Actually, FF does not support SNI on Windows XP. The OP is correct, at least, last time I checked this out (I made an edit to the SNI page on Wikipedia years ago regarding this).
FF however does NOT use the built-in Windows Certificate store, which makes it very painful to use in environments where trust relationships need to exist, but built automatically.
I'd be surprised if FF4 didn't support SNI under XP, as FF3.6.15 seems to on my home machine (tested by accessing https://sni.velox.ch/). I don't know if earlier versions supported SNI on XP or not - I might fire up the portable versions (available from http://portableapps.com/apps/internet/firefox_portable under the heading "legacy versions") and try the test with them later.
The current stable version of Chrome seems to support the feature under XP too (though earlier versions didn't, see http://www.google.com/support/forum/p/Chrome/thread?tid=0403... - and other relevant discussions & bug reports come up in a quick search too).
IE8 on XP fails as expected (certificate error due to the name not being presented to the test site which means the wrong certificate is handed out).
All three pass the test fine on my Vista machine here.
Also will probably bleed into future platforms - people stick with what they're comfortable with, so once they switch to FF/Chrome/Safari they'll likely seek out that browser on whatever new machine they get after their hardware dies.
It's at a point where I don't see anything wrong with going either way.
I don't think IE9's abandonment of XP is going to slow the adoption of new web technology all that much, something that Mozilla seems to suggest. There is a lot of momentum behind them with mobile devices supporting them so well. Still, 40% of web users, even if those are among the lowest value web users, is worth developing around.
Mozilla isn't around to make a profit, so spending on XP users makes more sense. Abandoning them would have also made sense.
Mozilla wants its cake and to eat it too. They'll whine that IE9 doesn't support XP, and then whine that IE9 isn't modern. If it's not modern then it really doesn't matter if it supports XP or not.
I personally can't stand XP as an OS. But I do understand Mozilla from a business perspective. Although I wish they weren't so sanctimonious about IE9 not supporting it since they really don't care and I think it just continues to make Mozilla look like whiners rather than doers, e.g., Google/Chrome.
Mozilla is trying to position FireFox a as superior alternative to IE9. Therefore, there's no cognitive dissonance at all with Mozilla saying that 1) IE 9 is worse than Firefox because it doesn't run on XP and 2) IE 9 isn't really modern while Firefox is.
Mozilla doesn't want IE 9 to be modern, but more than that, it doesn't want users to think that IE 9 is modern. Mozilla doesn't want IE 9 to have any compelling features whatsoever.
A negative marketing campaign will always backfire to some portion of your audience.
Mozilla do want IE9 to be "modern", that's entirely in line with their mission as a non-profit that promotes the open web. It's ironically Microsoft that's scared most of a truly modern IE9.
What Mozilla certainly don't want is marketing BS confusing users into thinking IE9 is modern, while it holds the web back and drags its feet about updates. Nor does it want any browser controlled by Microsoft to have dominant market share while Microsoft still sees the web as a threat that it can fight mostly by go-slow tactics like having IE6-8 hanging around preventing new web technologies hitting critical mass.
They don't even have spellchecking in their browser because the Word guys feel threatened.
They don't even have spellchecking in their browser because the Word guys feel threatened.
I suspect it may have more to do with the fact that some features are easy to do as addins. There are several spell checkers available for IE. Given that the IE team seems resource constrained, it seems more useful to work on things like perf, compliance, security, than features that are easy to add by 3rd parties. I'm not saying it wouldn't be nice to have, but given that IE9 still has HTML/CSS gaps, I'd much rather they work on those.
Why would the IE team be more resource constrained than the Firefox team, who managed to add a spellchecker years ago? I'll admit that I've not gone looking, but I've not ever heard of anyone using one of these third party IE spellcheckers - it's just been a feature which I expect a decent browser to have that IE clearly lacked. Surely that perception hurts them more than the cost of a few developer-months to take one of their existing spellcheckers and drop it into IE?
Why would the IE team be more resource constrained than the Firefox team
Because they have fewer people working on the product? I don't know. But presumably they have some constraints on resources that prevent them from doing everything -- hence having to prioritize.
Surely that perception hurts them more than the cost of a few developer-months to take one of their existing spellcheckers and drop it into IE?
So then what do you cut? Dev/test time doesn't just appear magically. Do you cut flexbox? Text shadow? History?
The IE team is resource constrained because it's not in Microsoft's interests to build a good browser. If it was in their business interests, like XBox, then they could throw tens of billions of dollars at it. But it's not, so no spellchecker and no WebGL and no IE9 for XP.
You do realize that the XBox team is also resource contrained. For example, you can't run any XBox or PC game on an XBox 360. Had to cut something. There's not even a web browser in the XBox 360.
Everything is about tradeoffs and constraints. You may not be familiar enough with the industry to know that they're there (ppl think that MS and Apple have enough money to simply do everything, but they don't), but trust me, these products have serious resource constraints across the board.
Strangely, you seem to think Xbox trying to conquer the living room without bothering to include a browser supports your case that Microsoft loves the web, but just consistently runs out of money across its various products before getting around to supporting it properly, whereas I would have had that right up there with the lack of spellchecking in IE as a clear example of Microsoft quietly hoping the web will go away if they ignore it long enough.
I'm not asking or expecting them to support everything and anything with their minuscule bank account and negligible market power. I'm just pointing out that the web is kryptonite to Microsoft and their PR to the contrary is laughable. If you want to believe it's simply mismanagement on a massive scale that leads them to fail at the web year after year then good for you, but I personally think it's just smart self-interest on their part.
Again, the position seems to not make sense. And it seems desperate. Is Chrome worse than IE mobile, because its not on WP7? Is Safari worse than Chrome because its not on Chome OS?
Again, I think its a fine business decision for Mozilla. But it seems odd to decry IE9's absence on XP when you already think it's not modern.
And the positioning against IE9 seems kind of dumb since IE9 isn't on XP. It's not going to beat IE9 on Win7 or Vista. But it can make huge inroads on XP. But not by taking IE9 share, but rather by taking IE6/Chrome share.
It just seems like Mozilla is so caught up at being anti-MS that that they still haven't understood that Google has already picked their wallet and has just started emptying their bank account.
Hardly surprising, but it's an interesting situation. The only company in the world that wants and needs Windows XP to disappear is Microsoft; for everyone else XP is an enormously bountiful source of users and revenue. It would be suicidal for Mozilla or Google to say "we won't support XP", just as it's fatal for Microsoft to linger on an 11-year old platform that eats into their revenue stream. Interesting dilemma.
ReactOS is just WINE on top of a custom Windows-like kernel. All of the real action in reimplementing Windows occurs in WINE, afaik.
It's pretty far from "full XP compatibility", but a lot of stuff works just fine. I'd love to see WINE considered a big investment center for people like Canonical. I think WINE and open graphics drivers are the two biggest parts holding back desktop Linux. Even if WINE could get .NET compatibility, that'd be a huge step forward (right now it has trouble running any of MS's .NET frameworks correctly). The reality is it's going to take a lot of money and especially time to cause the lock-in of Win-only apps to fully regress; it'd be a lot faster to work with it and cause almost everything to work perfectly on WINE. Office and .NET are the biggest hold-ups atm.
The problem with major investment in WINE is that as soon as MS thinks that suing CodeWeavers/any other WINE investors or distributors would be less harmful than allowing the project to continue without the extra attention, they're going to drop some big patent litigation all over that. MS is not just going to sit back while WINE plays a fundamental role in killing Windows -- as soon as it becomes serious, look for a major, relentless attack by MS.
Yes, the ability to use native drivers on ReactOS stems from writing a kernel that mimics the Windows kernel API/ABI as closely as possible. The ReactOS userland, however, is provided primarily by WINE. You can think of ReactOS primarily as an open kernel compatible with Windows's kernel API, and WINE as an open userland compatible with Windows's userland APIs. ReactOS isn't rewriting WINE, they use WINE extensively. The main unique work of ReactOS is the Windows-compatible kernel.
Mono is helpful for pure .NET directly from a Unix environment. However, Mono is more of an open .NET development environment than a .NET compatibility environment; while Mono can run a lot of .NET apps that are made for Windows, that's almost incidental of compatibility with the CLR bytecode. There are important implementation differences that Mono is aware of and has no intention to correct because, unlike WINE, Mono is not really a compatibility layer.
Additionally, many Windows programs launch .NET programs directly, and afaik it's not very easy to get them to use Mono by default. For instance, I recently purchased C&C 4 on Steam, but it's a pain to make it work because circumventing the .NET-based launcher wants me to include an EA ID (I could use a fake auth server but just don't feel like putting that much effort in), and launching directly doesn't work because of the .NET-based launcher (.NET doesn't work). Further, more and more products are relying on .NET components, including Visual Studio, Office, and other big things. I think getting the .NET runtimes working well should be one of the main goals for WINE devs.
It's hard to make money selling stuff to cheap people, and it's hard to gain users by chasing people who are ok still using XP, for whatever reason.
It's nice that Firefox is still supporting XP, but I don't think the decision should be based on the % of web users still on XP. It should be based on the % of them likely to be interested in downloading a new browser.
So what's all this nonsense about XP being 10 years old and so customers who bought it not being worthy of a newer browser? Do you know when Microsoft actually stopped selling XP to customers?
Trick question, they haven't. Despite several passed deadlines the latest plan, as of July 2010, is to continue selling downgrade licences to businesses up until two years after Windows 8 ships!
Right. Very large companies are still tied to XP, and therefore Microsoft is. IBM's standard load is still XP, and I haven't seen any Vista or 7 plans yet. I do suspect that they'll skip right to 7 when it's considered "mature" and they allocate the resources to clean up the enormous technical debt accumulated with the in-house applications.
You can't always make decisions based on dollars and cents, sometimes its about just being good to your users and that is one thing that Microsoft has proved incapable of over the years.
For them its always business ... which is why I hope they eventually lose the browser battle.. They act like people on Windows XP didn't also pay money for their copies of windows ... but all this will do is give people a solid excuse to use other browsers on Windows XP as IE 8 starts to age.
Thanks for letting Chrome and Firefox an unencumbered run at your market share Guys!
The problem is that Microsoft has too many stakeholders. The IE9 team would certainly love to have more people in the world use their product, but company-wide strategy leads to them dropping WinXP support.
This is slowly going away (see the various MSFT iPhone applications), but this mentality will certainly remain with the company. Actually, I'd argue this is a strength under certain circumstances.
If it is so much work for Microsoft to support IE9 on XP, they should make it and offer it for sale. Since XP is so important, and IE9 is so important, people will surely pay for a browser upgrade, right?
What's the market price for a modern browser? $79? $29?
According to a friend that just got a new work laptop, Windows XP is still standard issue at IBM. I imagine plenty of other large companies are in the same position.
Forcing an OS upgrade is in ms's interest since its likely people will likely stick with the built in one, which is Microsofts browser.
Keeping backwards compatibility is in mozillas interest since people are likely to change their browser than their OS. This means xp users are likely to switch to mozilla than Microsoft.
37 comments
[ 4.9 ms ] story [ 77.6 ms ] threadedit: https://secure.wikimedia.org/wikipedia/en/wiki/Server_Name_I... seems to indicate that Firefox supports it on all platforms.
FF however does NOT use the built-in Windows Certificate store, which makes it very painful to use in environments where trust relationships need to exist, but built automatically.
The current stable version of Chrome seems to support the feature under XP too (though earlier versions didn't, see http://www.google.com/support/forum/p/Chrome/thread?tid=0403... - and other relevant discussions & bug reports come up in a quick search too).
IE8 on XP fails as expected (certificate error due to the name not being presented to the test site which means the wrong certificate is handed out).
All three pass the test fine on my Vista machine here.
i'm sure it wasn't easy to dedicate all the extra work up front and over the life of the project...
I don't think IE9's abandonment of XP is going to slow the adoption of new web technology all that much, something that Mozilla seems to suggest. There is a lot of momentum behind them with mobile devices supporting them so well. Still, 40% of web users, even if those are among the lowest value web users, is worth developing around.
Mozilla isn't around to make a profit, so spending on XP users makes more sense. Abandoning them would have also made sense.
I personally can't stand XP as an OS. But I do understand Mozilla from a business perspective. Although I wish they weren't so sanctimonious about IE9 not supporting it since they really don't care and I think it just continues to make Mozilla look like whiners rather than doers, e.g., Google/Chrome.
Mozilla doesn't want IE 9 to be modern, but more than that, it doesn't want users to think that IE 9 is modern. Mozilla doesn't want IE 9 to have any compelling features whatsoever.
A negative marketing campaign will always backfire to some portion of your audience.
What Mozilla certainly don't want is marketing BS confusing users into thinking IE9 is modern, while it holds the web back and drags its feet about updates. Nor does it want any browser controlled by Microsoft to have dominant market share while Microsoft still sees the web as a threat that it can fight mostly by go-slow tactics like having IE6-8 hanging around preventing new web technologies hitting critical mass.
They don't even have spellchecking in their browser because the Word guys feel threatened.
I suspect it may have more to do with the fact that some features are easy to do as addins. There are several spell checkers available for IE. Given that the IE team seems resource constrained, it seems more useful to work on things like perf, compliance, security, than features that are easy to add by 3rd parties. I'm not saying it wouldn't be nice to have, but given that IE9 still has HTML/CSS gaps, I'd much rather they work on those.
Because they have fewer people working on the product? I don't know. But presumably they have some constraints on resources that prevent them from doing everything -- hence having to prioritize.
Surely that perception hurts them more than the cost of a few developer-months to take one of their existing spellcheckers and drop it into IE?
So then what do you cut? Dev/test time doesn't just appear magically. Do you cut flexbox? Text shadow? History?
Everything is about tradeoffs and constraints. You may not be familiar enough with the industry to know that they're there (ppl think that MS and Apple have enough money to simply do everything, but they don't), but trust me, these products have serious resource constraints across the board.
I'm not asking or expecting them to support everything and anything with their minuscule bank account and negligible market power. I'm just pointing out that the web is kryptonite to Microsoft and their PR to the contrary is laughable. If you want to believe it's simply mismanagement on a massive scale that leads them to fail at the web year after year then good for you, but I personally think it's just smart self-interest on their part.
Again, I think its a fine business decision for Mozilla. But it seems odd to decry IE9's absence on XP when you already think it's not modern.
And the positioning against IE9 seems kind of dumb since IE9 isn't on XP. It's not going to beat IE9 on Win7 or Vista. But it can make huge inroads on XP. But not by taking IE9 share, but rather by taking IE6/Chrome share.
It just seems like Mozilla is so caught up at being anti-MS that that they still haven't understood that Google has already picked their wallet and has just started emptying their bank account.
Firefox 3.x works perfectly fine on WinXP too.
It's pretty far from "full XP compatibility", but a lot of stuff works just fine. I'd love to see WINE considered a big investment center for people like Canonical. I think WINE and open graphics drivers are the two biggest parts holding back desktop Linux. Even if WINE could get .NET compatibility, that'd be a huge step forward (right now it has trouble running any of MS's .NET frameworks correctly). The reality is it's going to take a lot of money and especially time to cause the lock-in of Win-only apps to fully regress; it'd be a lot faster to work with it and cause almost everything to work perfectly on WINE. Office and .NET are the biggest hold-ups atm.
The problem with major investment in WINE is that as soon as MS thinks that suing CodeWeavers/any other WINE investors or distributors would be less harmful than allowing the project to continue without the extra attention, they're going to drop some big patent litigation all over that. MS is not just going to sit back while WINE plays a fundamental role in killing Windows -- as soon as it becomes serious, look for a major, relentless attack by MS.
But yes, it will never be Windows and will always have technical and legal problems.
Mono is helpful for pure .NET directly from a Unix environment. However, Mono is more of an open .NET development environment than a .NET compatibility environment; while Mono can run a lot of .NET apps that are made for Windows, that's almost incidental of compatibility with the CLR bytecode. There are important implementation differences that Mono is aware of and has no intention to correct because, unlike WINE, Mono is not really a compatibility layer.
Additionally, many Windows programs launch .NET programs directly, and afaik it's not very easy to get them to use Mono by default. For instance, I recently purchased C&C 4 on Steam, but it's a pain to make it work because circumventing the .NET-based launcher wants me to include an EA ID (I could use a fake auth server but just don't feel like putting that much effort in), and launching directly doesn't work because of the .NET-based launcher (.NET doesn't work). Further, more and more products are relying on .NET components, including Visual Studio, Office, and other big things. I think getting the .NET runtimes working well should be one of the main goals for WINE devs.
All three will support hardware acceleration on XP.
Only one that left tens of thousands (maybe millions?) of XP users out in the cold is Microsoft.
It's nice that Firefox is still supporting XP, but I don't think the decision should be based on the % of web users still on XP. It should be based on the % of them likely to be interested in downloading a new browser.
Trick question, they haven't. Despite several passed deadlines the latest plan, as of July 2010, is to continue selling downgrade licences to businesses up until two years after Windows 8 ships!
For them its always business ... which is why I hope they eventually lose the browser battle.. They act like people on Windows XP didn't also pay money for their copies of windows ... but all this will do is give people a solid excuse to use other browsers on Windows XP as IE 8 starts to age.
Thanks for letting Chrome and Firefox an unencumbered run at your market share Guys!
This is slowly going away (see the various MSFT iPhone applications), but this mentality will certainly remain with the company. Actually, I'd argue this is a strength under certain circumstances.
Windows XP users would never pay for IE 9, and Apple claim they can't give XCode 4 away to users from an OS released 18 months ago.
What's the market price for a modern browser? $79? $29?
$0?! How on earth did that happen?
Keeping software modern is a good thing.
Keeping backwards compatibility is in mozillas interest since people are likely to change their browser than their OS. This means xp users are likely to switch to mozilla than Microsoft.