a source? I love to read that kind of stuff.

I think it's a dead tree source. Might still have it.

Thanks! I like (reading) dead trees tho. I might find a copy or someone might have scanned it.

Edit: Comment made me look like an enemy of forests and trees so, clarified the situation.

Windows 2000 was the first rock solid OS from Microsoft, targetting both desktops and servers,but the desktop version was targeted at business users, not consumers. The consumer version at the time was the horrible Windows ME.

Windows NT was -very- solid on good hardware. Would stay up for months with zero issues.

Ha! I had the same with windows 2000 as compared to NT4. I stayed on NT4 as long as I could, and never really liked windows 2000 because it took twice the resources to do the same job at half the speed. Windows XP was 2000 with more (ugly) eye candy and a lot of bloat, so I didn’t like that either, but hardware improved enough that by the end of its reign it was quite fast.

i changed my windows 10 color scheme to bright red so I always feel like I'm in danger.

To remind you of how much telemetry they're sending out about you to God knows who? :)

I still have a desktop running windows XP that my parents refuse to upgrade, partially due to a piece of software licensed per computer that costs >300$. Thankfully chrome no longer opens half the websites so they are limited on the web side at least. 14+ years of service, the only times it blue screemed were due to hardware breaking down. Rock solid.

Did you try to virtualize it and run it into a VM only for that app?

This indeed

No I'm not touching it with a 10 feet pole. Besides the fact that I would need new hardware for that( it's a core 2 duo system) which they are unwilling to fork the money for, anything that happens with it in the future will be my fault.

But it would be a time when the HD dies. Are you able to re-install? If no, you should try to backup/virtualize.

No I'm not able to reinstall. I am also not willing to deal with tech support. If and when it dies I'll just order a dell for them and they can sort the software issue themselves.

> First, it was the first solid (in relative terms) consumer operating system from MS.

I used Windows 3.1 for years and I never had to reinstall it for stability reasons. At some point I also switched to XP and after 2 years it wouldn't boot anymore because of a blue screen appearing before the graphical interface would show up. That made the decision easy for me to use Linux exclusively - which I had on dual boot already for a considerable time.

To me Windows 95-XP were really frustrating to use since random parts of the system (CD-ROM, Word, Games...) would stop working at random times. Or for instance on Windows 95 sometimes the system would only boot if I opened and closed the (empty) CD tray.

At work I sometimes used Vista and 10, I think they are okay but macOS and Linux are far more fun.

Before committing to GNU/Linux in 1998 or 99, I've had a similar experience with moving away from DOS/Windows for workgroups 3.11 to '95: Windows 95 tried to achieve too much in terms of backwards compatibility between 16- and 32-bit software, and was very flimsy as a result.

However, I've briefly used NT4 as my "development box" (Delphi addict for a while, though in my defense, I was like 15 at the time) for 6-12 months, that was as stable as anything you could get. But it had the same problems as Linux: it was very specific as to what hardware it supports. The same "soft modems" (called "winmodems" at the time) were equally unsupported in NT4 just like in Linux, so you had to get yourself external hardware modems which were like 5-10x the cost (whatever came out of US Robotics?).

So since I was more into the hacker mindset, I moved into GNU/Linux world and never looked back.

But I did grow an appreciation for MS engineering: they can develop quality software, and did it even in 96, but their consumer OSes strive for unparalleled compatibility and allow anybody (literally anybody) to write system level code that runs on the kernel ring level due to that.

A couple years down the line, I've even seen an article how many workarounds they had to introduce into '95 to allow old software employing their own extended/expanded memory management tricks to continue to function (think stuff like DOS games, WordPerfect...).

For all the fun we make of it, Windows 95 was a marvel of engineering when you consider the constraints and requirements. Just oppose that to the new web app approach to software engineering: breaking backwards compatibility is such a common theme that nobody bats an eyelid.

I admire what the community has achieve with Linux. I think is really great and I have used Linux in the past (Redhat, Fedora, Suse, Ubuntu, Mint).

However, I always wonder what makes your experience so different than mine.

For me I can simple replace what you said and it will still be valid:

To me Linux were really frustrating to use since random parts of the system (CD-ROM, Word, Games...) would stop working at random times.

(well the part about the CD tray never happened to me with any OS).

I really like the customizability of Linux and such, but in the end, Windows 2000, XP, and 10 just works 99% of the time. At least for me who I consider an advanced user.

Of course there are things I don't like, like telemetry, but the pros vs the cons, at least for me, make me always return to Windows.

I have also worked with macs and for 4 years my wife had a MacBook. Not only do I find the prices crazy expensive, compared with ThinkPads, but I never found anything that was better in OSX than in Windows.

Again, I can use macOS if needed be, and I find it just fine. But I prefer Win/ThinkPad.

Of course this is my own opinion.

For me the main difference between (early) Windows and Linux is, Windows is quick to setup and quick to break. Linux is hard to setup and hard to break - I cannot tell how much time I spent tweaking my Linux desktop. Also I could for instance use an old SCSI scanner on Linux much longer unlike Windows where they eventually stopped developing drivers for new Windows releases.

But now things are of course a bit different, especially when choosing well supported hardware. On the other hand driver support tends to be more stable and long-term oriented on Linux since they tend to be maintained centrally in the Linux kernel. I agree that macOS is a bit mediocre in some sense, but somehow it unites the worlds of Linux and Windows.

It could be a matter of experience. If you are more experienced with Linux, it could be easier than if you are more experienced with Windows.

For me, Linux is easy to install, and it has been for a while, but then trying to customize it to do what I want, is much more complicated than Windows.

For example, I want RAR support. There is no LinRAR, so I had to download a library and install it. It wasnt that bad, but in the end I ended up with a tool that was not as simple to use as Winrar.

Want to use Total Commander? Either install Wine or run Midnight Commander or some other similar program that are not up to par with TC.

Want to use Office? Wine or Crossover. Want to use Illustrator? Same.

Need to install VMware? In Linux you have to import some signing key first and do some additional steps (don't completely remember). Nothing major, but you need to find out first how to do it. In Windows, just download the installer, double click and follow the wizard.

For me it seems like it is always more complicated to do simple things in Linux, and you end up with some not as good solution.

Also, I don't understand how can I download an installer from Windows XP and will 99% of the time install in Windows 10. However, if I want to install something old in the latest version of Linux, it won't install, will have to fix dependencies or similar.

And for me the worst thing is that energy management for notebooks is not exactly the same. If my battery lasts 4 hours in Windows, it lasts 3 or 2 in Linux.

Drivers it could be what you say for old ones where the manufacturer did not update to Windows 10, but if you have a supported device, usually the support is better on Windows.

Of course there are advantages, like no telemetry, no antivirus, but for my use case, I found the experience worst.

I am not saying totally worst. Maybe it is a 10 to 20% worst.

Of course, I have used maybe 90% of my time Windows and 9% Linux and 1% MAC, so maybe I am just more experienced.

A lot of it has to do with experience. Linux is definitely rougher around the edges than Windows and OSX, but it also allows you to be free to do what you want with the OS as a hacker or programmer. Also I don't like being tracked while I'm on my computer. OSX is much better than windows in that respect so I'll throw that out there for the record. Windows tends to be more opinionated and tries to bind you to it's work flow and update schedule. On linux you can have your system update everyday or once a year; not recommended but if you're airgapped it doesn't really matter. I choose freedom over walled gardens. I still play games on windows though, but that's the only reason :)

The issue there is you're trying to run Linux like it is Windows. If you do that then you're never going to have a good experience. For example if you're after Windows software on Linux or insist on downloading software rather than using the package manager then you clearly prefer the Windows-style workflow. Which is fine -- everyone has their own preferences. But if that is your preference then the issue isn't Linux doing things wrong, the issue is you just prefer Windows.

For what it's worth, I equivalent teething pains when using OS X for a while. It took me a few months before I finally learned to adopt Mac-isms and I still don't feel at home on it like I do on Linux and BSD (my preferred platforms). Windows, however, always felt somewhat alien to me (and not through a lack of experience, I used to be a Windows developer and have written some pretty low level software for the platform. But even with all of that experience I still couldn't see eye to eye with the OS).

By the way, there are Linux-native builds of WinRAR.

Thanks. I know what you mean.

In the case of downloading vs package manager, the problem is that there is lot of software that I use that is not on the repositories. If everything was there, I have no problem on using 100% package manager.

Regarding the apps, there off course the issue is that some apps have no equivalent in the Linux world.

There is no MS Office, LibreoOffice is good but not so good, and same for Adobe Suite. Total Commander I can more or less replace with Double Commander, or TC under Wine.

Afaik, there is a unrar.dll equivalent in Linux, but not a proper WinRAR.

Thanks for your opinion!

If you buy a system made with known linux compatible hardware you'll have zero problems. Just like a mac. Windows has the advantage that it gets first tier service and thus a greater chance with being stable with a broader swathe of hardware. OSX to me is much better than Windows because it doesn't get in your way. You want a new app download it and drop it in Applications. I don't know WTF windows does but some installs take forever; on Mac OSX and Linux install times are so much more reasonable and transparent. Obviously it matters WHAT you do with the OS, for some situations only Windows will have the software you need, and on those occasions I reluctantly use it. As far as stability on good hardware, for the average user, there's very little difference in the 3.

I have always used ThinkPads, which are among the best supported notebooks for Linux afaik.

I totally agree that Windows has an "unfair" advantage here as manufacturers make sure that their hardware works in Windows while for Linux most of the time they don't even provide drivers.

I have helped friends and family configure their MACs and honestly I never felt wow this does not gets in the way. Not all apps in MAC are installed the same way IIRC. For example Adobe apps. Never had issues with installers that take forever in Windows. Same with Transparency.

About the availability of the software, 99% of it is on Windows, so that is a big plus. But I am not talking so much about that, I am talking about the complexity to do things in Linux.

Most of the time, at least for a user like me that tries it/uses it sporadically, I need to search for a way to install or do something. It is good that most of the time there is a response, but I find it weird that all is so complicated.

For example, this is how I installed virtualbox some years ago (it might be better now)

wget -q https://www.virtualbox.org/download/..._vbox_2016.asc -O- | sudo apt-key add - wget -q https://www.virtualbox.org/download/oracle_vbox.asc -O- | sudo apt-key add - echo "deb http://download.virtualbox.org/virtualbox/debian xenial contrib" | sudo tee /etc/apt/sources.list.d/virtualbox.list sudo apt-get update sudo apt-get install virtualbox-5.0

In windows it was download, double click, next, next, next. Done.

Here is how to install MKTOOLNIX (again, a couple of years ago)

wget -q -O - https://www.bunkus.org/gpg-pub-moritzbunkus.txt | sudo apt-key add - sudo sh -c 'echo "deb http://www.bunkus.org/ubuntu/xenial/ ./" >> /etc/apt/sources.list' sudo apt-get update sudo apt-get install mkvtoolnix mkvtoolnix-gui

In windows, download, double click, next, next, next.

Again, if you use it all the time, maybe it becomes natural to you.

That's not really a fair comparison though because most of those stages are just part of the download stage you glossed over in Windows (I really hate having to browse the web looking for software then having to manually download the thing after I've found what I want).

Also Windows has no managed way to keep applications up to date. Which means you either have to manually repeat all those steps yourself or run background agents for every publisher to keep that software up to date. But by far the most annoying in Windows is software that doesn't check if it's up to date until you run it and then demands you update it there and then (great way to kill productivity). Yet those steps you've exampled is a one time pain and then all software is managed for you. Updates are easy.

I used to be a hardened Slackware user and was prepared to download and often compile my software manually. Then I discovered package managers in Linux and from that point onwards I lost any respect for an OS that didn't manage software for me. It's amazing that it's two decades from me discovering package managers, ten years since iOS App Store was announced, and yet people are still stuck on shitty first party solutions on Windows and macOS (thank god for homebrew et al).

Again, I find incredible cool what the Linux community has accomplished. And yes, the update process in Linux is much more simple in theory (sudo apt update). However, at least for my use case, the pros do not outweigh the cons.

Firs cons: lack of key apps (Microsoft Office and Adobe Creative Suite).

Second cons: lack of OEM produced drivers, which means that some things work worst than in Windows. Example, energy management for my ThinkPad.

Third cons: it seems to me that most things are more complicated to do than in Windows. Case in point, even if we do not consider the wgets, installing mkvtoolnix and vmware.

Fourth cons: Why do I need to look for the specific distribution version of a program? I can install a Windows XP program in Windows 10. I can even install a Windows 95 program in Windows 10. But I have to look for the specific version of Ubuntu 20.08 of a program. The version for Ubuntu 17.06 would not work.

Again. I can for sure use Linux if needed to (I am currently trying BSD on my ThinkPad), but if given the chance, I feel much more comfortable and productive in Windows.

> Firs cons: lack of key apps (Microsoft Office and Adobe Creative Suite).

Propitiatory commercial applications, sure. But there's plenty of open source counterparts. I've been using LibreOffice (and OpenOffice before then) for years and while it does lack a lot of polish compared to MS Office most of the issues with it are really just cosmetic. Plus I never liked the ribbon bar anyway. Adobe Creative Suite is a bigger problem. I have heard of professionals uses open source counterparts and getting on well with it but the last thing a person wants is technology getting in the way of their creative process (a sentiment I'm intimately familiar with from a music composition perspective).

Ultimately though, if you're unwilling to try changing muscle memory to use open source applications then Linux is never going to be the right fit. The problem there though isn't that it's not possible to do something on Linux but rather than you just more comfortable with Windows. And that's fine -- but it is literally just habit rather than a technical reason. > Fourth cons: Why do I need to look for the specific distribution version of a program? I can install a Windows XP program in Windows 10. I can even install a Windows 95 program in Windows 10. But I have to look for the specific version of Ubuntu 20.08 of a program. The version for Ubuntu 17.06 would not work.

> Second cons: lack of OEM produced drivers, which means that some things work worst than in Windows. Example, energy management for my ThinkPad.

I honestly prefer Linux for driver support. Windows was such a pain in the arse with having to install 3rd party drivers (though that's less the case these days) where as on Linux everything just worked with the default install.

That said, you do sometimes find edge cases where drivers are missing functionality but it's not all that common.

With regards to battery life, I a lot of that will also be down to tuning. Windows ships some pretty aggressive tuning features where as Linux basically ships untuned. You can do that manually and it is a bit of a joke that users have to do it manually. But I've found losing an hour or two off a 10 hour run time isn't all that painful.

> Third cons: it seems to me that most things are more complicated to do than in Windows. Case in point, even if we do not consider the wgets, installing mkvtoolnix and vmware.

I actually find the opposite to be true. Most things are more complicated to do in Windows. Developer tools, programming language support, containerisation/docker, debugging tools, networking tools, automation (this is by far the worst thing about running Windows for me), resolving issues (far easier to fix a broken Linux install), documentation (far better documentation in Linux), error reporting, etc.

Updates are a key thing too. Windows: you need a thousand different update managers running on start up. You need to regularly reboot due to Windows update (which forcefully locks you out of your system whether you want it to or not). Linux: everything is managed from one package manager and no reboots required (unless it's a kernel update but even then it installs it without locking you out and the reboot is just to load the new kernel)

Another bugbare is having to browse for applications on the internet, avoiding phishing sites for popular software, then manually downloading and installing it. Repeatedly for some applications that don't have update managers too. Users shouldn't be doing this in 2020 -- it's just a crap way to manage software.

> Fourth cons: Why do I need to look for the specific distribution version of a program? I can install a Windows XP program in Windows 10. I can even install a Windows 95 program in Windows 10. But I have to look for the specific version of Ubuntu 20.08 of a program. The version for Ubuntu 17.06 would not work.

That's why you use package managers ;)

> Again. I can for sure use Linux if needed to (I am currently trying BSD on my ThinkPad), but if given the chance, I feel much more comfortable and productive in Windows.

BSD is going to be much more painful for you than Linux given the complaints you're making about Linux. You're better off trying a Mint Linux. Or if you don't mind the one time pain installing Arch then go with that (Arch has no installer, you have to do it all manually. Which sucks for non-techies but once you're over that hurdle it's probably the most painless OS I've ever ran over long periods of time)

I don't rate Ubuntu much as a distribution. Canonical spend too much time reinventing the wheel and not enough time fixing the common complaints with Linux (though `snap` is an attempt at that but unfortunately even there Canonical's solution is hamfisted).

Ultimately though, it all boils down to habits and personal preferences rather than X being better than Y.

Let me start by saying that I don't want to convince you to use Windows ;) What I am trying to do is show how it all depends from where you are coming from, where your experience is at.

> Ultimately though, it all boils down to habits and personal preferences rather than X being better than Y.

I have no problem changing systems if I see the value/advantage. I have changed systems in the past. Started with Ti994a, DOS, OS/2, Windows 95, Windows 2000, Windows XP, Windows 10. In the middle I tried Linux, may variants: Mandrake, Redhat, Suse, Gentoo, Fedora, Ubuntu, Mint, Manjaro, and many others just to see how they looked/worked. I even worked with AIX and Solaris at work. What I mean is that I can adapt and work if needed be. My main point is that I don't suffer all of the pains, with the exception of telemetry, that Linux users have against Windows.

> Propitiatory commercial applications, sure. But there's plenty of open source counterparts. I've been using LibreOffice (and OpenOffice before then) for years and while it does lack a lot of polish compared to MS Office most of the issues with it are really just cosmetic. Plus I never liked the ribbon bar anyway. Adobe Creative Suite is a bigger problem. I have heard of professionals uses open source counterparts and getting on well with it but the last thing a person wants is technology getting in the way of their creative process (a sentiment I'm intimately familiar with from a music composition perspective).

I don't care if it is commercial or free, proprietary or open source. I care that is the best software for the job. That said, Office has become easier as I can use the Online version or Gsuite.

> I honestly prefer Linux for driver support. Windows was such a pain in the arse with having to install 3rd party drivers (though that's less the case these days) where as on Linux everything just worked with the default install. That said, you do sometimes find edge cases where drivers are missing functionality but it's not all that common. With regards to battery life, I a lot of that will also be down to tuning. Windows ships some pretty aggressive tuning features where as Linux basically ships untuned. You can do that manually and it is a bit of a joke that users have to do it manually. But I've found losing an hour or two off a 10 hour run time isn't all that painful.

The problem is that I don't see the point of expending a week trying to fine tune Ubuntu, or whatever version I want to use, when I can just install Windows 10 and it works perfectly out off the box. At least for my use cases.

> I actually find the opposite to be true. Most things are more complicated to do in Windows. Developer tools, programming language support, containerisation/docker, debugging tools, networking tools, automation (this is by far the worst thing about running Windows for me), resolving issues (far easier to fix a broken Linux install), documentation (far better documentation in Linux), error reporting, etc.

I never found using developer tools, programming languages, debugging tools, networking tools, are more complicated in Windows. Same for resolving issues (I had to reinstall Windows very phew times, maybe one or two because I couldn't fix something and it was easier to just reinstall). I don't use containers/docker, so I cannot comment there.

> Updates are a key thing too. Windows: you need a thousand different update managers running on start up. You need to regularly reboot due to Windows update (which forcefully locks you out of your system whether you want it to or not). Linux: everything is managed from one package manager and no reboots required (unless it's a kernel update but even then it installs it without locking you out and the reboot is just to load the new kernel)

This one I totally agree. Linux is much better at updating than Windows. No discussion here. The main problem is when there are non-official repositories.

> Another bugbare is having to browse for applications on the internet, avoiding phishing sites for popular software, then manually downloading and installing it. Repeatedly for some applications that don't have update managers too. Users shouldn't be doing this in 2020 -- it's just a crap way to manage software.

I don't see it as a big problem. I know the apps I want and I can just install them. But I understand how this could make life easier for some.

> That's why you use package managers ;)

The problem is that not all apps are on the repos ;) In my experience if the repo has the app, and it is installed regularly, all fine. If it not updated or does not exists, it is a pain.

> BSD is going to be much more painful for you than Linux given the complaints you're making about Linux. You're better off trying a Mint Linux. Or if you don't mind the one time pain installing Arch then go with that (Arch has no installer, you have to do it all manually. Which sucks for non-techies but once you're over that hurdle it's probably the most painless OS I've ever ran over long periods of time)

I don't mind. I tried OpenBSD, NetBSD, and DragonFly. I have OpenBSD running. As a desktop OS I don't see the advantage over Linux, much less over Windows.

> I don't rate Ubuntu much as a distribution. Canonical spend too much time reinventing the wheel and not enough time fixing the common complaints with Linux (though `snap` is an attempt at that but unfortunately even there Canonical's solution is hamfisted).

Don't get me there :)

Anyway, I always have a Linux laptop running around. I currently have Manjaro, but I will be trying Tumbleweed soon. I would love to one day totally move to Linux. I am not loosing hope.

Good discussion. I enjoyed this. Thank you.

I don't really much else to add aside that the boundaries these days are definitely a lot more blurred due to WSL (bringing many of Linux's tooling to Windows) and web applications (levelling the playing field somewhat in terms of application support). Both of which are largely a good thing for users.

Thanks again for the conversation. These topics often end badly and it's been nice to have a discussion like this without any of the flamewars :)

Same to you :)

Yeah but this is like comparing sausages to coconuts.

Can you open a coconut with your bare hands? Now go and tell vegetarians that sausages grow on trees, because they do https://en.wikipedia.org/wiki/Sausage_Tree

For me that actually was Windows 2000. Never really saw a need to switch to XP, and instead switched to Vista and later 7, eventually.

XP, when it was first released, was a more resource hungry and uglier version of Windows 2000.

Win 2000 was amazing when it was released. Games (albeit not DOS ones) worked well, it was stable and really your only other options were worse (Mac OS 9 was less stable than Win 98, ME was a dumpster fire, Linux wasn't quite ready for the desktop and BeOS was awesome but barely supported).

Then when XP was released, it had twice the minimum and recommended hardware spec than 2000, it had those stupid skins and it didn't offer anything functionally more beneficial over 2000 (aside faster boot time if you had hundreds of fonts installed).

Granted XP got better with SP2 or SP3, but it definitely wasn't the first solid nor even lightweight when it was originally released.

My migration from 2000 to XP meant blue screens. Most likely caused by drivers. But that left me with a lasting impression that 2000 was way more stable than XP.

It was largely the same kernel. Issue was largely because drivers for retail hardware weren’t written well enough.

It wasn't. It was just as stable as 2000 as long as you weren't a tweaker and stuck with stock. Almost every single OS becomes unstable with shitty drivers and/or cheap hardware that wasn't designed for it.

Yup. When saying "XP was great", people mean "XP SP2 and on" - for me, the move from W2k to WXP was "the same thing, now with Fischer-Price aesthetic!"

> Fischer-Price aesthetic

That is the perfect description. It’s one of the ugliest UIs I’ve ever experienced.

XP also shipped with the 2000 skin and it was trivially easy to switch over.

Yep. Even switched it over when you went to the lower resources usage settings.

It was easy to switch but there is a lot to be said for sane defaults. With each new iteration of Windows I'd find more and more insane defaults (for me at least). It started with file extensions being hidden by default. Then system files too. By the time XP was released I was finding that I had to spend an hour just configuring the system how I liked...and that was before I installed any software.

Users will sometimes reach a point where they say "If I'm having to spend this much time configuring new installs to function how I like, then I might as well install something else that already ships with the defaults I like". And this is why many power users started drifting away from Windows in the decade of 2000 to 2010.

Windows Server having file extensions turned off by default still drives me crazy. I get it for the desktop versions, but it would be fantastic if server editions had it on, or it saved and picked up the toggle from your AD account/profile.

I agree generally about the importance of sane defaults, but the theme was just soooo damn easy to switch, and a lot of non-techy people I know did so.

I actually don’t really associate Windows XP with its theme in my memory, because I so rarely saw it in use.

People said that (as well as Skittles OS) since it launched.

I have to stick up for XP because i have so many fond memories from my young adulthood of using it. I never had a problem with it's default theme but it also shipped with a few other themes that were better. like olive and then later the zune theme which was arguably the best one.

I say this as somebody who really enjoyed W2K, but the complaints about WinXP's default theme are very difficult to understand when:

- The classic theme was just a few clicks away

- Of all the problems with XP, I would not even list the "easily changed default theme" in the top.... 100 or so. I would not expect to see such sentiment on HN, of all places.

Hey man, let people complain about what color the bike shed was painted.

XP was a pretty damn good OS that came out of a time when there wasn't a lot of options.

It was pretty stable. You could change network settings without restarting. It was built on the NT core instead of whatever 95/98/ME hot mess was.

The main issue was it's exposure to viruses. Which was address in 8+ (windows actually started acting more like linux with regards to user/admin).

Windows7 IMHO had the best ui out of the box. 10 grew on me. Let's just not talk about 8.

> XP was a pretty damn good OS that came out of a time when there wasn't a lot of options.

That's more true of Windows 2000 though. By the time XP was released Apple had released a few iterations of OS X and it was looking pretty decent. Not to mention desktop Linux becoming stable. Neither of which were true in 2000.

In fact with Windows 2000, I used to dual boot Win 95, Linux Slackware and BeOS 5 but I always came back to Windows 2k for day to day work. Then when Windows XP was released I gave it a try and got so fed up with it that rather than installing Win2k again I just switched to Linux full time instead (sadly BeOS had ceased development by that point).

The look was superficial, sure: For someone moving from Win2K to XP, this was not the main frustration, but a easily recognizable sign for the myriad other issues (note: original XP, XPsp2 was the release that should have been).

(We did some printing. In the range of 1M pages/workstation. It...worked...on XP. Eventually.)

Many games did not work so well on 2K, and XP brought some substantial improvements specifically in that department.

W2k had a w98 compatibily mode which could be enabled from the command prompt.

I hear that said a lot about Windows 2000 but my experience was that it was only really true of legacy DOS games that didn't work right. Anything written for OpenGL or DirectX worked fine. In some cases even better (for example Quake 3 ran so much smoother on my BP6 board running Windows 2000).

I remember several of my Win9x games didn't work fine, but not which ones. I think Carmageddon 2 was one?

There were plenty of games for 9x that assumed its lax memory and permissions model, especially for stuff like DRM.

Totally my experience as well, Windows 2000 really was the only Windows version I actually liked, it had very few faults for its time. Windows 7 probably comes second.

What I remember from XP is that it took until SP2 before all the problems were worked out, and SP3 actually introduced more annoyances in the name of security. I also think the default theme and general visual style of the XP is downright terrible and completely unworthy of a large professional company like Microsoft. The very first thing I always did was to switch back to the classic UI for everything (theme, start menu, all control panels, etc).

I don't care for any of the 'modern' Windows versions. They work, behind the scenes the technology has improved, but I they still annoy me in many ways, and the UI is IMO the worst it has ever been in the history of Windows. It looks cleaner but behind the veneer its just a giant heap of historical accidents piled on over the years, with no consistent direction of vision to be found. I get no joy using it, but I guess if you are used to working in Windows the OS is perfectly fine for what it does.

Agreed about the UI. Even at a distance of twenty years, Windows 2000 looks more pleasant and professional. The Windows XP UI managed to be somehow childish in its aesthetic and stressful in its complexity.

Of course every Windows OS eventually gets stable and fast on current hardware, at which point everyone starts praying that they'll get to use it forever and they'll never have to deal with a new major release again.

Agree, Windows XP was sluggish when it was released. Same applies to Win2K though when it first came out compared to WinNT 4.0. 2K was the first Windows to include a shadow effect under the mouse cursor. Oh the extra compute cycles it had to do to render that...

I do agree but don't forget that the difference between NT4 and 2000 was night a day. 2000 wasn't just a reskin of NT4, there were lots of compatibility and usability improvements too. The problem was most people simply couldn't run NT4 outside of an office environment because very few games would run on it, driver support was relatively limited, etc. Windows 2000 was the real bridge between NT office machines and home machines, not XP. Granted consumers wouldn't have seen it that way because most stores didn't sell 2000 on home machines (instead opting for the dumpster fire that was ME) but technology wise XP (original release -- pre-service packs) wasn't a huge advancement from 2000 where as 2000 from NT4 was.

I thought the issues with games on Win2k was that it didn't have an uptodate DirectX version. I recall various hacks to try to get around this after XP showed up but Win2k was always the odd one out for games if memory serves.

Don't forget that a lot of people compare XP not to Windows 2000 but Windows ME which was probably the worst Windows version ever :)

A few clicks and XP SP3 looks exactly like 2k, with a few useful features thrown in.

A decent number of not-ancient floss software supports it as well. Not so with 2k. Add thousands of security patches and firewall, and the choice is clear.

> A few clicks and XP SP3 looks exactly like 2k, with a few useful features thrown in.

There wasn't much added to Windows XP over 2000 in the original release of XP (service pack 3 was released 6 years (!!!) after XP was originally launched. So it's not really fair comparing XP SP3 to Win 2k). I remember this era in computing vividly and it took a few service packs before XP really came into it's own. However you still paid the price running XP because it demanded double the hardware requirements to run when compared with Windows 2000. So sure, you could disable themes in XP but for many people sticking with Windows 2000 was more appealing because you had more preferable defaults out of the box plus less bloat / hardware overhead too.

This is also what I remember, but then I also remember that the specs of my first and last XP machines were miles apart. The first had 256MB or RAM...

Funny that, because Win2k was a glossy, less stable version of WinNT 4 that we had to move to, to get USB support. Had USB been backported to NT4, I never would have moved.

Did you run any specific hardware or something? W2k was rock solid for me from RC2 and onwards. The only thing that really didn't run stable was the drivers for Via's crappy chipset, and they released improved drivers fairly quickly.

No same hardware I ran NT on, although most of my machines were dual processor which was rather rare at the time.

Ah, that could be it then. I've never experienced W2K as anything but rock solid, but then I never had anything faster than a K6-2 back then. :D

Do people not recall that XP, even after SP3, was the OS of over a decade of botnets, ransomware, and worms that crippled the web? Even Vista and 7 were guilty if you were lax and disabled UAC.

All of the things people are complaining about here, like Microsoft becoming the de facto admin of of their Desktop Windows 10 products, are direct results of a period they manufactured.

All anyone who waxes nostalgic should do to test is install anything earlier than 10 on an internet facing computer and then wait. It's terrifying how fast they get compromised.

Always amazing to hear people sing praise of Windows XP when you remember how everyone was praising Windows 2000 when they were calling out Windows XPs insanely ugly themes and designs.

I still have an XP laptop at work because it happens to have a serial port. Every time I use it I'm briefly shocked at how responsive everything is. That's on period hardware.

God the future sucks. IT has been destroyed by complexity fetishists and resume driven development.

Not entirely. There's still Alpine, Gentoo, Void, Debian/Devuan -- All of which still run on reasonably low specced hardware, and can be loaded with XFCE/Tiling WM's for a lightweight UX.

Outside of Linux and other FL/OSS OS's I have to agree though.

You say that as though everything above the Linux kernel isn't a nightmarish hodge-podge of software with interconnections ever increasing in complexity that creates conflicts so readily it needs third parties to carefully maintain specialized builds of each component to keep it all from blowing up.

There's also the FL/OSS BSD's, although I don't know how viable they are for desktops outside of ThinkPad's.

They suffer from the same problem of requiring third party maintainers to keep software working.

I guess I've just given up on having a sole first party maintaining a sane system. It makes more sense to me to have something like Linux/BSD vs a first party system that falls apart worse while still being closed, despite being a consistent maintainer.

Which I guess brings me back around to where you started.

KDE runs in less than 400MB which would be fine on 2GB systems of the era. It's browsers and electron apps (browsers in disguise) that chew up all the memory, and once you start swapping it's over.

I would disagree.

XP is certainly not the proper OS to be running today, it's been passed by especially in a security sense.

But when it was released, it was the first really truly stable and consumer focused operating system MS had released. The NT series prior had been stable (to a point), and was good for business uses, but try gaming or doing much that a home desktop user wanted to do and it really fell down there. XP picked up the mantle of being the successor to both Windows 98 and Windows 2000, and it really did a great job of it.

Was it perfect? Of course not. But it did a job and did it better than anything else in its time for its specific uses.

Not a single person got the watchmen reference

Completely agree. I was very slow to update XP, until both enough of a gap appeared, and Windows 7 established itself as the next high-usability Microsoft operating system.

Even then my use of Windows was low. I now have no need for it at all. From what I've heard Windows 10 is not in the same class.

Stability wise it's just as stable. People forget that even back in the day windows updates could break your machine and wax fondly with their memories rather than realistically.

> Windows XP is probably the greatest operating system...

I tend to disagree. In my opinion it is the worst OS ever created and the reason it's popular has nothing to do with quality.

XP was released in 2001, It was really good for its time. Most people kept using it(beacuse of Issues with Vista) till 7 launch(2009).So for 2009 XP may seem old and outdated. But the fact that after failure of Vista XP was able to keep OS marketshare for Microsoft is astonishing.

  > did not expect proprietary operating
  > systems to be actively working against you.

What did I miss? I've last used Windows in 2005, I'm pretty sure that was XP Service Pack 2. What does it do now to actively work against you?

Windows 10 has become much less open for non enterprise customers. It't incredibly hard to change settings or disable features such as Onedrive, Cortana, and others.

Agreed. I think MS has decided (for better or for worse) that the average user is not a good system admin, and/or a consumer OS shouldn't require one. Our IT maintains W10 pretty well. I have a test box in my office and I regularly get months of uptime without any nags of updates and other nonsense. I work in biotech and you don't want your cleanroom computers rebooting in the middle of an expensive batch. =)

Or maybe Microsoft just uses consumers, who pay less per license, as field testers for new features before those get rolled out to enterprise customers later?

I guess Telemetry and Ads is what they're referring to? And maybe automatic updates that wipe your desktop.

If I remember correctly, the trend of anti-consumer features was already well underway with windows XP. Integrated DRM was there. The limited activation count was there. Creating themes without signing was not possible, and the 'fisherprice' theme was rolled out even if a lot of users hated it.

Things were still overridable, but XP started the trend where the end user was not an allmighty sysadmin responsible for its own deeds good or wrong, but a consumer that should follow the vision of Microsoft and not touch the holy internals.

I can partially defend microsoft here. Everybody and their dog was messing around in the registry, was installing all kinds of system-breaking software, etc... But it also signifies a point where Microsoft realised it got powerfull enough to ignore the wishes of its own clients.

Virtually every third party theme came with instructions on how to patch uxtheme.dll to disable the signature check. It was not a big problem in practice, and I don't remember ever having to do the procedure more than once on a single XP install (unlike Windows 10).

To me there is a step difference between things like DRM it activation count limit vs disowning the user from his PC. One is creating a channel for bona fide purchases of licenses (you don’t need to buy music with DRM, it’s fair enough if vendors want to limit distribution of a product to the original buyer). The other builds spyware into everything you do with your computer and takes away control of it from you.

The analogy would hold if, dunno, buying some DRM Music also made you relinquish the ownership of your speakers to the vendor. It’s not quite as bad.

I can't agree with you here, and in fact, I think this comment is proof of a shifted overton window. You are correct that things are not as blatant as now, of course

DRM is, for example, a way to kill the second hand market or lending. I can sell my table without approval from its manufacturer, or drag it to my neighbours for a day. I can't sell my songs.

In the same spirit, if I as a computer enthousiast want to upgrade my computer hardware or reinstall windows, I should not depend on microsoft's goodwill.

I see your point, but I still think a difference remains. “Do you want to buy this non essential thing from me, that you can also buy elsewhere, with some restrictive conditions on license and reuse” is, in my eyes, a valid and honest proposal (potentially unattractive but that’s a different story). Take it or leave it. Exploiting a monopoly to force people, largely unknowingly, to surrender huge amounts of data and control over their computer, when there isn’t a practical choice (for most people it’s not really an option to switch to Linux or Mac, they are not fungible products) is a different cup of tea.

Aside from other stuff others already mentioned, I would also want to add:

Windows Vista and onwards, MS, while claiming it was for technical and stability reasons, nixed a lot of features that were annoying to movie studios, basically they changed how driver stack worked in a way to make DRM easier and make harder to pirate movies.

But as consequence of that, a lot of cool stuff outright stopped working:

* Some interesting video modes

* Some obscure ways to output data

* And the one that annoyed me the most: Soundcards became useless.

Before that, Soundcards could do a LOT, they could have their own processor and do a lot of fancy stuff, most famously they could do "3D Audio", and it was super cool, Thief for example is a game that uses it well, it would use its own processor to do same thing GPUs do for graphics, and do it for audio, it would even raytrace sound (ie: suppose a character is on the other side of the wall, but there is a door on that wall, the card would calculate the sound of that character footsteps reflecting on the walls and reach you, then if you were using headphone or 4+ speakers it would make you hear the sound coming from the reflections, thus a person with good hearing could deduce where an enemy was, on the other side of a wall).

That stunt basically killed off the soundcard market, people then were making cooler and cooler cards, after that the cards became little more than just output ports soldered on a tiny board.

Sound Cards pre-vista: https://upload.wikimedia.org/wikipedia/commons/thumb/9/95/Cr...

Sound Cards post-vista: https://http2.mlstatic.com/D_NQ_NP_768672-MLB41729093507_052...

A lot. Others have mentioned most of them already. Ads, telemetry, etc.

I'll just add—forced upgrades to Win10. Still hard to believe how much (some divisions of) MS despises its own customers. I call them the "Balmer camp" as a shorthand.

Despite parts of MS turning over a new leaf, the sociopath wing persist.

I considered Windows 2000 Pro the best version and switched to Linux after that - never looked back.

Windows XP's source code will never be leaked. Don't fool yourselves. Microsoft will never disclose the source code of theirs commercial software. It's against theirs law.

It’s illegal so it can’t happen?

If I'm going to play with a toy OS haiku is neater IMO.

Much more fun is writing your own.

I remember I had to reinstall it at least once per year to keep it fast. Except that it was OK. And yet my last XP laptop performed much better when I eventually ugraded to Ubuntu 8.04. Fast, no need to reinstall, pen drives immediately mounted and ejected, no drivers to install for printer and scanner, etc.

> So it's great that this source will not be forever lost.

What a WARPed perspective. (hehe)

It's great that this was stolen and distributed? There's nothing great about that. This source was not lost and has essentially zero risk of being lost. It would have been held for all time immemorial within the vault of its owner.

> Windows XP is probably the greatest operating system if you factor usability and how many people it has reached.

Sorry, but that honor must certainly go to Linux? The OS itself perhaps doesn't have usability, but it has flexibility and everything in modern life depends on it. Every single person in the world, everyone, today depends somehow on Linux.

I'm not poo-pooing XP in anyway -- it was and is a great OS. Both for its time and still today. But I think it is far in Linux' shadow.

Why and how do codes leak? I don't understand. Is it that some insiders are giving out the information?

It is still difficult for a new Comer like me to understand exactly what is going on here.

Could just be a disenchanted employee.

Creator of the torrent posted on reddit about it: https://www.reddit.com/r/windowsxp/comments/iz46du/the_windo...

>And also, Microsoft claims to love open source so then I guess they'll love how open this source code is now that it's passed around on BitTorrent.

One can love open-source and still be against license violations. What a warped world view.

License is violation!

I'm hereby revoking your license to downvote my blissful freedom of expression, thereby messing with my pursuit of happiness, causing stress!

(Marshal! To the yard! Switch that ugly element out of my train of thought!)

Yeah, if it were a violation of an _open source_ license, then the parent comment would make more sense.

Microsoft's love for open source is surface level; they love how easy it is to embrace, extend, and extinguish.

He was obviously joking.

I mean, he uses the term "OG Microsoft source code". He's just a kid.

That's some serious reach.

This sounds like a joke.

Wait, so is this something new, or just a re-release of whatever was going around on the net for years that was called the WinXP source?

Oh there's a lot of stuff like this floating around.

Important source code acquired and leaked by a script kiddie who has zero clue how valuable it is. Probably uploading and posting it to some small forum, where everybody else has even less clue.

Then it can take years until an actually knowledgeable person stumbles upon it and makes it known to the world.

One can get a list of files included without ever downloading any content (e.g. from the command line) and based on that, it looks like a troll trying to ride on the current interest. Happens all the time.

It seems that the list of files doesn't include anything that appears new (there were reports about the older sources leaked before and the list matches these), but includes the high troll insanity material about prominent "control" via "vaccinating the world".

The creator also included a bunch of wild conspiracy theory stuff in the torrent as well about Bill Gates and vaccinations etc which blows the torrent size out by a fair chunk.

It's actually quite interesting to watch, That James guy does a really good job with the research.

I still keep my 12 year old legal copy of Windows XP available to be run as a VM via VirtualBox for occasions when I need to try out some Windows/DOS stuff - works pretty well :)

Not sure why you got downvoted (too off-topic? no idea), because this is imo actually a rather interesting point. And 'pretty well' feels like an understatement when it comes to performance (or at least snappiness): I have XP on a VM for a while just for software which didn't support 7. Despite being in a VM (VirtualBox), the thing feels snappier and faster than current day desktop OS (XFCE on Ubuntu, Windows 10, OSX) on beefy workstations. It does lack quite some usability though in comparison with modern counterparts.

Yeah - probably considered off-topic - I just couldn't help sharing the anecdotal tidbit :) My xperience(!) has been the same - boots within seconds and basically gets out of the way. No advertisements/analytics (that I know of), great to quickly check a Word document that doesn't render well in LibreOffice etc. I don't use it often though - Linux has truly come a long way that such occasions are getting rarer as time goes.

One could also try ReactOS for such cases.

The ReactOS team might gain some insights from this leak...

and they get prison time.

If they copy code line for line yes, but might get insights they otherwise couldn't have. Or just pretend they happened to come up with solutions by themselves.

https://github.com/reactos/reactos#contributing--

" Legal notice: If you have seen proprietary Microsoft Windows source code (including but not limited to the leaked Windows NT 3.5, NT 4, 2000 source code and the Windows Research Kernel), your contribution won't be accepted because of potential copyright violation. "

That's a really honest approach, thanks for posting it. I stand corrected!

Edit: Might of course also just be covering their backs legally... but who am I to say.

Archive 40 GB Source code leaks 5 GB

Conspiracy videos about Bill Gates 35 GB

Just to put it in perspective... something the Gizmodo reporter miserably failed to do

Hmm, I had a look, the videos were more like 3 GB IIRC. There is a lot of stuff in that torrent, with the source to DOS, win2K, some Xbox stuff that's 5 GB itself, etc.

I did open a couple readmes, and the poster didn't appear that well-versed with tech, despite meaning well. It's an amateur job, but those tend to do the bulk of the work.

https://twitter.com/RoninDey/status/1309275918943301636

https://anonfiles.com/Ffn8u2Yaoa/Microsoft_leaked_source_cod...

Interesting, this one adds

    nt5src.7z
    windows_research_kernel/Windows Kernel Source Code like.7z
    windows_research_kernel/wrktools.7z

to a version of this torrent that was released a few days ago. I wonder if this is going to be an ongoing leak.

All these files have been out there for ages. Especially the WRK, which anyone with a .edu could already download.

I have an .edu, is there a place that I can grab this?

WRK is available on GitHub for years:

- https://github.com/9176324/WRK - https://github.com/ntkernel/wrk-v1.2

Strange that nobody noticed and sent strike to that repos

Interestingly, that ntkernel/wrk-v1.2 seems to be a public fork of the non-public "https://github.com/Trietptm-on-Security/wrk-v1.2" repo.

Might have been accidentally made public?

also ddk (driver development kit) which is publicly available sdk (now part of platform sdk instead of separate)

looking at filenames in top one there's a bunch of built dlls and exes there which looks like it is just installation of built code instead of source..

There are some source code archives in there:

https://pastebin.com/LyLH9bLR

There is a lot of filler, though. I assume the videos are trying to convey some point, but I'm not sure I care.

This is still illegal to obtain, right?

I'm not sure if it's wise to post the link as it may encourage others to download it and might be against the ToS on hackernews.

Is it, if you don't share it further?

It's a torrent though. I thought about downloading it, but it seems unwise when you're automatically sharing it with others.

You can prevent it from seeding and delete the tracker after downloading. It's not nice to seeders but it's an option.

It probably still is, but I doubt that will really stop anyone here.

you ever download a loaf of bread?

I am not sure this is against the guidelines. I don't think it's illegal to post it, but then there are some illegal numbers: https://en.wikipedia.org/wiki/Illegal_number

However, the commenter might have just posted the magnet link, which is more to the point, and unlikely to disappear (torrent 2.0 notwithstanding), unlike some third-party website (which could also turn out to be an infection vector). It also shortens the trust chain.

magnet:?xt=urn:btih:3d8b16242b56a3aafb8da7b5fc83ef993ebcf35b&dn=Microsoft%20leaked%20source%20code%20archive_2020-09-24&tr=http%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce

In what jurisdiction?

depends on where you live.

Interesting looks like theres also Xbox source code in there? Is that old or new? I don't really keep up with the reversing scene

Having not looked at any of the files yet, I wonder if it's just the changes that were made to WinXP for Xbox-use. IIRC throughout its' history the Xbox OS has just been a modified version of whatever version of Windows Microsoft is pushing at the time, with the OG Xbox being based around WinXP.

I'm very curious to see more come out as the code is picked apart piece by piece, and whether or not it'll confirm my suspicions. I'm also very interested in that Xbox folder, with any luck it'll be time to dust off my old OG Xbox and do some hacking!

AFAIK, the original XBOX was always based on a heavily stripped-down win2k kernel, and evolved very little. The "API" games used was a jump table to kernel functions, everything ran in kernel-space. I don't know much about later consoles, but I wouldn't be surprised if they used the same generic architecture.

https://xboxdevwiki.net/Kernel that site has quite a bit of info, with ongoing reverse-engineering and clean-room reimplementation projects (reactos, xquemu)

reactos being clean is a funny joke

I'm surprised it hasn't leaked before. Microsoft used to give access to their OS source code as part of collaborations with academics (I don't know if they still do). I was surprised at how lax they were. It seems they didn't care that much.

> And that could be used for good reasons — like people trying to create Windows emulating software on Macs, for example

Not really. It violates the major principle of clean room reverse engineering. That's why Wine developers are strictly avoiding such kind of situations. It can taint the result with legal problems.

The way the original IBM PC BIOS was rewritten was one team analyzed the code and documented what it did, and another team implemented it from the documentation. Having source code to start with helps with the first step.

That sounds like a violation of that principle, but if original authors don't care, then no issue.

> Would you believe more than 1% of computers worldwide are still using Windows XP?

Probably higher -- XP is still popular in China and supported by third party companies there.

I saw XP quite a bit while I was there too.

Are these airgapped machines or machines with networking? Internet connection?

Can't speak for Chinese systems, but I'm aware of "jump hosts" (used for remote access to industrial control systems) that are running windows server 2003 and are internet-exposed.

What's the demographic that's using XP in China? I find this super interesting...

Form my experience, mostly government and state-owned companies, which account for a big part of China's paid software market. Even some Shanghai government systems support only IE6 to 8, some only love IE11.

Most state-owned banks too, popular browsers are all duel-engine. There's little to no incentive to change what's working, and customers experiences don't really matter.

> customers experiences don't really matter

To consider it from a different perspective, perhaps customer experiences matter more in China. Technology in the US is super cool and cutting-edge but the constant shift to new technologies and the rapid deprecation of slightly old software leaves a significant body (perhaps even the majority?) of consumers in the dust.

Industrial equipment control computers mostly - the computer is part of the machine and cannot be upgraded. You see this quite a lot outside China too.

Honest question: Would using Linux have been a better idea then? An old unpatched Win XP has tons of vulnerabilities but so does an unpatched 20 year old Linux Kernel.

Using Windows XP on those devices often gets very bad rep but is there a viable alternative for machines that are air gapped and technically won't be upgradable for several decades?

Remote vulnerabilities don't really matter for air-gapped machines.

"Attacks like Stuxnet, the computer worm deployed against an Iranian nuclear facility a decade ago, shattered the myth that air-gapped systems are impenetrable fortresses."

source: https://www.cyberscoop.com/duo-labs-air-gap-radio-mikhail-da...

Links to: https://duo.com/labs/research/finding-radio-sidechannels

It wasn't really air-gapped if someone plugged an USB in it, was it ;)

I'm aware of the "over the air" methods to _read_ data from air-gapped systems (The blinking hard-drive light!), however I was referring to doing damage to industrial systems running legacy software, where you usually find this (think a big machine with a "terminal" attached to it that just controls the hardware and has no inputs).

That's exfiltration, not infiltration.

In the XP days Linux was not seen as a serious things.

At that time using XP was a very viable thing to do. You would pretty much never do it now. Think about it from a business perspective. Something goes wrong in the OS. At that time you would have had to have linux guru on staff to take care of it. OR you could pay 30 bucks for a OEM embedded copy. If something went kinda bonkers you could call microsoft and they would fix it. Their support on things like that is pretty good. At that time getting that sort of support for linux would have been basically nonexistant. Today you could get that kind of support for linux. Back in 2000-2004 it would have been tough. At this point no company is going to re-write the firmware on a machine they sold 20+ years ago unless they happen to be still selling that exact same model and you are paying for some awesome support contract.

Not if Linux won't run the programs you want to run or you don't have IT staff to get it to run with wine or something. Average users will just "take their chances" if they're behind a firewall and not surfing the web then chances are very low for getting a virus if you glue up the usb ports.

No, there are tons of regular office computers and even home computers that run XP. Industrial computers in China would actually be running fairly up to date OS's compared to, say, US.

I visited a number of factories in China last year and saw lots of XP on measurement, production and test equipment. All the office desktops were on W7 or W10. I'm sure lots of XP desktops exist, but I didn't see any in the (quite many) offices I passed through. I can't speak about home computers, I only saw a small handful.

You'd be surprised how many KIOSK's out there are running XP under the hood. Especially ATM's.

statcounter reports it to be 3.11% in the period Aug 2019 - Aug 2020

https://gs.statcounter.com/windows-version-market-share/desk...

Right, as I've mentioned elsewhere, I have one of those XP machines still working.

Also, as I've mentioned ages ago, a few years back I visited a factory with a very expensive 5-axis industrial milling machine (I was told its value was about $300k). I was very surprised to see that it was still running Windows 2000—and that's something I'd definitely notice immediately, as W2K has always been my favorite Windows version.

When I asked the engineer in charge if they were going to upgrade Windows 2000 to a later version, he said without any hesitation whatsoever that they definitely would not. Moreover, without further prompting, he said that the machine had a design life of 25 years which meant that it would be in operation until at least 2025 if not longer.

He added that I wasn't the first to ask this question and went on to say that upgrading was irrelevant as W2K was still fully fit for purpose.

What is even more surprising is what I saw at a brand new industrial plant that I visited even more recently. This time the company was a large international operation with various plants across the world that is involved in the manufacture and assembly of heavy industrial equipment and parts. One of the reasons I was visiting the factory was that the plant was brand new, in fact tooling setups for some of its operations still had to be fully commissioned.

What I saw there were literally dozens of PCs still running Windows XP! The PCs were integral to the company's factory automation in that they controlled various industrial equipment and processes most of which were all brand new (after all, it was a new plant).

Here too, I quizzed the engineer in charge of showing me around about the use of the now very-dated XP on the company's PC. His reply was that by using XP in the operation of the new plant would ensure that this plant would still be identical with their other plants across the world and that this was strategically very important for many reasons, especially so at times of maintenance and breakdowns (as they could call on common practices and procedures already in use across all of their plants worldwide).

He added that whilst they still had considerable stocks of normal commercial PCs that were hardware compatible with XP, they nevertheless were aware of the potential that at some future point they might run out of 'normal' XP machines. To solve this potential XP/hardware obsolescence problem, they had an easy long-term solution, which is to use mobos that are normally used for industrial control processes (embedded systems, etc.) and there is no shortage of ones that are compatible with XP (in fact I've catalogs that have hundred of similar boards, so many in fact that it's a pain to differentiate or select the best board for one's application). Incidentally, many IT people have never seen these boards due to their niche nature. Even I think it strange that so many types of x86/64 PWAs exist as I can think of better or more appropriate processors for embedded systems, TI, Motorola and ARM chips for instance).

If anyone thinks old Windows technology is actually obsolete then let them think and think again. It's only naive or disinterested Windows users who are cajoled or coerced by Microsoft into upgrading Windows and those forced by default to do so when they buy new PC hardware who actually concern themselves Windows upgrades.

Of course it's only we IT professionals who are permanently preoccupied with OS upgrades and worry about such matters—not to mention the fact that we live in a perennial state upgrade-itis and always have security matters/patches permanently ingrained in our minds. The fact that we think this way lulls us into believing that most PC and smartphone users similarly think this way but in fact nothing is further from the truth. Moreover it's only a very small percentage of 'normal' PC users who ever bother to give their OS a second thought and that's only when it's a means to an end. Others including many industrial users are much more pragmatic and sanguine about upgrading and only do so when they absolutely must—and never at Microsoft's whim or wish.

Like it or not, expect to see ancient museum-piece versions of MS Windows still in use for decades to come.

My immediate thought is whether researchers or investigative journalists will find cold hard US government backdoors. This is potentially big.

They won't. Such backdoors would have to be hidden from Joe Average Coder at Microsoft first and foremost. Coding for MS is not a livelong thing, you know? So any backdoor would a) be obfuscated and b) have some form of plausible deniability. If I would have to make them, they would look like strings of two or three bugs.

Likely you're right, but let's see if anything can become plausibly demonstrable after obsessive scrutiny. Like many situations in life, it may depend on whether someone determined / resourceful enough wants to do this. There may be no one with sufficient motivation.

Also, it's not just (allegedly) all of XP source that's been leaked:

https://www.bleepingcomputer.com/news/microsoft/the-windows-...

It's also Windows Server 2003, MS DOS 3.30, MS DOS 6.0, Windows 2000, Windows CE 3, Windows CE 4, Windows CE 5, Windows Embedded 7, Windows Embedded CE, Windows NT 3.5, and Windows NT 4!

That's a huge amount of stuff to analyse.

Do you know if there is source code for classic Windows apps like the Calculator, Notepad or Paint? Would love to recreate those simple apps and chuck the lousy Linux/Mac equivalents.

I am not sure if it is what you are looking for, but the calculator is now open source under the MIT license https://github.com/Microsoft/calculator

As a recall, making notepad is like a homework assignment for a visual basic class. You just drag the text editor window and add the menus, there isn't a whole lot there!

Paint would probably be a bit more work, but there are a few clones out there already

Notepad is simple to build in Visual Basic because it uses the controls that are inherited from... notepad.

What GP meant, is whether the source for those controls is available.

Notepad is simple to build in Visual Basic, Delphi, heck even assembly because the controls are inherited from Windows[1].

[1]: https://docs.microsoft.com/en-us/windows/win32/controls/indi...

There is no reason to build them for Windows. I just use the ones already available for Windows. I want to build it for Mac so I can get the great simple tools on Windows but on the stability of Mac. For Paint I have resorted to buying a Mac Store app called Paint2 made by some Chinese developer. It is still too complex but is the best I can do now.

Really looking to build a tool with the same look and feel. Launches fast, can just paste text stripping formatting and then allow you to save it as txt or copy it as is to other apps. A typical Ubuntu install gives you something like gedit2 or pluma. Too much for a text editor. Same goes for Paint. I can't believe that I have to install a heavy app like Gimp, Pinta or Krita. They are significantly more heavy and less stable in my experience. The simplicity of these tools make their brilliance.

The same goes for Mac. Mac has "textEdit" and "Notes" apps but it handles formatting, is slower to launch and is more of a Wordpad clone. For Paint, their Preview tool is terrible. I just want to take a screenshot, paste it in a file, maybe add a red circle or square and save it as png. Quick and fast.

It would be nice to have a Github repo where you can browse the history like for Unix at https://github.com/dspinellis/unix-history-repo . Not going to happen, of course.

There's always Ken Thompson's hack of the compiler to insert a backdoor into the login program on Unix. See Reflections on Trusting Trust: https://www.win.tue.nl/~aeb/linux/hh/thompson/trust.html .

Obfuscated like NSAKEY?

https://en.wikipedia.org/wiki/NSAKEY

There have always been rumors about such back doors, but even the public ones, encryption schemes, have never been proven by pure analysis. As far as back doors in code goes, have a look at this: http://underhanded-c.org/. It's a pity it seems to have been short lived.

“Never been proved” is correct, but this proof is quite a tall order.

Remember, almost everything in the Snowden disclosures was known before snowden - but generally dismissed as “conspiracy theories”.

None if it was proved, until it was.

Then it needs some authoritative source, a leak, or result of a raid. Which was my point: analyzing the Windows XP code most likely is not going to prove there's a US mandated backdoor.

That would surprise me. Governements all over the world already received access to the source code to perform audits.

Besides, probably the quality was still low enough that you didn't need this, there were plenty of bugs that would grant a big organisation access.

They got access to the source code, but not the ability to compile themselves.... my university had such access.

If it really does exist, the backdoor can simply be inserted during the last-minute compile before release. It would be invisible in the code repository, the vast majority of internal developers at Microsoft won't even see anything unusual. Also, I heard anecdotes that Microsoft already allowed governments to audit the source code of Windows under NDA on multiple occasions in the past.

But if you cannot guarantee the correspondence between source and binary releases, such a review only helps a little. Reproducible build is crucial for auditability.

And that's why modern compilers won't have that. It was exactly so the same source, even compiled a second later, it will generate different binary file.

I hate it. It was so easy in DOS era - same source, same binary file, easy peasy.

The reproducible builds folks are working on fixing that:

https://reproducible-builds.org/

The major variables in modern compilers are just automatic timestamps, exploit mitigation random seeds, and toolchain versions, it is possible to make them immutable. The problem can be fixed, and there are already major projects to address it. Do you know that 90% of the Debian packages are already reproducible [0]?

[0] https://isdebianreproducibleyet.com

That’s true. But that may also be a plausible deniability thing - you create a place to hide binary modifications by making sure no two builds are exactly the same.

It could be chalked to some lack of care; however, up until 2000 or so, non reproducible builds were considered a bug in at least two places I worked in. The fact that it has become so hard to make builds reproducible could be Increased “entropy” (because no one cares To fix it) - but it could also be orchestrated by someone with a vested interest.

E.g. - suppose you are a three letter agency, and want to implement a “reflections on trusting trust” attack. Non reproducible builds become a pre-requisite.

No disagreement. It's why the problem needs to be fixed, although it's not a silver bullet (the compiler bootstrapping is still vulnerable).

Most modern compilers have switches to disable this behavior.

> I heard anecdotes that Microsoft already allowed governments to audit the source code of Windows under NDA on multiple occasions in the past.

I can confirm that back in the Windows 2000 days, MS let a major global bank have access to the Windows 2000 source code to aid them in coding some low level bespoke software.

Back then, if you were a gold customer, you could pretty much get anything out of MS under an NDA.

'MS let a major global bank have access to the Windows 2000 source code to aid them in coding some low level bespoke software.'

Examples of this are comparatively well known but it seems to me it's really not relevant here. In those instances it's extremely unlikely that said institutions would have access to even the majority of the source code let alone all of it.

All they need are API hooks and or various security code that's relevant for their purposes, etc. If I were the Microsoft person responsible for interfacing with these banks, I'd do what I've done with unrelated stuff, which is to tell them just sufficient to do the job (that's to say only on a need-to-know basis).

This is correct. Back at least in the Vista days each public release had to be passed to the NSA to validate the encryption algorithms and implementation. I’d imagine they poked around a few other parts of the OS

> each public release had to be passed to the NSA

That sounds a lot like a conspiracy theory. Any references on this?

The NSA has a dual mandate - to improve security for the US government (and by extension, US businesses to an extent), and to peek into communications outside the US.

They have been pretty negligent about the first (or even malicious about it - e.g. the dual drbg case), letting the second take over - but officially they still have the first mandate.

In fact, DES was considerably strengthened in its day by the NSA review - at that time for reasons not understood by industry or academia. It was later discovered that the change required by the NSA made DES much more resistant to differential cryptanalysis, a technique that was (re)discovered by academic cryptographers much later.

Every “conspiracy” I’ve heard about the NSA and friends turned out to be true, most with definite proof from the Snowden releases.

> Every “conspiracy” I’ve heard about the NSA and friends turned out to be true

And therefore every next thing anyone on the Internet concocted has to be true as well?

I get what you're saying though, and I'm well aware of the dual mandate. But you haven't given me anymore reason to believe this particular one, and I find it strange because I've never heard of any company in any country having to give their software to an intelligence agency before being allowed to release only the modified version. Implanting backdoors is not unheard of, but after they had great success with the clipper chip it's usually done without the company in question knowing about it.

You say the excuse was to validate some implementation, but in that case Microsoft wouldn't need to publish any modified versions, the nsa would just point out "this contains too strong crypto, can't export this" or "you made a mistake in algo X allowing attack Y". Not "please substitute this dll with our version and better not tell your customers!".

You are reading too much into cududa’s comment and mine, quite a bit of things neither of us said.

Cududa mentioned Microsoft let the NSA review it - that it was procedure, not that it was law. Furthermore, no one claimed that NSA recommendations or replacement DLLs had to be used by law - though that makes little difference. I mentioned that this would be in line with well known history of DES development. That’s basically all we said about it.

I have no idea what you are referring to with the Clipper chip. But RSA, NIST and others were definitely aware they were peddling NSA recommendations With the dual-drbg fiasco.

I'll reply with my own doubt myself:

There's a possibility the leaked files are tampered from the original code anyway, i.e. backdoors were removed before being initially leaked. Rationale being that Microsoft / government wanted to control the situation long-term by letting something tampered be what leaks out underground instead of the 100% full thing.

Torrent poster also discusses that possibility: https://www.reddit.com/r/windowsxp/comments/iz46du/the_windo...

Nonetheless, it's interesting if anything plausible is found.

EternalBlue was probably the biggest backdoor used by the US government for years and even Microsoft (at least officially) didn't know about it. Finding and not reporting bugs is much easier than getting a company to put in backdoors without anyone blowing the whistle or objecting.

Plausible deniability where people can look is a huge thing. Even where people can’t Easily look - e.g. Intel ME “firmware”, it’s likely done as shoddy/buggy coding, so Intel can just look incompetent (and not downright malicious) when it does come out.

Russia and China aren’t buying any of this, and are fabricating their own chips. 5-Eyes are likely in on the thing so no reason for them to set up their own fab facilities.

But setting up your own software ecosystem is much easier (especially given Linux / BSD), even though it’s still expensive - so intelligence agencies would rather not give countries incentive to do so just because there’s a back door.

Here's the original source: https://boards.4channel.org/g/thread/77879263

And the archive when it gets pruned: https://archive.is/eg9nw

"That leak is already turning up some interesting stuff, like the NetMeeting user certificate root signing keys." -- https://twitter.com/gsuberland/status/1309366364537266177

Maybe people can take this code, fix a bunch of vulnerabilities, and keep developing a fork ofvit where MS left off Unlikely, but would be cool :)

It would be very handy if someone were to generate new software drivers for newer equipment from scratch.

I can't see a copyright issue if new drivers were brand new code. The trouble's always been MS has deliberately denied providing new drivers for newer hardware as deliberate means of obsoleting its older versions.

Using older versions of Windows (such as Win 7) on new hardware is a highly desirable idea for those of us who are mightily pissed off that Microsoft has essentially converted is Windows platform from an operating system to a spyware and advertising platform. New drivers would solve the problem.

Would be interesting if it can actually be compiled into a usable ISO - possible implications for digital preservation.

Would love to see the actual compile/build process for this. I don't think we'll see bootable ISOs (from this).

"tools\razzle offline" sets up the environment

"buildx /z"

"postbuild" should make an iso

Does razzle dazzle contain any easter eggs?

I can't spare mobile bandwidth to try this out, otherwise I would happily poke around myself.

I wonder how this will affect other projects, such as ReactOS and Wine. Exciting.

I'm not sure if they want to go this way https://en.wikipedia.org/wiki/ReactOS#Internal_audit https://wiki.winehq.org/Clean_Room_Guidelines

They can't look at it without running the risk of copyright infringement [1].

[1] https://en.wikipedia.org/wiki/ReactOS#Internal_audit

It's radio active for those projects if some reason they tried using this info.

That's so sad that we have to waste time implementing bug-compatibility with the proprietary software.

There should be some law about interoperability.

How would such a law protect you against a copyright lawsuit where the rights holder claims you looked at leaked sources?

AFAIK there are already laws about reverse-engineering for compatibility. May extend that and add something like the right to repair, so the ones who sell software or software services also have to provide documentation and more access to the actual features.

Like the thing is a windows binary is not following necessarily following any standard like POSIX. So implementing a compatibility layer without a spec or source can be tough.

That would be really bad for these authors to download or look at leaked code. That's exactly how you get yourself in legal troubles.

That's why they have policies to not look at any leak, and even with this ReactOS and Wine already had their share of controversies in the past.

But realistically, how will they know if some random Wine contributor is "inspired" by leaked code?

It seems like it would just go undetected.

Both ReactOS and Wine are open sourced, Microsoft engineers and lawyers look at their code. That's not a hypothetical situation, it happened in the past, the most recent news event on this was: https://www.theregister.com/2019/07/03/reactos_windows_resea....

With the risk of copyright infringement you want to be sure that you're taking active measures to defend your position in case you're facing legal charges.

(Usually projects of this sort will make you guarantee that you have not been influenced by the closed-source code they are trying to clean room.)

Yeah, I actually once worked at a place that was clean-rooming. They went to absurd lengths, even WITHOUT the old source code.

If you were designing the new app you couldn't even -look- at the existing UI; you could only get descriptions. Every term was checked by legal to make sure that it was a 'standard' industry term vs something that was specific to the existing software. They even used a different tech stack wherever they could.

I think the other company still tried to sue, IDK how it played out (I was gone by then) but I remember at least thinking they were in a pretty safe spot with how they handled it.

However, maybe having someone external to the project look into, and understand those undocumented functions or system calls, and explaining them to the people which will actually implement them could work.

This is not the first time Windows code leaks. I also have a vague recollection of one of these projects actually having to rip out and rewrite a subsystem because it turned out to be too close to original Windows sources found in one of those previous leaks.

In short, this leak likely won’t affect any project that is seriously trying to reimplement substantial parts of Windows, if they care at all for US law.

To FOSS projects, leaked source code is highly toxic and infectious. The FOSS community's attitude on leaked source is similar to the corporate world's attitude towards GPL in the early 2000s, but unlike GPL, this is not just exaggerated FUD - the dangers of leaked, unauthorized or proprietary source code is real. Once leaked sources found its way to FOSS projects, the entire project may become illegal and face potential lawsuits (see the history of how BSD was almost killed by AT&T's lawyers). And once leaked sources found its way to your brain, it's not 100% safe [0] to work on FOSS anymore and the best option is banning yourself from participating similar projects, it's almost a memetic virus.

Projects on reverse-engineering and reimplementing proprietary technologies are the most vulnerable. A decade ago, ReactOS's development was suspended for years until the codebase has been reviewed.

[0] It's not illegal. If you can absolutely guarantee your work is not a derivative work based on the leaked source, but just a reimplementation, it's fine to proceed (the clean-room approach is not always needed, it simply offers the strongest legal guarantee). But for a big project with numerous contributors, the risk is high.

Regardless of what you do, I'd be extremely careful about looking at leaked commercial source code. While I think you could legally get away with writing code that was just inspired by something you've seen, I not sure I'd want the hassle of having to potentially defend myself if it came to a lawsuit.

It will reduce the number of eligible contributors. If you have seen the windows source you are not allowed to contribute to wine. Unless someone uses this as an opportunity to document the interfaces and create a test suite and thereby allows a clean room implementation the source code isn't just useless. It actively ruins the wine project.

Is this all of it, or just portions like many of the other leaks?

Get it, try to build it, see if you can make your own installer ISO and test it in a virtual machine. Ain't rocket science, just time and if you have it (which I don't) knock yourself out.

And seeding it is highly illegal and unethical.

Taking a cynical view (and how can you not these days), the inclusion of the anti Bill Gates videos could be actual goal here. The source code leak as the headline makes a lot of people interested to download it, and then maybe they watch those videos as well. This could be part of a misinformation campaign aimed at spreading discord around COVID-19. Given the timing and with the upcoming US election, everything needs to be looked at with a critical eye.

Does this mean we’re about to see a wave of new exploits?

Or has this code been out there for a while?

Governments, Universities, and large corporations have had ongoing source licenses for Windows back to the 95 days.

https://www.microsoft.com/en-us/sharedsource/

Worth noting that "open-source" doesn't imply a license to use the software in any way. You can put up a project on Github with "all rights reserved" (though you obviously grant Github rights to distribute it when you upload it).

You are mistaking "source code available" for open source, which means a specific set of things, and does ensure you can compile and run the code.

I chose the most literal definition because it's complicated. Debian calls it "free software," but that has its own issues because there's "freeware" where source might not be available and the license isn't permissive. Then there's the JSMin "no evil" (unless you're IBM, its customers, partners, or minions) license.

> I chose the most literal definition because it's complicated.

That's not the most literal definition, and it's not complicated once you know what it is. Open source[1] is a term coined by people decades ago to refer to something very specific, not "source code can be seen". There are specific licenses that are considered open source, and others that are not. Open source specifically requires a license that allows certain capabilities.

You can read more about this at https://opensource.org/. There's been some misunderstanding over the years where people (including me) have misapplied the term to mean you can access the source, but that's not really what it means, so we shouldn't dilute its meaning by allowing that.

> Debian calls it "free software,"

Free software is also a term used to refer to open source.[2]

I understand it can feel pedantic to have this terminology so strongly enforced, but the reason people do so is because there have been efforts by commercial entities to co-opt the term for their own version of software and software as a service which does not conform to what open source stands for, and ride on the coattails of the movement and good disposition towards open source. Your first comment actually seemed to be trying to bring awareness about this. What you can say as a shorthand going forward if you want is that "releasing the source does not mean it's open source, which if it was would mean you had a protected right to actually build and run the software in question". That's both more correct, leads people to more information about the actual issue in question, and lets you offload some of the more complex explaining to the site made for it.

1: https://opensource.org/

2: https://opensource.org/faq#free-software

> Or has this code been out there for a while?

According to the torrent creator's reddit post "[a]pparently the file had been passed around privately for years among hackers".

How much of the source code is carried over from release to release of Windows and how much is new?

I'm reasonably sure that large parts of windows were re-written between 2003 and Vista; so I'd assume the vast majority is new.

I really wish the 3.1 code had leaked, for the benefit of all the retrogaming out there, but the source code dump has a lot of interesting things in it.

Win32 is probably the biggest set of games that you can still play as abandonware.

Gizmodo's baking metaphor is really really awful.

How is it awful? Can you give me a better analogy?

I didn’t mind their analogy, but I think an analogy is a poor way to explain such a simple concept. Why not just say, “the code that the developers wrote” or something along those lines? I don’t think the concept of source code is inaccessible to laymen.

> Why not just say, “the code that the developers wrote”

Because some readers might not know what is the source code. Or they might have a vague idea but they might not understand what it means for the source code to be released.

> I don’t think the concept of source code is inaccessible to laymen.

I don’t think the article claims that. On the contrary; how I see it, the article explains the concept.

Well, was that an early draft, or is it after the editors went in and made changes/corrections? How many chapter are in the manuscript?

Because the source isn’t a recipe in this analogy it would be akin to buying a cake mix and following the cooking instructions.

A further development of the analogy would be one could try to replicate the cake mix by reading the list of ingredients and using their knowledge of baking and trial an error to replicate it as closely as possible vs getting their hands on the exact formula used by the maker of the cake mix.

Blueprints. Source code for a program is analogous to blueprints for a building.

I found the whole article very low quality. I don't recall the last time I read anything on Gizmondo, but I wasn't expecting that.

> Can you explain what a source code is to me like I’m five?

I’m astonished by the amount of people who choose to leak code on 4chan.

As opposed to where? It is one of the few platforms that would allow a post like this other than security info forums, but if you aren't looking for credits on one of those 4chan is about all that is left.

There's literally nowhere else to leak things like this without removal (major platforms) or tampering by the administration (wikileaks et al).

Friendly reminder to use a VPN in the hypothetical case where you would want to download such a leak for _research purpose_.

Always a good advice but I wouldn't worry about MSFT going after downloaders of Windows XP source code. If it was about Windows 10 or even Windows 7 then maybe, but I don't think they care much that this is out in the open.

Downloading it would probably not even be a crime, in the same way a journalist getting hold of confidential IP wouldn't be a crime either. They'll go after you only if parts of the code end up surfacing in competing products, or if you belong to the distribution chain that leaked it in the first place. Microsoft is traditionally pretty skilled at dealing with piracy and leaks in more sophisticated ways than "sue them all".

You do realize that technically, while you are torrenting, you also distribute, right? Because that's how torrent protocol works.

You can easily disable upload in every usual torrent client

I think this applies only to US. In EU there is no way you can even enforce those EULA screens at install or prosecute someone for downloading warez (distribution is a different topic).

It's a torrent. So you are a distributor under all common clients.

I'm not sure you can get a client that would distribute nothing. You also have to make sure you don't advertise others in the swarm or any info.

Researchers have done it, not sure how technically.

Why My Printer Received a DMCA Takedown Notice - http://dmca.cs.washington.edu/

It was OK, but it was the thin end of the wedge of many things bad about computing since then. I'm referring to product activation and updates.

The internet and operating systems have had an evil feedback loop. To quote from The Outer Limits, 1963: "We will control the horizontal. We will control the vertical. We can roll the image, make it flutter. We can change the focus to a soft blur, or sharpen it to crystal clarity. For the next hour, sit quietly and we will control all that you see and hear. We repeat: There is nothing wrong with your television set."

Even supposedly free browsers like Firefox are getting in on the act, pestering the user with updates, and make it difficult to turn off. My dad complains about this constantly.

More and more, we see external agents dictating policy.

I don't think even Richard Stallman originally predicted the true nightmare that computing would be become.

I'm actually not familiar with the legality of downloading this torrent. Is this considered legal, or is it like stolen goods, where the mere possession is a crime, even lacking evidence of the actual theft?

Also, what would be the legality a group of people "maintaining" this code? (developing security patches, etc)

A simple rule of thumb from someone who is not a lawyer. Is the thing you wish to download the property of someone else, and have they granted you the right to view/use it. I think the answer is pretty clear here.

The first answer probably depends on where you live. In the US, I would not bet on it. In Germany, I could at least argue that you legally cannot steal source code.

For the second part: that's probably legally impossible world wide due to trade secret protection agreements.

Another interesting question to ask may be: Would downloading/seeding this torrent be considered a worse offence than if it were, say, a movie?

I can't see why Microsoft would bother pursuing people for downloading or seeding this. The cat is out of the bag. And people will use this in ways MS don't want regardless of what they do. It is not like a movie where there is a loss of hypothetical revenue. Source code is not exactly an easy was to install a pirated OS!

Plus it's not like someone's gonna run XP SP1 on their main machine. It's been 18 years.

i would add that the biggest problem is not the downloading part, but the upload which you are also doing when using torrents.

> the upload which you are also doing when using torrents.

Unless you set your seeding limits to 0kb. You take a hit in speed/availability (as more peers will likely not do business with you), but you also don't distribute anything.

Honeypots can still record your IP address and ask ISPs which customer is behind that IP.

> patches

People did this for the last version of opera that still used its own engine.

Depends on the jurisdiction and the legal system where you reside. Copyright is a complex matter.

Long story short, there are two major ways in how copyright violations are approached. Either the other party starts a civil suit in court and tries to claim damages (much like if you would damage someone else's property); or state law considers this a prosecutable felony or crime, in which case you look at punishments such as fines or prison time. Moreover, you may run the risk of being liable to both scenario's.

In the United States, criminal copyright law is a thing:

[1] https://en.wikipedia.org/wiki/Criminal_copyright_law_in_the_...

However, it does state:

> Criminal copyright infringement requires that the infringer acted "for the purpose of commercial advantage or private financial gain." 17 U.S.C. § 506(a).[9] To establish criminal liability, the prosecutor must first show the basic elements of copyright infringement: ownership of a valid copyright, and the violation of one or more of the copyright holder's exclusive rights. The government must then establish that defendant willfully infringed or, in other words, possessed the necessary mens rea. Misdemeanor infringement has a very low threshold in terms of number of copies and the value of the infringed works.

> An individual may be liable if the infringement was committed: (B) by the reproduction or distribution, including by electronic means, during any 180-day period, of 1 or more copies or phonorecords of 1 or more copyrighted works, which have a total retail value of more than $1,000; or (C) by the distribution of a work being prepared for commercial distribution, by making it available on a computer network accessible to members of the public, if such person knew or should have known that the work was intended for commercial distribution. 17 U.S.C. § 506(a)(1).

> Without establishing the threshold value, legitimate infringement, or the requisite state of mind, there can be no criminal liability. If the defendant can show they had a legitimate copy or use – such as through the first-sale doctrine – then the burden of proof falls on the government.[9]

Directly using the code would imply liability to copyright infringement. When you want to compete with the product of a competitor, that's something you want to avoid. A classic tactic to avoid this is "clean room design":

[2] https://en.wikipedia.org/wiki/Clean_room_design

> Typically, a clean-room design is done by having someone examine the system to be reimplemented and having this person write a specification. This specification is then reviewed by a lawyer to ensure that no copyrighted material is included. The specification is then implemented by a team with no connection to the original examiners.

ReactOS is a contested example of "clean room design" for copying elements from Windows:

[3] https://www.theregister.com/2019/07/03/reactos_windows_resea... [4] https://news.ycombinator.com/item?id=20341933

> "I think it's a ripoff of the Windows Research Kernel that Microsoft licensed to universities under an agreement that was obviously violated by some, as the code has been uploaded to numerous places, some of it on GitHub," Rietschin wrote. "I glanced at the ReactOS code tree, and in my opinion, there is absolutely no way on Earth this was written from a clean sheet only from the available public documentation." He says that "internal data structures and internal functions, not exported anywhere and not part of the public symbols, have the exact same names as they appear in the Research Kernel."

The issue with this code leak is that any engineer looking at this risks being considered "burned" if they are ever asked to implement similar features in a comparable product and rely on their knowledge. I wouldn't touch it with a 30 feet pole.

"You want to update XP so it can run with new hardware? Go for it." - I would really like to see a version I could use with 2013-14 MacBook Pro/Air.

I have a MacBook Pro 2007 multi-booting XP, 7, Ubuntu, 10.4, 10.5, 10.6, 10.7, 10.8, 10.9, 10.10, 10.11. Surprisingly, XP is the OS I use most frequently on that laptop, by choice.

I got some old HP laptops running XP for the homestay family kids, and have been playing network games and chatting with Escargot MSN.

Windows XP still supports 16 bit DOS games. If the Luna theme seems childish, it's even more appropriate for young people using computers for the first time. Buttons have icons, which is easier when you're 6 and can't read quickly.

I hope this source code can allow me to boot XP natively on my main laptop, so I don't have to carry around another computer just for compatibility.

> I hope this source code can allow me to boot XP natively on my main laptop, so I don't have to carry around another computer just for compatibility.

In my experience, Windows XP runs fine on relatively modern hardware. The biggest offender is generally storage drivers (AHCI). You can download a driver and re-pack the installer with it.

Note that I haven't tried this in a long time, it might not be practical anymore. But with today's HW-assisted virtualization, you could just boot it in qemu or any other accelerated virtual machine, and you couldn't tell the difference.

'In my experience, Windows XP runs fine on relatively modern hardware.'

How modern? And what hardware models? Was the newer hardware Win 7 compatible or later?

I wish you'd be more specific and tell us how as it's never worked for me even after mucking about with different PC hardware and various different AHCI and other drivers.

I'll try again soon. I definitely remember running it on:

- Intel Core 2 Quad and Xeon E5472, DDR2. Old-ish (the xeon came out in 2010), but I used that system as my main until 2018 -- though with Linux.

- Samsung N220 (or quite similar) netbook. That thing originally came out with Win7 starter, and was a slog with it.

A cursory search ("Ryzen XP") gives plenty of results: https://forums.guru3d.com/threads/installing-windows-xp-on-a... https://community.hwbot.org/topic/189776-xp-works-on-ryzen/

Though by now ReactOS might be better supported, especially on modern HW.

Thanks for the info. I'll check that out.

I wouldn't hold my breath with ReactOS. It seems to me there's so few developing it that it'll never reach critical mass. Shame really, as alternative non-MS O/S that runs x86/64 Win APIs would be useful.

An interesting question here is how did Microsoft's systems get compromised such that their source code was leaked?

I imagine it all went down in a manner very similar to the sense/net hack from neuromancer.

I've heard that governments got access to source, maybe from there?

On this note, does anyone know where to legally obtain windows xp, hopefully for cheap ? I should note that I've tried at multiple places and none seemed fine so kindly speak from experience. (Or sell me your own, I'd trust anyone here).

The internet archive can help you out here, I used this recently for an XP VM

https://archive.org/details/WinXPProSP3x86

Technically I'm looking to buy a serial so I can tell myself I didn't steal. Thank you for your link, once I have a serial I'll know where to get a safe copy from.

I have an educational MSDNAA (or Dreamspark, or whatever it is called this month) key I can send you, if that's legal enough for you.

You are so kind for making that offer but probably not, thank you :). I want to buy from someone who has paid money to get the key. Thank you again.

Why, if you don't mind my asking?

Even if you paid anything for such an old product, your money wouldn't be going to MS anyway. Not sure what difference it makes, paying for an abandoned product.

While you’re not wrong, I think no one should have to explain themselves for wanting to stay compliant with the law.

Even when the law has no basis in reality? Windows XP is going to stay in copyright for... how many decades now? Even though it has no commercial value anymore?

Laws which make no sense and have no value to society should be ignored. If software is abandoned, it should be distributed freely.

While the whole OS may be abandoned, parts of it may still have commercial value.

Imagine you’re a subcontractor who wrote, for example, a PRNG driver, and licensed it to Microsoft so they can include it in Windows XP. Does the subcontractor lose their rights to the driver just because one of the OSes happened to become abandoned?

Why would those rights be tied to commercial value at all?

Great metaphor

"Here’s where the metaphor really breaks down but stay with me: imagine if you wanted to break into a house but you didn’t know how. If you knew they had to open their windows when they baked, that would help you sneak in."

If someone can successfully build this jalopy, it would be nice to get a version compiled with optimizations for modern processors (>= core2). Maybe run the address/memory sanitizer over it.

> Torrent Size : 42.93 GB

well, ok. let's do this.

It's like leaking the plans to the Death Star. Now they'll find a vulnerability.

This will be interesting. I learned a lot from the mainsoft leak.

Would you mind sharing some of those things your learned? I'm sure people here would be interested to hear about it (I know I would!).

I learned how the win32 common controls combo box worked, which was handy as I had to fix a bug related to it, how they handled compatibility between windows versions and lots of little bits of weird design info in the comments. I spent a good couple of weeks picking interesting bits and analysing them.

Back when Slashdot was still somewhat popular (early '00s), the source code of Windows NT and Windows 2000 was leaked.

I wonder if there are any consequences from a security standpoint. Does the community of hackers out there believe this will open the opportunity for new exploit?

Good ReactOS developers don’t have to guess anymore. They can use this code as a reference and build their own version better than hasty M$ Devs.

No. If any maintainer of ReactOS evne just opens the source code, the entire project is doomed to get killed by lawyers.

The Win2k source code has been leaked before, as have several WinNT sources. These leaks pose more risks to projects like ReactOS and Wine than they benefit.

How would they know that they "opens the source code"? How would they prove that in a court of law that they didn't reverse engineer it? Seems very fishy to me....

If they did, Microsoft sues, and the programmer is asked under oath, everyone knows. Unless the programmer decides to lie under oath, which will only make matters worse eventually.

If they claim they just reverse engineered the code, they're in an equal amount of trouble. Reverse engineering is illegal in many jurisdictions, unless you manage to do it in a very specific way which should still results in different code than what Microsoft wrote.

The more source code leaks, the harder it becomes to claim plausible deniability. Ethically, I think that's wrong and should be fixed, but the copyright and software protection law is against compatibility projects like these when it comes to legislation.

'The Win2k source code has been leaked before, as have several WinNT sources. These leaks pose more risks to projects like ReactOS and Wine than they benefit.'

Correct, but does anyone really care? Many of us will have died from old age long before ReactOS actually produces a UI without crashing! It seems to me that even if the team used ideas from XP, by the time they implemented them and tried to disguise the fact, that it would be so old and irrelevant no would care. Probably not even Microsoft.

I would love ReactOS to succeed but the only version that I've managed to get working on my various test machines is about four years old. All the newer versions, and like all those before my one and only working version, simply crash.

ReactOS developers have been using the old Windows research kernel leak for years as a source lol

this is going to be great for wine support! cant wait to play postal 2 flawlessly on debian!

Ha! Shame it wasn’t Win 7 that was released as I still use it on a few of my machines. ;-) Win 7 is the last MS OS that I now permit on my operational hardware. These days, Linux and rooted Google-free Android are my main OSes of choice.

Not that I care much nowadays, as XP's code release is unlikely to affect me in any practical sense other than for just curiosity to see what goes on beneath its surface.

Having experience with almost all Windows releases from the grotty, horrible Windows 1.0 of 1985 to the current Win 10, I've no doubt whatsoever the best and most practical version of the OS for me was Windows 2000, and from the numerous post I've read here many of you agree with me. (Incidentally, to be more precise, Win 3.1 for Workgroups was the first version I used in any practical sense as I found that Win 1.0, 2.0 and even 3.0 were essentially useless.)

My second choice is Win XP and it's still installed on one of my older working machines. In some ways, my next choice would likely be NT4, I had a lot of experience with it but once you got it working, which wasn't always easy, it was reasonably solid and straightforward to use. Of course, its major downfall was both the lack of drivers and not having any effective plug and play infrastructure (hence installing it was often messy and time consuming).

That said, I doubt this XP source code leak will be a unique event. I fully expect that we will see a great deal more source code becoming available from proprietary Windows-like software in the not-too-distant future—even if these 'leaks' do not originate from within the software companies in question (as all that will be needed is access to the compiled code).

There is little doubt that soon we will see competent deep-leaning AI disassemblers become available that will be able to learn how to disassemble large programs such as Windows with ease whilst simultaneously attributing its disassembled code with meaningful and intelligible code comments and descriptions. Moreover, I see little way that software manufacturers will be able to do to stop this from happening.

(I'd however add a caveat to the above, which is that whilst AI disassemblers will likely be able to eventually disassemble most existing compiled code that's been developed for past and present-day processors—from say the earliest 4004, 8085, 8051 and upwards to present-day ones as well as those likely to be developed in the immediate foreseeable future—this may not always remain the case, as it's possible that neither critical parts of compiled code nor newer specialized processor instruction sets may be available once manufacturers become aware of the threat.)

Furthermore, I'd wager it's likely that eventually AI disassemblers will actually learn on code such as this leaked Win XP. After all, what better sample would be available to a third-party AI developer or researcher to train his or her AI on than the source of the once-famous XP? None I suspect! Moreover, I'd add that I would be very surprised if companies of the likes of Microsoft and Google aren't already training their own AI on disassembling their own code. (After all, they've 'brilliant' examples to hand and a vested interest in seeing how capable AI is at the task of code disassembly—and moreover they need to know how much lead time they have before this new technology poses a serious threat to their existing products).

That said, with about a half century's worth of existing software compiled from many trillions of lines of code that will be available come the time AI is ready to examine and disassemble it, the snoopers among us will still have great scope! Frankly, I reckon the large-scale application of AI to code disassembly would be a truly great thing: just think of all the bugs and shitty bloated code we'll find. Moreover, exposing source code would put manufacturers on notice to improve their products—as not only would their multitudes of bugs and crappy code be exposed but also they wouldn’t be able to hide dirty tricks and corrupt practices as Volkswagen did with its engine emission technology.

AI disassemblers may even force many manufacturers to adopt the open-source model for their software development.

I've little doubt that over the long term code disassembly by AI will change the current paradigm of software development forever. Unlike its current undisciplined laissez faire approach to product development that encourages sloppy code full of bugs, software development will have to adapt to become a much more professional and rigorous engineering profession such as chemical or civil engineering both of which adhere to much stricter rules and disciplines—and I reckon that will be a good thing.

A more rigorous approach to software development is something I've advocated for many years—as I class myself a victim of poor software having lost thousands of hours of my valuable time—right, thousands of hours out of my life that would have otherwise been used to better purpose—resolving problems caused by poor, buggy and dysfunctional software—of which the majority thereof would never have occurred had standards comparable to those of other engineering professions been applied to software development. Given AI technology is of itself a product of software, and that it inevitably will be turned upon its own means of creation, seems to me to be poetic justice.

In the interim, I eagerly await revelations and 'meaty' pronouncements from eager XP code examiners.

Is this legit?

Link to code.

It might be convenient to have a wine distribution with some of the higher level user space replaced with components windowsXP. I wonder if Visual studio and .net would run better.

This is something that the wine devs themselves shouldn't do for obvious reasons.

Windows XP's source code will never be leaked. Don't fool yourselves. Microsoft will never disclose the source code of theirs commercial software. It's against theirs law.

'Windows XP's source code will never be leaked'

It won't have to be. As I've said elsewhere, eventually deep-learning AI disassemblers will do it for us complete with commented code!

(That's if anyone is still interested.)

Reply on news.ycombinator.com