ntauthority

It does, under the accessibility options. 'Contrast themes' allow fairly fine-grained customization, still, and a lot of legacy values that have been there since Windows 95 are still supported otherwise to some extent albeit sometimes hidden.

This would be fine if not for even the canonical C/C++ compiler on GNU/Linux having issues building older versions of itself without patches, let alone other applications.

As with older Windows binaries on newer Windows versions, when it doesn't work outright, there's always fixes - but that's not the point this subthread tries to discuss.

A more interesting variant of your question would be 'how come Wine is better at running decades-old Windows software (on systems such as Linux) than Linux is at running decades-old Linux software on Linux?'

... perhaps Win32 is a more stable ABI/API by design.

> (f) presents a risk to consumer safety or (financial) wellbeing

By some measures, PayPal itself does. Should PayPal ban itself?

(for example, anecdotally from an acquaintance in Germany a few years ago, there are some flows that can and will, without the appropriate checks, lead to a customer ending up in a debt to PayPal, and they have some of the worst debt collectors in this regard)

This 'modern' Notepad is actually not 'a wrapper around a Win32 edit box', it's (likely to support additional use cases like macOS TextEdit in the future - most the modernization seems to be to haphazardly 'compete' with arbitrary features from macOS) actually a variant of the Office rich text control, which in some cases actually is more performant.

The WinUI parts and initial launch cost, however, are somewhat hard to forgive.

Somewhat regrettably given the connotations also associated, cryptocurrency. In countries where there are legal, regulated exchanges, this pretty much meets the requirements there.

MRT is the Windows Malicious Software Removal Tool which has been distributed via Windows Update since around 2005.

My favorite issue with phone number-based identity verification is that it doesn't account for multiple different countries well at all. In some regions, getting more than one 'legit' phone number (i.e. not on any of the VoIP blocklists) per person is nearly impossible, while in other regions, even in the Western world, it's very much possible to buy a lot of pay-as-you-go plans, keep the numbers from expiring, and use them to retain perhaps dozens of accounts at no to minimal cost, making it yet another measure that affects legitimate users disproportionately compared to it affecting illegitimate users.

For Windows, there once was a similar tool that synthesizes a .pdb file using the DIA API from a list of names: https://pefrm-units.osdn.jp/pe_debug.html

> I’m touching lightly to scroll, but pausing for a moment to finish reading

Just checked on my own behavior in this regard, and I seem to have grown a habit of just scrolling slightly enough to not make it appear to be a long press before actually resting my finger on the screen.

This might be a more common habit and would explain the consensus of long presses not being seen as too problematic.

Anecdotal point of data: M247 seems to run a lot of bad-faith traffic as well - while a service I run tries to keep block lists minimal even for frequently abused endpoints (eg credential stuffing) their ASNs are a mainstay in there.

The 'normal' system being complex is a bit exaggerated here - if you have to travel one-off via local multi-mode transportation you'd usually just get a regional Deutsche Bahn day ticket which will generally by accepted by local transport companies as well, and I believe various variants of these would also exist for longer-timespan tickets.

Still, that would not be as handy as the mostly unified billing system using NFC cards we have in the Netherlands since the late 2000s, where if you do not have any entitlement valid on a particular mode of transportation, it uses a charge card balance as such.

While this isn't as innovative nowadays as it used to be, it's still notable as it works in the entire country, which, while small, is still not a city-state.

By 'affectionately named', I rather think someone thought it sounded cool to name it such. Perhaps CloudCore proper allows 'upgrading' a running system to the 'Win3' composition in case one would want to run GUI diagnostics, but other than that the run level terminology doesn't mean much as it's rather a distinction made at image staging time.

FactoryOS is from an entirely different set of compositions (composable core/WCOS, meant to reduce the amount of MobileCore-based SKUs) which was meant for a few new products (10X, HubOS, Andromeda) that never really shipped and got changed around and canceled numerous times. It is curious, however, as to why there is no WSK released for the 'Nickel' release of Windows (with build 22621 as final), and instead we got this image dropped instead.

The VM was rather an image of a similar composition to this one (ContainerOS shares some component roots) running in the Host Compute Service which actually is properly documented nowadays after a while of only being partially documented in Docker for Windows code: https://docs.microsoft.com/en-us/virtualization/api/hcs/over...

It's a different composition, this seems to be a derivative of some 'CloudCore' product used internally with Azure, with layered compositions affectionately named 'run levels' (the full GUI boot also known as ClientCore is apparently '3', but '2' would be a console-only boot like Nano Server).

There's an interesting 'hello and enjoy' text file left in here describing this run level:

               HELLO!
  
  The OS is booted to RunLevel 3 (CLIENTCORE).
  
  Services are STARTED at this run level.
  
  All OneCore compliant apps can run.
  GUI apps (that depend on user32 and gdi only) can run.
  
  Run 'memstat' to see memory usage.
  Run 'tlist -s' to see running processes and services.
  
  'njoy.

With Chrome, it's taking away. With Safari extensions, it's adding functionality that didn't exist before.

Compare 'some ad blocking' to 'no ad blocking at all' - you go from 100% of ads being shown to maybe 5-10% (and a lot of broken sites due to less advanced rules). Still not the ~0% people are used to on an open platform, but the iOS default was worse.

The verbatim email:

---

Our limitations due to new United States sanctions

Dear sir,

Due to recent sanctions from the United States government, we have reviewed the organizations we have been communicating with to understand to which organizations the new laws apply.

Our understanding is that your organization is located in either Russia or Belarus. Because of this, we must pause interactions with your organization until the sanctions are lifted.

This includes: We must pause on selling, issuing or renewing a Mattermost subscription or license key to your organization. If you have a Mattermost license key, a trial key, or a Not-for-Resale license key we ask that you not use it. We are not permitted to provide you technical support for our software. Moreover, we cannot answer questions about our software. We are not permitted to provide our software, including any updates, upgrades or cloud hosted versions to your organization. We must ask that you do not use any software or online services provided by Mattermost, Inc., including any versions that have been provided online to the general public. We deeply apologize for the inconvenience. We must abide by United States laws. We hope after sanctions are lifted that we can support your interest once again.

The sanctions applying here can be found online at: https://www.bis.doc.gov/index.php/policy-guidance/lists-of-p...

If you believe we have made an error, we ask that your legal counsel please review the U.S. sanctions rules and send their interpretation to compliance@mattermost.com

Sincerely,

Mattermost Export Controls Compliance Team

---

Counterpoint for this rant (showing it's again not very objective, and just what your or your environment's expectations are): it's easier to cross-compile binaries for Linux (using Clang and a sysroot on Windows) than it is to natively compile on Linux (not using a sysroot, as that is the 'default' flow there) if your environment is mostly Windows already. We do this for one of our products, in fact.

The post meanwhile seems to be part about open source policy (nasty, but not a technical issue - also including a new-to-me and hard to find on a Google or DDG search fact about the VS gallery endpoints having led to legal threats), part issues induced by 'weird' modern languages not caring to support Windows (as expected?), part Git for Windows not bothering to handle symbolic links cleanly (and the weird admin-only default thing that remained from Vista), and part... concerns where if people would spread the proper way of doing stuff (curl.exe bundled by default for downloads, or long file names being 'weird' - albeit less broken than claimed here) instead of just ranting this'd be fine too... but not really a coherent whole.

Every existing library follows the standard which specifies host byte order, which usually means little endian. The Rust library cited a few levels up this chain ignored that, somehow assuming big endian, and then had to correct for this mistake.

Even those points can be argued around in various ways, so they're not even that clear-cut.

> uniform, comprehensive, robust package management, including for the system software

Every Linux distribution has its own package manager, so that somewhat fails on the 'uniform' point. Package management for system and apps being shared (comprehensive) is an arguable point as well, and 'robust' seems to not be the default on more traditional Linux distributions.

The approach seen in for example Fedora Atomic matches the macOS approach a lot more, including rollback and consistency, but also splitting the system image, overlays, and user apps.

> GNU coreutils

Not all Linux distributions have these, either, busybox distros seem fairly common these days (and offer most the popular switches you'd see with GNU utilities).

In addition, of course, the GNU utilities run under any UNIX-like, so if you do need any of the convenience switches you're used too, they exist anywhere as well.

Again, of course, subjective.

> good filesystems

APFS seems 'better' than the ext* default seen on most Linux distributions, and has had less of a history of 'random corruption' than early btrfs.

> 'root is root', ...

Red Hat-like distributions (and even Ubuntu) seem to by-default enable some kernel security module, which have the same effect.

The implicit UID 0 bypass is nasty for security.

> lots of .. terminal emulators

I'm not a picky terminal user, but macOS Terminal.app nowadays seems to match, say, gnome-terminal or Konsole quite well in behavior.

Perhaps it doesn't support some fancy feature set, but I'm not sure if the main DEs do either.

> some choice in desktop experience

Valid point. However, if you prefer something as uniform as the macOS experience, you can't get that on generic freedesktop Linux.

> popular apps don't just entirely stop working

Depends. The typical Linux distribution has a nasty ABI as well, often only guaranteeing source compatibility.

This is indeed often easier to work around even for binary-only Linux apps (on a multilib amd64 system, you can usually run some libc5 app from the late 90s fine. on macOS, this is 4 CPU architectures ago and you won't even stand a chance).

> pretty much everything is discoverable and configurable if you're determined

Same goes for macOS and even Windows. Reverse engineering is a thing that can be done with sufficient determination, and in many legislations has exemptions for these kinds of use cases.

macOS and Windows also offer symbol names for a fair amount of OS components, it's just perhaps a little more work than 'grep the source code for a flag'.

In the end, however, it's all a matter of tradeoffs and what you're used to (i.e. won't have to waste time learning before being able to get work done), and if you're used to something that varies from the default of any system, this is less and less likely with macOS or Windows.

Yeah. The bundled ones in the main OS image are the worst. Who the hell knows what nefarious acts lie behind IOPCIFamily.kext?!

/s, though not entirely, moving more stuff to unprivileged contexts would be nice

I barely even knew Medium was a platform that had its own post feed.

I only knew it from clicking link submissions here that looked like they were from a corporate blog, but then turned out to be Medium with a custom domain, mainly as I would see a header saying 'by using Medium you agree to the privacy policy', which is a statement that is vague in a few ways:

1. Am I 'using Medium' if I'm reading a blog post on blog.somecorp.io that only says it's 'Medium' in the privacy policy header?

2. Have I been 'using Medium' by simply opening an article on it? Why have I not been given any time to not agree?

3. If I want to read the privacy policy, it seems to be hosted in a Medium article itself, with the same self-referential banner 'by using Medium, you agree to [the privacy policy]' which is a link to the same page.

4. If 2) applies, by reading the privacy policy I agree to the privacy policy. The escape in 1) doesn't apply either as the privacy policy is (or was?) hosted on a subdomain of medium.com. By trying to read the privacy policy (or interacting with the link that leads to the privacy policy?), I therefore have implicitly agreed to it.

This not only seems illegal in various jurisdictions, it's also very immoral.

Qualcomm has even ended up in a legal quarrel with Apple over a modem contract that in the end led to Qualcomm getting what they wanted.

They're a ferocious company when it comes to legal.

A year ago, I had to go to a bank office and engage in some verification process that also required me to use the physical bank card with its preassigned PIN.

No matter what I tried, I hit the 3-try limit for the day, and opted to have the same preassigned PIN sent by mail to my home address.

When walking back home from the bank branch I realized the mistake I had made: I had entered the correct PIN, but had typed it in calculator/numeric keypad order, and not in phone/PIN pad order.

While I've never used that PIN on a numeric keypad for a PC, somehow my brain associated the numbers with their order on a PC keypad, since I had used my PC keypad to unlock my PC with a PIN numerous times more than I had used any card terminal with a PIN.

So, the next day, I returned to the bank branch office to try the same operation again, and indeed - I had correctly entered the PIN and the online banking transfer limit ended up adjusted just fine.

Of note is that pairing the desktop APU with one of the higher end desktop chipsets will still have the chipset (which is just another legacy IO die) have the excessive 10-15W idle consumption it would anyway.

I'm not sure if the Bx50 chipsets are any more efficient or if that needs the absolute low end of A series chipsets (that I believe are instead some ASMedia designed IO stuff).

Comment threads everywhere (including on commits and lines of code), these comment threads then having emoji reactions, and the inability to disable any of that (same for the pull request feature, which also leads to more comment threads) can get somewhat annoying when a group has decided to target you for arbitrary reasons, and such potential 'needing to see harassment' being deeply embedded in one's code workflow (i.e. in your own repository) can get tedious.

Not everything needs to be turned into an embedded discussion forum.

First off, nobody deserves any sort of abuse.

Secondly, the Sneedacity incident was abuse targeted at a fork of the Audacity project, not at the original 'hostile takeover' project itself. The fork was intended to remove said abuse you feel is deserving of more abuse.

From the article:

> Starting today on Android and iOS globally, it’s now easier than ever to travel back in time right from your phone.

So yes, this is possible now.

I accidentally clicked this on my phone. Skimmed it. Read the last paragraph. Ignored it. Thought about it. Can confirm: it worked.

Piracy and such hoarding are inextricably intertwined. If you want a particularly uncommon variant of some pirate content (like an obscure movie in any quality, or a somewhat popular movie in original BD format), you better grab it when it's available, as further down the line getting the exact same such content may be more and more difficult.

Streaming services have a similar issue (re-encoding, bad rereleases due to licensing issues, delistings, crappy format choices in the first place) but generally a bit better SLA-like and a longer timeframe to watch certain content.