I was trying to post a comment to Stan Beer’s article Microsoft promises IE 8 Beta 1 with a smiley face in 2008, but I ran into the dreaded “There is a problem with your post – Try removing some punctuation, sorry” error so I’ll just post it here; its big enough to be a full blog post anyway ;)
“It really should be an embarrassment to Microsoft that in the Internet Age, the company is taking almost as long to develop new Web browsers as it takes to develop new operating systems.”
Eheh… Actually that’s not an accident; I’ll explain.
Recently, a Microsoft employee released a report saying Internet Explorer was more secure than Firefox – his claims were quickly discredited (e.g. see links on Asa’s blog post) but I had a look at the original and the thing that jumped out at me right away was this:
Microsoft generally releases a browser in conjunction with a new operating system release and commits to supporting that version for the lifecycle of the product – now 10 years for business products. [...] Internet Explorer 5.01 SP4 is also still supported for those Windows 2000 users that have made the decision never to upgrade their browser to a different release.
Yes, seriously it says that; Microsoft still supports IE5!
I almost didn’t believe it, but I had the opportunity to test Windows 2000 recently and after installing Service Pack 4 and doing the updates, sure enough, there’s IE5. They WANT you to upgrade to IE6, calling it an important security update and microsoft.com doesn’t render well at all in it, but they still support it. Why?
corporate enterprises [...] sometimes have custom web applications and are hesitant to upgrade between major releases very often, and even then may have a relatively long transition plan.
“custom web applications”… is that a euphamism for ActiveX? (Update March 5: Ah, I get what “custom web applications” are now; basically forms and such that only render properly in Microsoft’s proprietary rendering engines because they either are buggy or because MS intentionally added proprietary ‘extensions’ to the web standards. Welcome to vendor lock-in.)
Another critical thing to note is the ‘look and feel’ of IE 5, 6 and 7. I blogged about this a while back; basically:
IE7 has Vista’s look and feel, regardless of if its on XP or Vista.
IE6 has XP’s look and feel, regardless of if its on 9X/2K or XP.
IE5 has the native 9X/2K look and feel.
Amazing that 3rd party browsers like Firefox and Opera can look like native apps regardless of their OS. As I noted: “So not only is IE7 an upgrade, its also a sales tool for Windows Vista, just as IE6 was an upgrade and also a sales tool for Windows XP!” (And it seems that Safari for Windows is a similar sales tool).
Between these two items (and anecdotes like this), we can see that Microsoft has historically not cared about the web any more than it was a good vector for its products. Of course, they will not admit this, making comments like:
Mozilla released Firefox 1.0 in November 2004, Firefox 1.5 in November 2005, and Firefox 2.0 in October 2006. Only Firefox 2.0 is currently supported with security fixes from Mozilla, as it is has been Mozilla’s policy to support a previous version for six months after a new (major) version is released. So, according to its original schedule, Firefox 3.0 was scheduled to ship in November 2007, which meant Firefox 2.0 support would end in May 2008 . To put this in perspective, if Microsoft had this same policy, then support of Internet Explorer 6 would have ended in May 2007, or similarly Internet Explorer 5.01 support would have ended in 2001. [...] shorter lifecycles mean more people may still be running an unsupported version and be exposed. To explain this comment, let’s look at an example using Microsoft IE6 SP2. Imagine that after IE7 was released last October that one month later support for IE6 would end. How likely is that everyone will have upgraded by the end of that month? What if it was six months? Is it likely some consumers or companies might not have upgraded by the end of the six month grace period?
The funny thing is that Firefox, which does not seek to hook itself deeply into an OS, has never had DRM (WGA) interfering with upgrades, tries to fit in rather than garishly proclaiming that its time to spend more money and (most of all) just wants to be a web browser, does not have the massive problem with people lingering on old versions that IE does!
I direct your attention to a stats page that shows a breakdown of FF versions in use, like for webreference.com While it unfortunately doesn’t seem to cache old numbers, I have watched rapid (less than a week) transitions to the latest version of FF. Here’s a sample of the stats from Dec. 21:
FF ALL 30.18%
FF 188.8.131.52 24.66%
(ALL other individual versions <1% each)
I did the sums on a spreadsheet and (after removing a couple possibly spoofed versions for a total FF share of 30.02% on that site) here’s what I got:
FF 0.x 0.03%
FF 1.0.x 0.31%
FF 1.5.0.x 0.82%
FF 2.0.0.x 28.46% (less 24.66% 184.108.40.206 = 3.8%)
FF 3 0.4%
Excluding development version FF3 for a total of 29.62%, the fully up-to-date Firefox ratio is 24.66%/29.62% or almost exactly 5/6 and as per http://www.mozilla.com/en-US/firefox/220.127.116.11/releasenotes/ the FF 18.104.22.168 point release was made on November 30, 2007 so only three weeks prior to these stats!
IE ALL 38.16%
IE 7 17.24%
IE 6 18.62%
IE 5 2.11%
IE 4 0.17%
I think only in Microsoft’s wildest dreams do they hope for 5/6 of IE users to be on version 7 any time soon.
And because Microsoft wants to do its ten-year-plans and support the “custom web applications” I can’t see IE 8 breaking the mold. The latest batch of web standards compliance, very late indeed and only limited to an internal development version (with Microsoft, and I have but to point at Vista for the latest example of this, never believe its anything other than vaporware until they’ve actually released! ;), is a good thing for the web, but not really for Microsoft; in hoping to stop the exodus from IE to FF by adopting real standards to make it more developer friendly like Firefox, they have now lowered the bar for people to migrate from Windows to Mac OS and Linux where IE doesn’t exist. Oops :)
Bonus Link: Microsoft has an insanely large (almost 600 MB; its over half an hour long @ 640×480) video in WMV format here where the IE team is interviewed about IE8. The first little bit is a neat time-lapse where the Acid test face slowly forms as the rendering engine gets better over time until it actually works.
Update March 4, 2008: It looks like IE8 will have three ‘standard‘ rendering modes, equivalent to the standard modes in IE6 and IE7 and a third even better (as in more standards compliant) one new for IE8. [Update March 10, 2008: Though its still FAR from perfect... including the addition of "quite a few evil things"] Now Microsoft had said that they wanted to make the IE7-type the default which drew howls of outrage and isn’t making them any friends in Europe where they hand out stiff fines for monopolists that don’t play nicely with others. So they backed down. To quote this page though:
I’m glad that they’re going to make “standards” mode standard.
I just wish they were doing so for the right reasons.
It really seems to me that M$ is out of touch with the web (or worse, thinks they are in charge of everything because they are the great and mighty Microsoft!) and goes along doing its own thing without thinking aboout how it affects others (or worse, intentionally thinking about how it will hurt others) until someone hits them in the back of the head with a cluebat. Its sad to think of a giant company as a spoiled child who won’t behave unless rules are imposed on them, but that’s Microsoft (and its glorious leader Steve Ballmer) for you.
Update March 5: There’s an interesting article here that suggests the “real reason” IE8 will render in a more-proper standards mode is that “the mobile web is coming” and that if MS doesn’t start using real web standards now, they’ll get left behind when a tsunami of people using cell phones, etc. to browse the web can’t view IE7-targeted sites properly. I rather suspect the aversion to punishment I outlined above rather than foresight in Microsoft’s decision, but its an interesting notion. Oh and BTW, I mentioned Ballmer yesterday; looks like he wants to stay on as head of MS until 2017! I suspect that shareholders will boot him before that! ;)
Update April 17: Regarding “Imagine that after IE7 was released last October that one month later support for IE6 would end. How likely is that everyone will have upgraded by the end of that month? What if it was six months? Is it likely some consumers or companies might not have upgraded by the end of the six month grace period?”
Let me give another example of how extremely large software populations can upgrade rapidly, even if they are proprietary: Adobe’s Flash. According to Adobe, “over 98.8% of Internet-enabled desktops” have Flash on their computers. There is a graph here that shows, as of September 2006:
“Before Flash Player 8, adoption typically took 12 months to hit 80%. Flash Player 8 hit 86% in 9 months – with so many sites using Flash and the new auto-update feature. Flash Player 9 is already to 50% adoption in 3 months”
So yes, it can be done… maybe it’s just that Microsoft can’t.
Update July 2: Firefox 3 was released on June 17; Asa Dotzler informs us that there were:
8 Million downloads in the first 24 hrs
12 Million downloads in the first 48 hrs
20 Million downloads in the first week
and Mozilla did NOT turn on the auto-update feature for FF2 users yet!
Firefox 3:2 ratios are quite good; on the pcpro.co.uk website, there were seven FF3 users for every five FF2 users only TEN DAYS after FF3 was released. An insightful quote:
“So Microsoft still has three out of ten people running an old version of its browser more than 18 months after Internet Explorer 7 launched, while Firefox has converted more than half of its users to the latest version in just over a week. That should set a few alarm bells ringing in Redmond…”
Another site, webreference.com shows almost 1:1 usage after two weeks (for “Tue Jul 1 23:50:07 EDT 2008″ there were 14.93% of visitors using FF 22.214.171.124 and 14.37% using FF 3.0). XiTiMonitor has an article up (Google translation; update: official translation) showing the immediate spike in European FF3 usage following its release.