Threads for bkardell

  1. 7

    That’s a very odd history. It asserts that browsers have never been sold, yet skips over companies that sold Mosaic (ISPs generally bought it and gave a copy to customers, but some didn’t). Opera was still selling their browser directly to customers until 2005, though with an ad-supported version available for the last few years of this. Even Netscape was only free for non-commercial use, they were selling directly to business users.

    Most of the browser wars, which ended up shaping a lot of this landscape, happened in a single paragraph of the article that completely skipped over all of the details. In particular, Microsoft’s decision to bundle IE (which was later found to be anticompetitive) made it impossible to sell a browser to Windows users. This was before the mobile web was a thing (remember WAP?) and back when before Steve Jobs’ return to Apple the Mac market share was approximately five people (KDE hadn’t even reached Alpha, desktop Linux was xterm and AfterStep), so Windows accounted for almost all of the client market.

    1. 4

      It seems to me that the article is indeed lacking a lot of nuance. Opera did, indeed, have a semi-subsidised model for a while – after 2000 or so it was ad-sponsored, it displayed some banners in the upper-right corner (and, later, it allowed you to see either that or some personalised ads from Google) which you could get rid of by buying it. But prior to that it was shareware.

      Things were also very dynamic at the other end of the web spectrum – web searching, for example, was itself a highly subsidised product. I think some super early search engines had some paid perks but by and large it’s always been “free”. AltaVista’s results were generally worse than Google’s, sure, but they weren’t exactly useless – the game changer was not so much that Google’s results were orders of magnitude better (although in many cases they were), but that Google figured out a way to turn that into a cash source to fund further development and make it even better. AltaVista, AllTheWeb & friends ground to a halt eventually, as the kind of money they got from e.g. search deals with Yahoo just wasn’t enough to keep them at the top. If Google hadn’t figured out the ad trick, it’s likely that they’d have joined the history books after a few years, too.

      A lot of the Internet ecosystem is effectively “subsidised”, the way the author puts it (correctly, otherwise – maybe the history part lacks some nuance but I think they’re pretty spot on about the current state of affairs). IMHO effecting change in this regard at the browser end is not just rather hopeless at the moment, but I also don’t think it would have the dramatic effect some people hope it would have.

      (Although the quality of browsing might change tremendously. On my macOS machine I use the Orion browser and, even though it’s still in beta, I honestly never want to go back to anything else ever again unless someone somehow resurrects Opera.)

      1. 2

        On my macOS machine I use the Orion browser

        Ooh in addition to being webkit it also has built-in tree style tabs. I guess I’ll be trying this thing out.

        1. 1

          Oh man this browser is amazing… except that the 1Password extension doesn’t really work rendering the whole thing unusable for me :(

        2. 1

          There are endless possible details I could have gone into here, for sure. FWIW, this piece was originally considerably longer, and included things about various models that didn’t win the day, or even remain viable - because I am a big fan of history and nuance… But in the end many of those didn’t actually seem to add to the story as much as spiral into side-stories and confuse the larger topic. I did my best to stay truthful here and suggest that we did have various flirtations with lots of things, but none of them were just sell the browser, the way that Microsoft just sold word, for example.

          FWIW, altavista was my engine of choice before Google came along and really I do feel like Google’s approach of basically indexing everything and introducing Page Rank was an incredible game changer. Its definitely true that them finding a successful monetization strategy created a real feedback loop to make it get even better faster though.

          1. 1

            Oh, hey, I know it’s hard to strike the right balance between offering all relevant details and sticking to the point. When I said the article was lacking nuance, I didn’t mean to say it wasn’t truthful and especially not that you were deliberately lying. In case anyone’s still reading I’d like to clarify this is absolutely not the case :-D.

            IIRC there were companies that tried to “just sell” the browser, but this was the early ‘90s and there were many different approaches to “just selling” it. Microsoft never needed to offer trial or restricted versions of Word in the 1990s. But before they could fund it with ads, Opera or OmniWeb were shareware, and IIRC early Netscape Navigator versions were offered for sale to some segments, too (but that was before my time, I’m likely wrong on that – their model was definitely subsidised later on). Earlier on, lots of companies, including big names like Fujitsu, tried to sell licensed repackaging of Mosaic (I think they sold at Infomosaic or something like that). Of course, with few exceptions, I don’t think there were any companies that only sold browsers – but then again, it’s not like Microsoft only sold Word, or Sun only sold Solaris and so on.

            As for Microsoft’s distribution model – I’m not sure if it was that clear that Netscape’s model was failing when Microsoft bought Spyglass Mosaic in 1994. Their decision to bundle it for free may have been a genius move from someone who’d seen the value of the subsidised model, though when they started doing that, in 1995, Netscape wasn’t really failing, either. But the fact that, by bundling it for free, they also avoided paying royalties, may have played a role in this decision, which later turned out to have other useful implications as well.

            This is the kind of nuance that I think was lacking and tbh now that I read all that, I can see why. I definitely don’t want to dispute the claim that web browser development was (and is) heavily subsidised starting with IE 4 or so, that would obviously be silly. I just think a) a lot more exploration happened before we ended up with this model (and we may have ended up with it by accident, to some degree) and b) that, when it finally happened, it was part of an already growing trend, as lots of stuff on the web was (and continues to be) subsidised, for better or for worse.

            1. 2

              It’s ok, no offense taken or anything - I don’t generally like to wade into comments on my own posts - I just wanted to say “you’re not wrong about memories of various kinds of attempts” etc and that this piece could have been longer - it was hard to not make it longer, in fact :). But, yeah, if you like this kind of history there’s a bunch more of it on my blog and we have a podcast too - this episode https://www.igalia.com/chats/ecosystem-health-iii talks with opera folks and gets into some of it, I think it’s a kind of fun listen. Project Code Rush is a documentary available on YouTube too with a bunch of the stuff about Mozilla founding - if you haven’t seen it. I’ll also be giving a long form talk on all of this at Web Directions Code in the fall, maybe an opportunity to say more there

              1. 1

                Oh, cool! Web dev hasn’t been my thing for a very, very long time now (holy crap it’s been 15 years!?) so I don’t really follow Web Directions Code, but that’s a talk I’d definitely wanna watch!

        3. 1

          KDE hadn’t even reached Alpha

          Not sure what year you’re trying to reference here. KDE was released in 1998 and that’s the year I came online so I kinda remember that time. Netscape 4 was the cool kid over IE as long as IE 5 wasn’t released, then they were kinda leading again for a while. Then I don’t remember the timeline 100% but I think it stayed at the helm until early Firefox came into the picture (guess we can’t count Phoenix here yet).

          1. 2

            Parent’s timeline mostly checks out IMHO, at least as far as the Mac marketshare and KDE are concerned:

            This was before the mobile web was a thing (remember WAP?) and back when before Steve Jobs’ return to Apple the Mac market share was approximately five people (KDE hadn’t even reached Alpha, desktop Linux was xterm and AfterStep),

            Jobs returned to Apple in early 1997. “Before that” would be 1996ish, and the first KDE announcements were in late-ish 1996. 1996 was definitely peak AfterStep with xterms :).

            IMHO the wider assertion about the effects of bundling IE with Windows 98 is spot on, too, although I think that was about an year and a half after both Jobs’ return to Apple (not that Apple’s marketshare wasn’t negligible at that point, either) and KDE’s first alphas :).

            1. 2

              Yeah maybe I indeed missed that part but I don’t think I heard 96-97 ever called the browser wars, I thought that was only beginning in 99, but it’s been a while and maybe my interest simply wasn’t focused on web dev before that.

              1. 2

                Eh, periods and wars are shady like that :). But IMHO it’s definitely “browser wars” period. IE 4.0, which prompted the infamous “From the IE team… We Love You” “prankcident”, was released in October ‘97, and was already a part of Microsoft’s web browser strategy.

        1. 4

          It gets worse. WebKit has specific quirks hardcoded for specific sites like YouTube, Google Maps, Twitter, and the NY Times. So websites can behave differently depending on the browser, and browsers can behave differently depending on the website. It’s a mess!

          Imagine being a developer at a company whose web site behaves differently when it’s running on a development server compared to a production server, with absolutely no changes in the code. You might not even know WHY your site is singled out since WebKit bugs are mostly kept secret.

          (Firefox and Chrome do not seem to have any hard-coded behaviors for specific sites though.)

          1. 2

            Yeah I mentioned that in the post I thought.. Definitely in the linked podcast. I have been that person. Firefox does and multiple chromium browsers at least do