2009/2/20 Aryeh Gregor <Simetrical+wikilist(a)gmail.com>om>: There are different levels of support. We should certainly make sure things fail gracefully for IE6, but a new feature not working on IE6 shouldn't be a reason not to implement it for everyone else. (I believe that is pretty much the current policy already.)

2009/2/20 Aryeh Gregor <Simetrical+wikilist(a)gmail.com>om>: I can't see any significant new features causing a problem that wouldn't be dealt with by the "fail gracefully" condition. As long as adding the feature doesn't make things worse for IE6 users (so they can still read and edit the sites), then there isn't a big problem. Of course, if you can cater to IE6 easily, then there is no reason not to.

2009/2/20 Hay (Husky) <huskyr(a)gmail.com>om>: Are the browser stats for Wikimedia anywhere? - d.

On Fri, Feb 20, 2009 at 9:49 AM, Sherool <jamydlan(a)online.no> wrote: I would be willing to bet the vast majority of IE6 users left are on corporate networks as mentioned above, and have little to no control over what browser they use. At my former job, XP and IE6 were the standard supported platforms, and they had given no indication of wishing to change this anytime soon. For what it's worth, I have a feeling end user services wasn't going to change their mind about what browser to set as standard just because some people complained that Wikipedia told them to change. Upgrading an enterprise installation must have a good cost/benefit ratio, and it just hasn't hit that point yet (from an EUS standpoint), or you'd see more institutions finally dropping IE6. -Chad

On Fri, Feb 20, 2009 at 2:02 PM, Thomas Dalton &lt;thomas.dalton(a)gmail.com&gt;wrote;wrote: I know a lot of shops are skipping the Vista upgrade with 7 on the way. They might be willing to hang on to a slightly-outdated product that just works until they get the new OS rather than pushing an IE update so late in their OS cycle. That being said: I think the number of IE6 users will drop yet again when Windows 7 comes out. But this is all conjecture. TBH: The 2010 date might not be a bad one to target for IE legacy support. When the company doesn't support it, we shouldn't bother either. At that point, people who are using IE6 are just too far behind the curve to be bothered with. If we kept this "must support it" mentality forever, we'd have never made it past <table> and <font>. All browsers die eventually, it's our job to kill support around the same time. Not while there's still a good percentage using it (and 20% is a fair chunk), but rather when it's practically un-used anyway. -Chad

On Fri, Feb 20, 2009 at 2:27 PM, Chad &lt;innocentkiller(a)gmail.com&gt; wrote: This is precisely it. We will cease to support IE6 when it's no longer used by a significant percentage of our users. Not when we think its users should upgrade, not when Microsoft thinks its users should upgrade, not when some militant anti-Microsoft brigade thinks its users should upgrade -- but rather when users *have* *already* upgraded. On Fri, Feb 20, 2009 at 4:37 PM, Tei &lt;oscar.vives(a)gmail.com&gt; wrote: That's a ridiculously gross exaggeration. In the particular case of MediaWiki, the real work in making us IE6-compatible was done long ago. The ongoing cost of IE6 support is fairly close to zero. Only the occasional feature is complicated by its lack of selector support or whatnot. It's slightly good for most users (very slightly in Wikipedia's case), and *very* bad for the large minority that still uses IE6. We are just not going to ignore a fifth or more of our users, period. To do so would be completely irresponsible. There would most definitely not be a net benefit to users, only to developers.

On Fri, Feb 20, 2009 at 9:19 PM, Thomas Dalton &lt;thomas.dalton(a)gmail.com&gt; wrote: If some new major feature is added that a given old browser doesn't support, a banner message could be displayed akin to the one Brion put in place recently for mobile users not using the mobile gateway. -- Stephen Bain stephen.bain(a)gmail.com

On Fri, 20 Feb 2009 11:16:06 +0100, Aryeh Gregor &lt;Simetrical+wikilist(a)gmail.com&gt; wrote: The Norwegian sites in question still "work" in IE6, they just urge users to upgrade and have stopped "wasting time" implementing hacks to make new features still work in IE6. The argument is that the people who still use IE6 now are not likely to ever change on their own because "it still works". It is fairly trivial to make most Intranet applications work in at least IE7, most IT departments are just plain lazy when it comes to implementing new technology as long as the old stuff "works" (and I can understand that to some extent, if the old stuff does what it's supposed to why waste time and money to upgrade just for the sake of upgrading, but at some point you have to bite the bullet or be left in the stoneage). So the point of the campaign is to show people that IE6 doesn't "just work" anymore, thus creating an actual reason to upgrade. If enough sites follow that model I'm pretty sure you'll see the market share of IE6 start droppign faily fast as people call computer savy friends to help them finaly upgrade or IT departments are presured into upgrading to IE7 or allow FireFox, Opera or other alternate browsers to be installed (for example I do have to use IE6 for some ancient Intranet stuff at work, but I use Opera for everyting else). I do believe that once IE6 (not to mention IE5) is dead and burried the web will be a better place, so IMHO we should not ask "what do we gain from this", but instead "do it for the betterment of humanity" (pompous enough for you ;P). I'm not saying we should stop making the site work for IE6 users, but I don't think a browser spesific "site notice" urging them to upgrade would be such a bad thing. -- [[:en:User:Sherool]]

On Fri, Feb 20, 2009 at 3:50 PM, Aryeh Gregor &lt;Simetrical+wikilist(a)gmail.com&gt; wrote: Exactly. Also, in many ways upgrading to IE7 doesn't solve many problems at all (except for some very obvious CSS bugs) because it's still riddled with bugs and wrong implementations of W3C specs, especially on the Javascript implementation. We just have to live with the fact that web development means spending 30% of your time on writing hacks for IE6 and IE7 (and probably for IE8 too). -- Hay