The Year on the Net (1996) | Hacker News Reader

The Year on the Net (1996)

wsj.com

82 pointsbookofjoe45 comments

For those interested in Internet history, there's a very interesting roundtable discussion hosted by Walt Mossberg with Steve Case, Marc Andreessen and Brad Silverberg at: https://www.wsj.com/public/current/summaries/hyperoun.htm.

If you scroll to the question: "Brad -- you seem much less interested than Marc or Steve in the prospect for non-PC hardware to access the Internet. Can you really make the PC simple enough and cheap enough to make it a ubiquitous access device?

Brad's response seems to foreshadow Microsoft's notorious mobile device strategy for the forthcoming decade leading up to the release of the iPhone: "...Given the choice of a Web appliance with a sub-optimal Web experience (because by definition it can't do everything a PC can) or a PC that gives the best Web experience, plays games, runs productivity applications and everything else you can do with a PC, we think the overwhelming majority of customers will continue to choose the PC. If the Web appliance is a 10x opportunity relative to the PC, Netscape's investments certainly don't match their rhetoric. Finally, everyone is in favor of the subsidy model, but no one has agreed to actually spend their own money to subsidize devices...."

https://www.wsj.com/public/current/summaries/hypefaq.htm

"Do I Have Privacy On-Line?

While you sign on to the Internet and blithely zap and receive electronic mail, visit Web sites and bare your soul in on-line discussion groups, you are increasingly being watched and tracked. Every move you make on the Internet can be followed, and the information gathered can be used against you."

Is this really from 1996?

EDIT: Looking at the site HTML, it looks very legit, or someone remembers way too well how '96 HTML looked like.

I wonder what the doctype is supposed to mean:

> <!DOCTYPE HTML PUBLIC "-//Dow Jones//DTD DJ-HTML//EN">

What is supposed to be a copyright notice or something?

The DOCTYPE element technically describes the Document Type Definition (DTD). The HTML there doesn't mean it's an HTML document: it means the root element of the document is an HTML tag. The "-//Dow Jones//DTD DJ-HTML//EN" is a Formal Public Identifier, which specifies the actual DTD. The vast majority of sites don't actually guarantee their HTML is valid or even well-formed. In the grand scheme of things, this is most likely cruft left over from the early days of when the site was published and a lot of these things were more up in the air.

For a bit more detail:

https://en.wikipedia.org/wiki/Document_type_declaration#Synt...

It's never really gone away either. Even now, provided you take the time to learn the format, custom DTD's are still very much a thing.

http://leancrew.com/all-this/2014/09/sgml-nostalgia/

That's quite remarkable. I can only assume that, like most people writing HTML by hand, they copy-pasted the boilerplate from some other page without really understanding it.

I can find no other references to this on Google, the best I can do is: http://xml.coverpages.org/mdml.html which links to http://xml.coverpages.org/dow-mdml-dtd20010105.txt : that refers to

<!DOCTYPE mds-results PUBLIC "-//Dow Jones//DTD MDML//EN">

The original Dow Jones page from that era according to the Internet Archive (and this is an amazing thing in itself!) just has <HTML>, no DOCTYPE. In 1997 it acquires <!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML//EN">

Yes, it's legit. See their more detailed story: https://www.wsj.com/public/current/articles/SB84858592582314...

Which references "Lexis-Nexis P-TRAK": https://www.alternet.org/story/6981/lexis-nexis_backs_down_o... (2000)

I remember this era of the Internet, and it's astonishing how little the debate has progressed on so many of the subjects. The EU has chosen in favour of privacy with data protection, boosted recently by GDPR. But state surveillance is still an issue, huge insecure private databases are still an issue.

People are still arguing about censorship: https://www.wsj.com/public/current/articles/SB84838144911430...

Broadband is still inadequate: https://www.wsj.com/public/current/articles/SB84811906424727...

The SEC is still trying to deal with fraudulent securities, most recently in the form of ICOs: https://www.wsj.com/public/current/articles/SB84953719050089...

People are still conspicuously flouncing off the internet when they don't get what they want: https://www.wsj.com/public/current/articles/SB84963869413930...

I mean, yeah. If you ran your company's mail server it was trivial to read everyone's mail, for example. The reason you didn't is the same reason you don't do the snooping you could do today -- it's too boring.

However, automated data collection has gotten far more advanced.

> it's too boring.

I don't know about that...I once had a compliance officer at a hedge fund I used to work at tell me to my face that any time he was bored, he'd open up my chat logs and follow the links. Most of the links were technical, some were memes, a few slightly unsavory variety (nothing nearly as bad as what floats around on a trading desk, though).

If you think that's prescient, take a look at the Privacy Protection Study Commission's Report to Congress twenty years before that in 1977:

https://www.epic.org/privacy/ppsc1977report/epilogue.pdf

https://www.epic.org/privacy/ppsc1977report/appendix5.pdf

Or Paul Baran's testimony to Congress in 1965:

"The initial questions are those of examining the proposed central file system, considering its weak spots, and creating a precise description of those safeguards required that are technologically and economically feasible. If the gap is too great, then clearly we should not build the system; But as a practical matter--we should realize that eventual development is almost inevitable. We would do well to concentrate on the more constructive and larger issue of how shall we control the development of the automation of all sensitive information files in order to [...] protect the rights of the individual and avoid a “1984” nation?

This may sound pessimistic, but if one can save money automating a data system it is only a matter of time until it happens. The only questions are: when and how? I would like to add the further distressing thought that we may already be well-along in the creation of the very system whose needs and dangers we are discussing today."

The browser wars link is great. It's crazy to see how fast Microsoft moved on Netscape back in 95. Hard to imagine such a short release cycle for ANY company today, let alone Microsoft.

> Microsoft has come a long way in a short time. The first version of Explorer was released in August 1995, for use with what was then the company's on-line focus, the Microsoft Network -- a proprietary network that was refocused toward the Internet in February. Explorer 2.0 was released in December 1995 and widely derided as an also-ran by the on-line community. But such critics were silenced by the quality of Explorer 3.0 [released in August 96].

The initial iteration of MSIE was based on the existing Spyglass Browser (see https://www.nytimes.com/1998/03/02/business/spyglass-a-pione... ). Spyglass made a deal with MS to get a % of the sales of MSIE. Unfortunately for Spyglass, MS gave IE away for free.

Wow, I never heard of this story so far. Back then, I was naive and young and believed Internet Explorer to be the better browser ;-)

It was a better browser. There was a story in one of the major PC magazines about how badly Netscape crashed on every platform. It use to be a point of nerd pride how well your operating system could handle a Netscape crash. It would crash the entire MacOS.

By the time its descendant Firefox came out, it was a much safer, stable, standards-conformant browser. Still have memories of just how complacent MS got when IE6 was released.

Ahh, classic MacOS. Those were the days. I once locked the entire system so hard neither mouse nor keyboard responded.

(14 year old me had opened the System file with ResEdit. Turns out opening a System menu resource for editing while it was in use was a really bad idea).

First proper computer. Learned a lot with it.

> I once locked the entire system so hard neither mouse nor keyboard responded.

I did that just the other month on a modern PC using WSL on Win10. Turns out McAfee leaks memory like a sieve when faced with WSL, and it leaks it in a way that the memory is never even returned to the OS. Worse, yet, the process using the memory doesn't even show under Task Manager! The memory just appears to disappear.

Never noticed at home as I don't run McAfee, but at work had to. It is very quick & easy to fully exhaust memory building even a moderately sized C++ project under WSL. I had hard-lockups several times before I was able to determine the cause, with help from the MS team working on WSL. I had to do a hard restart; there was no way to restart cleanly. Especially problematic when I first ran into it working remotely and RDP'd into my work desktop.

Disabling McAfee on my work PC resolved it.

(edit: add quote to what I was referring)

It was a better browser. It’s fashionable to hate on IE6 these days but it was the first browser fast, stable and feature-rich (XmlHttpRequest) for serious use. That’s why it became so ubiquitous and hard to unseat.

Also IE could print. Netscape crashed hard talking to a HP laserjet running on a Netware network running on either 95 or NT4.

When I first started running www.ibm.com in 1994 the only realistic browser options were Viola and Mosaic. IBM WebExplorer (OS/2 only at first) showed up and was fairly good but quickly lost out to the Netscape betas. While I was aware of Spyglass I can't remember ever using it.

When MSIE first came out I was explicitly prohibited from using it (typical of IBM at the time). Netscape gets slagged for becoming bloatware but the initial versions were fairly decent and cross–platform. IBM WebExplorer was genuinely not too bad, but it became a victim of IBM internal politics and died a slow death of neglect.

True tidbit which I have no public sources to cite: IBM WebExplorer was the first browser to support scripting, albeit in Rexx, IIRC a couple of months before LiveScript came to Netscape. IBM, in its wisdom, decided not to release that browser, I believe out of concern it might undercut the market for Lotus Notes.

Note: I could be wrong, but I believe LiveScript was what JavaScript was initially called. I realize there's something now called LiveScript which is completely unrelated.

I was first online during the summer of 1993. I remember using a browser called Cello, which was developed at Cornell Law School. I suppose it's name must somehow be inspired by the name of the Viola browser that you mention. I did not know about Viola at the time.

I also remember using a shareware winsock implementation to get online, Trumpet Winsock.

I remember Cello, I think I even tried the beta port of it to OS/2. IIRC it died fairly quickly after Mosaic appeared.

TCP/IP on OS/2 2.x and Windows 3.x was provided through shims or DLLs. I think out-of-the-box OS/2 came with no networking at all (you could add SNA/APPC) and Windows had a couple of TCP/IP packages you could add. Neither operating system was prepared for the explosion in TCP/IP networking in 1994-1995.

yes, and the millions poured by Microsoft into advertising. magazines were filled with stylish internet explorer ads i.e. one-color muppets characters ads in rolling stone.

Found this great Marc Andreessen keynote from the Netscape Developer Conference in 1996. Netscape wanted to be everything simultaneously. Youtube, Slack, Gmail, Amazon, AWS, Dropbox and about a dozen other businesses. Wonder if they just focused on one product. Say "groupware for Fortune 500 intranets". They would have increased chance of survivability. Even considering Win98 bundling its competing offering in their OS ;)

https://www.youtube.com/watch?v=e8wdc9qkkV4

Wonder if they just focused on one product. Say "groupware for Fortune 500 intranets". They would have increased chance of survivability.

Others think that it was one of the causes of their demise. I've deleted the link to jwz's article on Groupware because links from HN seems to be unwanted for some reason, but anyone interested surely can google it.

Yes, jwz has a massive hateboner for Hackernews.

I had guessed that much :) Oh and I can totally understand it. I just wonder if there was a specific incident that sparked that.

Did Netscape really die, or am I using its (wonderful) offspring to post this comment?

The company is dead and almost all the code has been rewritten. Is it your grandfather's axe if you've replaced both the blade and the handle?

At least about:mozilla still works.

Is it your grandfather's axe if you've replaced both the blade and the handle?

I take it you're opposed to prison sentences lasting more than seven years?

>Is it your grandfather's axe if you've replaced both the blade and the handle?

It might not be /his/ axe, but it is one that he /created/.

Yes.

Firefox is to Netscape as Teenage Groot is to OG Groot.

It's 22 years later. The presentation is infinitely better. The content is infinitely worse... or at least the content that greets the average net user.

You really have to dig for good quality content today.

It's disgusting how bad its gotten. It's assuredly partially nostalgia, but I think also partially truth, but the internet of my younger years I remember as being filled with labors of love, pasionately created content most places you look. Now it's difficult to find much that isn't lowest common denominator pandering or essentially white noise.

I don't think it's just nostalgia and survivorship bias. The median quality of content on the Internet has become objectively much worse, mostly due to a race to the bottom and a race to make the most "viral" thing to cash in quick. Media has always been a bit of a deflationary race to the bottom but this is quite extreme.

Some survival bias maybe, but I agree, content was better, and I believe mostly because there was more work put into the content. Today a lot of work goes into presentation, into using words in common enough dictionary ( see https://blog.xkcd.com/2015/09/22/a-thing-explainer-word-chec... )

But don't forget that there was less content, and soon the era of blogs rose, for which I read a fascinating description lately:

"Everyone said too much and said it poorly. It was incredibly entertaining." / http://nymag.com/selectall/2018/05/i-dont-know-how-to-waste-... /

There are still plenty of labors of love all over the place, but my own personal experience is that if you create something as a sincere act of self expression or love for [x], people don't bother with it. I think a lot of what the internet has turned into has also primed us as users to seek that sort of stuff out. I've been trying to distance myself from a lot of services and the meme-o-sphere in favor of occupying the types of sites that used to matter to me (personal web pages, no social media, a couple blogs, rss from a few news sites, SLSK). I have partly HN to thank for this, after reading everyone's disgusting unsympathetic responses to Seattle's "homeless problem," I've distanced myself from the site and don't regret it one bit.

Edit: it really did feel like we were on the verge of the future, but the rise of communications technologies has hobbled all of us, and Idiocracy may be the only future that's possible if we don't slow the ocean currents to a halt first.

As to the presentation, I wouldn't go so far as saying it's "infinitely better." I think there's a lot of over-designing these days. Many websites are too flashy, distracting, complex.

As to the content, there's the good and the bad.

First, the bad: I feel like the websites on the Internet of yore were akin to the counter-cultural zines of the pre-digital era. They were made by passionate, creative, intelligent individuals and consumed by the same. Modern websites are more like mainstream magazine-tabloids: a strange blend of glossy, clickbaity journalism that prioritize going viral - or, at the least, being provocative - over everything. E.g., even as a liberal, I find myself groaning whenever I read slate.com.

The good is that there are places like HN, through which we can find thousands of high-quality sites - many of them created by individuals like the creators of those old school zines. (There are also places like reddit, which foster all types of communities, but I'm a bit conflicted on that website.) Anyway, if anything, the good content is out there, we just need places like HN to help us find it!

This timeline is wonderful: https://www.wsj.com/public/resources/documents/hypetime.htm

The IPO of Yahoo, the creation of MSNBC.

I read through most of the content, and the quality of reporting was top-notch. The browser wars article, the reporting on the state of online privacy, the discussion of IPOs... all of it turned out to be the major important topics of the early web. Great post, OP!

Thank you.

I love this! It's prophetic

This reminds me of mirc 16bit for win 3.1

Reply on news.ycombinator.com