Does Microsoft care about Webmasters?

synja wrote:

You misunderstand what I am saying. I think my post was a little too tongue-in-cheek, sorry.

Just because the W3 decides to post a standard to their web site does not make it an absolute vital part of the web, nor does it automatically become a top priority for MS to implement.

SVG is a great example. Its got a great "gee whiz" geek appeal for being entirely XML based, but its overall usefulness is pretty much nullified once you actually try to use it for something other than creating charts and graphs... you know, things that your average internet user isnt even intelligent enough to interpret in the first place.. trust me, ask anyone who has actually developed web apps that utilize SVG (myself included) its really quite useless for consumer based web sites.

So is it really worth getting pissed at MS about for not implementing right away, or at all for that matter? In my opinion, no.

• Microsoft is a member of the W3
• Chances are Microsoft made the specification or a good part of it.
• Many web developers, of which a good portion are Microsoft customers, will want to use it.
• Corporations that are huge customers of Microsoft might be able to save money (could save bandwidth or compress very well, which would mean the server will close connections faster lowering the CPU load thus allowing the server to do more with less) by utilizing the specification
• Microsoft has control of 95% of the market so it has an obligation to do so or risk having another antitrust lawsuit.

Flash is useful. One of the reasons is because its graphics can be rescalable. That is why SVG is useful.

When you consider the benefits of a new specification (if it wasn't useful, it wouldn't have been made) people everywhere have every right to be upset with Microsoft for not adding support ASAP (ASAP != years).

Edit: By the way, this was a general post meant for standards in general. Specifications such as CSS and PNG need full support ASAP while others such as SVG do not require support immediately but should have some priority that will make sure that they get support long before people are ready to start pulling their hair out. I would edit my post to clarify that but I've been coding a CMS for my website for a good portion of the day and I'm ready to call it a day.

TonyCh wrote:

Hi folks, I know I'm a little late to the party, but here I am. Before I begin, I should introduce myself. I'm Tony Chor, the Group Program Manager for the IE team.

With respect to this discussion, it might help to provide a little context. After IE 6 shipped in the fall of 2001, parts of the IE team went off to focus on different web browsing challenges. For instance, the UI team went off and built MSN Explorer. While Microsoft's overall investment in the web and web browsing probably actually went up during this time, it's probably fair to say that we defocused on Internet Explorer proper. It's easy to Monday morning quarterback this set of decisions, but whatever, here we are.

Anyway, we've started the get the band back together as it were. So far, the new IE team really has been focused on security and taking care of key corporate customer issues. A big part of the security effort has been our push around XPSP2. This has been all-consuming to the point where we've essentially stopped our Longhorn work for now.

That said, we are starting to plan our future again (and there will be a future for IE outside of security releases.) We've heard the standards discussion loud and clear. I can't commit yet to what our position on specific standards will be, but here's a set of principles that I think we might use to decide.

1. Fix bugs in our current standards support that block developers from getting things done.

2. Adopt approved standards that enable or unblock interesting new scenarios and uses.

3. Standards need to be well-formed and have clear IP / license as well as technically possible / affordable.

4. Our implementation of the standards should either be backward compatible (or be unrelated to) with sites and applications developed for our existing browsers. Or, they need to provide such overwhelming benefit as to merit a break with old apps/sites.

So, that's my take. I'd appreciate your thoughts on these principles.

Thanks.

• CSS Level 1 - support is good but there are bugs
• CSS Level 2 - support is terrible and MSDN claims IE does not support it
• CSS Level 3 - support is non-existant
• PNG - There is a petition with thousands of signatures requesting proper PNG Alpha Transparency support (currently requires a propeitory filter which, devalues PNG, in some cases makes using transparent PNGs impossible and prevents webmasters that code for standards from using Alpha Trasparency)
• XHTML Specifications - I stick with XHTML 1.0 Tranisition so the only issues I really know of are with the doctypes but if my understanding is correct, there are issues with 1.1 and 2.0
• XML
• XSLT

Here is a list that need full support after that list:

• P3P
• SVG

After that, I really suggest making sure that you have full standards support by speaking to people in the standards community (e.g. webmasters of standards websites, w3, etc), I'm sure they can point out where Internet Explorer is still lacking.

By the way, as soon as the IE team makes a decision, please let everyone at Channel 9 forums know about it, many webmasters here are ready to pull their hair out over IE's standards support so any news would be more than welcome.

Edit: If you implement support for CSS attributes that Microsoft has thought up and are not currently part of the specification, do not name it what the W3 would name it when they add proper support assuming that they will. Use a prefix like Mozilla did.

For example, lets assume that it is 1997 and you thought up of the margin attribute. When implementing support, do not have IE's CSS parser parse for "margin," have it parse for "-ie-margin" (and tell people that this is temporary until it is in a final specification). That way when it is in a specification that has been finalized, you can implement proper support for the "margin" attribute.

Knute wrote:

Shining Arcanine wrote: Regarding 4, you should forget about backwards compatibility when it comes to implementing standards (the way they should have been orignally implemented) as that is what quirks mode (please tell me IE has a quirks mode) is for. When in standards mode (triggered by a proper doctype), everything should be 100% compliant with the W3 specification. There should be no concern with backwards compatibility because the doctype states that the webpage abides by the specification (how many webpages are coded with nonstandard workarounds for IE's bugs and have a proper doctype?). If the pages are coded with nonstandard workarounds and have a proper doctype, that is not IE's fault, it is the webmaster's fault for using a doctype with invalid markup in the first place.

You would be the first one to rant when pages broke, give me a million breaks. You know there is no way in the world they would not do backwards compatibility.

I think another main reason you don't see a huge effort for compliance is that the user's aren't the ones complaining about standards. All the end user knows is that their internet experience currently works whether pages are in full compliance with standards or not.

I personally don't think you will ever see a full 100% compliance with the standards but perhaps a direction that is aligned as close as possible.

Also if you are unhappy with browsers then perhaps you should just code up one on your own... Then perhaps you would see what type of herculean effort is required to dot every i and cross every t when it comes to standards...

~ Knute

travis wrote:

Knute wrote: ...I think another main reason you don't see a huge effort for compliance is that the user's aren't the ones complaining about standards. All the end user knows is that their internet experience currently works whether pages are in full compliance with standards or not....

So much for Balmer's "Developers! Developers! Developers!" then.

Knute wrote: ...Also if you are unhappy with browsers then perhaps you should just code up one on your own... Then perhaps you would see what type of herculean effort is required to dot every i and cross every t when it comes to standards...

First of all I should probably state that my unhappiness with IE has driven me to use FireFox as my default browser.  I have written an extension for it, and it was very easy.  I think that if you saw the herculean effort that has been required to code around IE's bugs you'd realize how much development time this would save.

TonyCh wrote:

Glad to see my post got read...

I definitely see your perspective re: doctype and standards. It's not a bad way to think about the issue of backwards compatibility, but we should be very clear that regardless of whose fault it is when a site breaks in a new version of IE, Microsoft is held responsible publicly and sometimes legally. The fact IE behaves a certain way today creates a set of expectations that is difficult to change. We're doing for security, but standards compliance may be harder. We'll see.

The list you have is pretty good. Per my guidelines above (which I'd still like feedback on), stuff like CSS3 would not be something we'd ship yet since it's not final.

Not much defense on the PNG transparency issue other than GDI and our architecture made it hard originally. We want to fix it very much as well. No ETA yet, but I'll try to let this forum know when we make some decisions on this.

Finally, to set expectations, no major standards chagnes will happen in months (vs. years.) Any big changes in Trident (our rendering engine) will take a long QA and beta cycle. Plus, our dance card for the next few months is full.

Any other thoughts?

• You could attempt to limit the amount of people complaining by documenting the changes to Internet Explorer (when the next version of IE is released after XP SP2 as either a service back or an operating system) but I'm not sure if that would prevent people from attempting to hold Microsoft responsible for their mistakes when designing their websites.
• Another thing you could do (probably won't be very effective but is a pleasant thought) is to point people to the w3 validator saying "there is it, it states where the mistakes are, use it to find out how to fix them."
• Yet another thought (this one actually might work but it goes back to the doctype) is to have two engines, one will be the existing engine and you can call it the quirks engine. The other will be a the same engine except it will have proper standards support and be what is primarily developed. The quirks engine will be utilized when no doctype is detected and the standards will be utilized when there is a doctype. That could reduce validation time however, it would result in quite a bit unecessary code so you might want to pull the quirks mode code out of the standards engine and the standards mode code out of the quirks engine. If you do that, you could discourage use of the quirks engine on MSDN.

on the topic of broad support, I understand that adding broad support in a few months is asking for too much, however I (do not know the amount of resources allocated to the IE team and) was trying to stress that it needs to be done ASAP.

On 2nd thought concerning CSS3, I agree with you, wait until it is done so that it can be implemented the right way.

warren wrote:

TonyCh wrote: Finally, to set expectations, no major standards chagnes will happen in months (vs. years.) Any big changes in Trident (our rendering engine) will take a long QA and beta cycle. Plus, our dance card for the next few months is full. Any other thoughts?

Perhaps offering "Technology Preview" releases for future IE versions could be a way to get the developer community onside and help make the QA process as productive as it can possibly be.   There wouldn't be an expectation from developers that the product would be "feature-complete" at that time in terms of UI, administration, etc., but at the same time, it would let them provide you with feedback on specific rendering issues that you need to hear about.

(I do IE web application development professionally, so you could say I have a vested interest in seeing the platform improve )


Changing subjects a bit; are you able to comment on what new features of  IE6 that are being included with XP SP2, will be made available to Windows 2000 users?

DouglasH wrote:

shineing,

I have one question on this.  Considering (looking at hte css website) that none of the css3 specs are recommendations. Isn't a little early to be critical of the state of IE in that regards??

Now css 1 and 2 I can see and agree with (not that css2 spec the way it is currently written isn't broken and the reason that css 2.1 and the css3 specs being in spec.

Now if we get to recommendations and it is implementable within 6 months to a year than I can agree with your statements. 

Considering that it typically takes about 4 to 5 years for these standards to finally become implemented in any large numbers on the web there is time.

(what is it about 50 to 60% of the web is still useing Html version 3 last I checked.

douglas

Knute wrote:

DouglasH wrote: shineing, I have one question on this.  Considering (looking at hte css website) that none of the css3 specs are recommendations. Isn't a little early to be critical of the state of IE in that regards?? Now css 1 and 2 I can see and agree with (not that css2 spec the way it is currently written isn't broken and the reason that css 2.1 and the css3 specs being in spec. Now if we get to recommendations and it is implementable within 6 months to a year than I can agree with your statements.  Considering that it typically takes about 4 to 5 years for these standards to finally become implemented in any large numbers on the web there is time. (what is it about 50 to 60% of the web is still useing Html version 3 last I checked. douglas

Douglas,
Good point, right on the money. Also interesting is that if you go to here and type in slashdot.org - cnn.com - apple.com - None of them are in full compliance with the specs either.

Standards are important in all that we do, but if there are better ways of doing things so suit our needs we take that route.

~ Knute