What we talk about on IEBlog


The information published in this post is now out-of-date.

—IEBlog Editor, 20 August 2012

I will paraphrase Jeff Davis, a developer on the IE team, when it comes to setting parameters for what we will and will not talk about IEBlog. 

We will happily post and discuss issues around what features should be in IE, how features work in IE, the importance of application compatibility in IE, tips and tricks in IE, answers to technical issues, security and extensibility in IE, web browsing in general, and what keeps us up at night.

For better or worse, there are some things we are just not at liberty to talk about. Unfortunately, due to our working with law enforcement around malicious hacking and attacks, we cannot comment on security issues currently under investigation. Another example: as much as we’d like to be “transparent” about what to expect when, we’re a publicly traded company and we have to abide by larger communication and announcement guidelines. After a feature is public, we will go into depth on it (e.g. see Jeff and Tony’s blogs about XP SP2 features they worked on).  

I also strongly recommend that general feedback be posted on the Internet Explorer Wiki hosted on Channel 9

Now, back to my day job.

– Dean

P.S. to to everyone providing constructive feedback – thanks, we do appreciate it.

Comments (78)

  1. Anonymous says:

    I just thought I’d draw a little attention to Dave Hyatt’s blog Surfin’ Safari (http://weblogs.mozillazine.org/hyatt/) which pretty much represents the ideal that all commercial browser-maker blogs should aspire to. I realise the IE team has more constraints than say Mozilla or Konqueror with regard to necessary secrecy, but as Safari is also commercial it seems a reasonable comparison.

    I look forward to seeing some content that isn’t just marketing from you guys soon (I know you’re just settling in).

    OT: Any chance of a preview function for comments?

  2. Anonymous says:

    Thanks, I’ll check it out. I didn’t know about that blog.

    I’m looking forward to the more product-oriented posts myself. Thanks for understanding that we’re working through the housekeeping issues here right now.

    I’ll ask around wrt the preview function for comments.

  3. Anonymous says:

    Dean, that sounds perfectly reasonable to me. Ditto to what Rory said; Dave Hyatt’s doing a good job with his blog. It would be interesting to see you two comment on each others developments.

    I hope you appreciate, however, that it is a little hard to talk about new features when the existing features don’t work properly.

    Dean (or anybody else) – does anybody have more details on the Content-Type HTTP header handling in the new XP service pack? The knowledge base article is a bit vague. Does Internet Explorer always honour the Content-Type header, no matter what the extension, no sniffing?

    Will this behaviour be back-ported to older operating systems?

  4. Anonymous says:

    Actual improved support for standards in SP2. Cool. Thats something I’d also love to here more about.

  5. Anonymous says:

    This is the description of Internet Explorer’s current behaviour:

    http://msdn.microsoft.com/workshop/networking/moniker/overview/appendix_a.asp

    This is the specification it violates:

    http://www.ietf.org/rfc/rfc2616.txt

    This mentions new behaviour:

    http://www.microsoft.com/uk/windowsxp/sp2/sp2.mspx

  6. Anonymous says:

    Hi Dean, just thought I’d check back here incase you missed my response to your comment on my own blog. Basically I think my attitude towards this blog (reserved optimism) is far better represented here than there. I sometimes lapse into Slashbot anti-MS mode but I genuinely believe that IE developers are no different at heart than Apple or Mozilla developers. I want IE to succeed by being good, not by being good enough. I just hope you can pull it off. Like Jim said, I’m interested in fixing old bugs more than making new features.

  7. Anonymous says:

    > I genuinely believe that IE developers are no different at heart than Apple or Mozilla developers. I want IE to succeed by being good, not by being good enough. I just hope you can pull it off.

    I agree 100%. Significant improvements to Internet Explorer would be fantastic, and I wish you all the best luck with it. I know I talk about Internet Explorer’s shortcomings a lot, but I don’t intend this to be an attack on its developers, IE’s stagnation is clearly a policy from higher-up and not developer laziness or anything like that. If nobody had been working on Mozilla for the past couple of years it would be practically unusable.

  8. Anonymous says:

    Hey Rory, thanks for the response. I hear you on good v good enough, and getting basics right before adding more stuff.

    There are some challenges around what you describe as "fixing old bugs"… basically how to deal with the fallout to everything and everyone who relied on the old behavior (however good or bad that behavior was). I talked about it to Scoble the other day. I’ll post about it as soon as I can so I can hear specific feedback about how we’re thinking about it now.

  9. Anonymous says:

    Well, one good thing (in this backwards compatability argument) about Windows/IE integration is that you can’t install IE 5 onto XP, or IE 4 onto 2k. People can’t just go to oldversion.com & click Internet Explorer.

  10. Anonymous says:

    JP- Actually it is possible, though not quite at a one click level. See…

    http://www.quirksmode.org/browsers/multipleie.html

  11. Anonymous says:

    And again, it would have been nice if that info came from MS rather than someones random blog.

  12. Anonymous says:

    "how to deal with the fallout to everything and everyone who relied on the old behavior"

    I can only imagine how hard that must be. IE is so ubiquitous that every aspect of it, including every feature and every bug, is probably in use by someone somewhere. And probably in enterprise apps to boot.

  13. Anonymous says:

    Dean, your browser is already great, it just needs some few improvements. Tabbed browsing, popup blocker etc… for users, some few additions for developers. It is already a fantastic browser, but that’s not enough. Firefox almost caught up with IE now, and in some areas it is even better. Its overall interface is great, but you can do a lot more for usability and simplicity. I think you just need to repeat your success with windows again. Don’t break the old applications, yet build the best browser again.

  14. Anonymous says:

    "how to deal with the fallout to everything and everyone who relied on the old behavior"

    Does anybody have any comments on my earlier suggestion of an X-Internet-Explorer-Standards: true HTTP header? It seems to me that it’s perfectly backwards-compatible.

    http://blogs.msdn.com/ie/archive/2004/07/21/190747.aspx#191539

  15. Anonymous says:

    Jim claimed that IE doesn’t support Javascript standards. He uses the fact that IE’s term for Javascript is JScript but not JavaScript, to claim that IE doesn’t support standards.

    Don’t change the HTTP header at all. Probably it will break so many applications, as Firefox breaks them because of that and it hurts Firefox a lot. Probably Jim wants you to change that just to hurt Microsoft, so that people get pissed off at Microsoft.

    Many servers do not properly configure the correct http headers, thus IE’s ability to determine the correct header helps a lot. Of course, if it doesn’t hurt in anyway, some features can be added. Just be aware of people giving you wrong feedback intentionally.

  16. Anonymous says:

    Alex I’m confused by your ascribing of malice to Jim’s comments. I don’t see any comments relating to javascript and what I have seen from Jim is the opinion that IE has deficiencies–particularly from a web-developer’s perspective–but that the IE team will hopefully address at least some of our issues. He has made useful inquiries and an honest attempt at suggesting solutions, taking into account the difficult situation the team is in of having to keep as much compatibility as possible.

  17. Anonymous says:

    "Alex I’m confused by your ascribing of malice to Jim’s comments. I don’t see any comments relating to javascript and what I have seen from Jim is the opinion that IE has deficiencies–particularly from a web-developer’s perspective–but that the IE team will hopefully address at least some of our issues. He has made useful inquiries and an honest attempt at suggesting solutions, taking into account the difficult situation the team is in of having to keep as much compatibility as possible."

    Have you checked out previous comments? Before charging me being malice, at least spare some time and make sure that I am being malice if you are really in good faith.

    It is no secret that some people do make malicious comments here. I am warning against the fact that not everybody who has an opinion on the issue will come here and comment here and it is also highly likely that some of those who come will come with malicious purposes as we see from the previous warning.

    You might have some assumptions about what web developers based on what you read on the web, but the web you read doesn’t necessarily represent the real world. So excuse me if I express my opinion on the issues as a developer. Many of the features of IE make the web work, I realize that a lot when lots of sites break with Firefox for small easily fixable issues. You have no excuse when things stop working. I don’t think a minority of "web developers" have the right to speak for all web developers, especially if those "developers" intentionally try to distort facts and claim that IE doesn’t support the standards because it named its Javascript as Jscript and it is no more Javascript!

  18. Anonymous says:

    Rory Parle’s own comment about this blog:

    "Microsoft just launched a new marketing site for Internet Explorer in the guise of a team blog. Why do I say it’s a marketing site? Try this gem:

    Some of us have our individual blogs today, but we also wanted to have one that was focused on what we do every day at work – make Internet Explorer the best way for browsing the web.

    Compare this to what an actual developer blog should be: Surfin’ Safari is Dave Hyatt’s blog about Apple’s Safari browser where he has recently discussed Apple’s additions to HTML for Dashboard. This led to responses from other web-related proffesionals, including representatives of Mozilla and Opera who made suggestions to make these additions more palletable. Dave made swift changes to Safari’s implementation to address these concerns and continues to work with the wider community to ensure Apple doesn’t break the web.

    Meanwhile Microsoft sits around patting each other on the back and saying how great their old and busted browser is."

    Yeah, it looks like Rory is a really nice person coming here to help IE Team and web developers to fix issues and move forward. Definitely I think IE team should separate those who really want to help, improve IE from those who come with malicious intents. No wonder why those bash IE turn out to be average Slashdotter who thinks that web developers want Mozilla/Safari/Opera. Ding ding, 90% of the users use IE and and many more develop just for IE. Rory, grow up first before accusing others for being malice. You are here just to disrupt what Microsoft developers are trying to do.

    Yet Rony deny that he is the typical slashdotter.

    "The fact remains that we’ve seen nothing of significant value from Internet Explorer in a long time. It’s far behind other browsers and even someone as biased as a team member should be able to see that right now it isn’t the ‘best way to browse the web’. Perhaps you meant that you hope it will be again some day. I have to agree with that; it could do nothing but good if Internet Explorer was to catch up with or surpass the current crop of great browsers.

    I’m not the Slashbot type who assumes everything out of Redmond is evil. I think it’s reasonable to assume that the developers (though not necessarily the management) are working to make a good browser. I just think some interaction of the kind seen from Apple would be better than hollow claims of superiority from the widely regarded "new NN4" "

    You disrespect Microsoft developers, you bash Windows users in your blog, you bash Microsoft, you praise Apple as if they are the only truth, you show your own zealocy and illogical way of thinking and then demand that Microsoft change IE the way you want, but not the way the real IE users and web developers want? I don’t like such apple zealots trying to dictate IE team how to write the browser. What is important is the real customers of Microsoft. When my friend who use IE can not access a web site because IE team decided to listen to slashdotters who bash them anyway and will continue to bash, I will be the first person to bash IE team for breaking things for the sake of satisfying people who hate them. For slashdotter this thing is never about giving feedback, it is an opportunity to bash Microsoft, as Jim and Jesus_Christ and you demonstrated. I don’t think you have enough dignity to apologize from IE team for accusing them being dishonest and putting this site just for marketing, but if you do so you may gain some respect other than people like Jim and other slashdotters.

  19. Anonymous says:

    "Many of the features of IE make the web work, I realize that a lot when lots of sites break with Firefox for small easily fixable issues."

    Yeah, possibly, because so many things are poorly developed/deployed. But this "making it work" is what created all that incompatibily mess in the beginning – browser vendors saying "oh well, the author must probably have meant this or that, so I’ll interpret it that way". If a user delivers wrong content, then it should not be displayed properly, period. (Guess why the XML standard is extremely strict in what it demands… and it works!).

    Same goes for meta-information like MIME types: If a server sends something declared as text/plain, then it should be treated as text/plain, even it contains html tags (and even if it has a .html extension!)

    Once again: fix these bugs, adhere to them standards – well, and then bring the usability of Mozilla/Opera to IE *SCNR*

    And no, this was not malicious.

  20. Anonymous says:

    well, take the mozilla kernel and put the ie frame around, it will work a lot better

  21. Anonymous says:

    I’d like the next IE to handle PNGs properly, that would make things a lot easier and full CSS2 support would be nice, most important thing ist i think not only to enhance the security-features but also to tell people how to use them. Most people i know don’t have the smallest knowlege what malicious websites can do to them. You can make software as good as possible but it will not help until the people understand how to use it properly…

  22. Anonymous says:

    "You disrespect Microsoft developers, you bash Windows users in your blog, you bash Microsoft, you praise Apple as if they are the only truth, you show your own zealocy and illogical way of thinking and then demand that Microsoft change IE the way you want, but not the way the real IE users and web developers want?"

    Look at the IE chat from MS. These are clear examples of answers, yes, but what’s the point when WE already know them? They have a lot of backward compat,., they don’t want to break it. Same with the Channel9 forum responses. And yet: I don’t blame the devs at all! Do you think that’s possible for someone?

    Now: is there anything wrong with doubting? NO. I find myself doubting if I will ever buy from Ford again since every car since the Villager has been crap value, crap quality. Do I HATE, DISRESPECT, <other NEGATIVE words> their developers? Guess.

    BTW your tone of posts all sound very sadistic and psychoish especailly in comparison to the people you’re responding to.

  23. Anonymous says:

    Come on guys, let’s keep this calm.

    As for the future of IE – yes, I’d like to see it move towards standards. Yes, this’ll break some websites. However, due to the size of IE’s market share the web site maintainers will have no choice but to fix their sites. Greater work in the short term for greater good in the long term. That’s an acceptable payoff I think.

  24. Anonymous says:

    "You can make software as good as possible but it will not help until the people understand how to use it properly…"

    Yes this would be good. There are so many people who would install the same things in Firefox, through XPI (although new versions almost eradicate this). MS brought computers to the consumer, I think they should educate them. Online forums can’t always do it, if you look at many security forums such as Wilders, they are BANNING Hijackthis logs from their site. If people learned what was safe computing…some people are willing to learn…some sort of tutorial when you first bring your new Dell computer home? I dunno…

  25. Anonymous says:

    > Jim claimed that IE doesn’t support Javascript standards. He uses the fact that IE’s term for Javascript is JScript but not JavaScript, to claim that IE doesn’t support standards.

    That’s untrue. I’ve stated that Microsoft’s implementation of ECMA-262 is called JScript and not Javascript. I’ve stated that Internet Explorer supports ECMA-262 through JScript.

    As far as other specifications go, I don’t think anybody is contesting the fact that Internet Explorer doesn’t implement CSS 2, PNG 1.0 or HTTP 1.1 properly.

    Alex appears to be holding a grudge against me. Feel free to assume I am some sort of enemy if you wish, guys; the points I make should stand or fall irrespective of who I am, Bill Gates, Theo de Raadt, or somewhere inbetween.

    For those interested, my only contact with Alex has been through the comments on this page:

    http://blogs.msdn.com/ie/archive/2004/07/21/190747.aspx

    If you have the patience, you can read through it and make your own mind up who the irrational one is.

    Alex, if you have a problem with me, at least try and confine it to a single entry on this website. I have no intention of persuing a flame-war across this website and getting in the way of productive conversation. By all means disagree with what I am saying, but shouting me down because you don’t like me is hostile and inappropriate behaviour.

    You remarked on the other page that you wished people would be constructive. Here is your chance. I pointed people towards a link describing Internet Explorer’s behaviour wrt. the Content-Type header, pointed people to the specification defining that header, and a document saying vaguely that there have been some changes to Internet Explorer’s behaviour post-XP SP2. Do you have anything to add to that?

    If not, do you have any technical basis for criticising my proposed header for deciding whether to use a new rendering mode? If you believe that there is software that will break upon transmission of this header, then by all means, please describe your concerns. But please, let’s remain productive outside of our little disagreement elsewhere.

  26. Anonymous says:

    "Yeah, possibly, because so many things are poorly developed/deployed."

    So? You are not producing software for standards, you are producing software for the average people to use and develop for. Software is not here to make things harder for people, we use/develop software to make life easier. Some treat standards as religion, but people don’t care about that crap. Your boss doesn’t pay you to support standards, he wants you to make things work. Period.

    "But this "making it work" is what created all that incompatibily mess in the beginning – browser vendors saying "oh well, the author must probably have meant this or that, so I’ll interpret it that way"."

    Have you worked on the history of how things have evolved? You claim that it is browser makers’ mistake because it is easy for you to say that. Many people still use text editors to edit their html files, including those using frontpage and dreamweaver users. Sometimes they try to copy paste some code, they mix up stuff and things stop working. Even you make mistakes, you may try your best, but probably your web sites also doesn’t show properly and probably you waste lots of time fixing things for browsers like mozilla. I know that I am more productive for IE than I am for mozilla. The rest of the browsers are not there yet. They should first catch up with mozilla. I have seen people struggling to make things work with HTML, XHTML and javascript. The number of such people are far more than me and you, and they are also collectively more important than you and me. Even so many slashdotters bashing Microsoft over standards are not good enough to make things work, they just think they are superior but in fact they are ignorant and arrogant. Overall it is much better to have tolerance. Besides I had seen cases where Mozilla hided my mistakes in html wheras IE didn’t. I literarlly corrected my mistake using IE. I mean your argument also doesn’t hold there. This is another myth that is not true.

    "If a user delivers wrong content, then it should not be displayed properly, period. (Guess why the XML standard is extremely strict in what it demands… and it works!)."

    XML doesn’t work on the web yet. People who use xhtml usually do so on simple sites like blogs and they really do very simple things. XHTML is one of the W3C’s biggest mistakes. David Hyatt also confirm that browsers are not good at XHTML as they are good at HTML. By making things more strict, they ended up hurting the web. That’s why Safari didn’t implement their stuff in XHTML and SVG, because it just doesn’t work yet.

    "Same goes for meta-information like MIME types: If a server sends something declared as text/plain, then it should be treated as text/plain, even it contains html tags (and even if it has a .html extension!)"

    I do remember having some issues with the MIME type, but nothing serious to talk about. I also remember having issues with Mozilla which was more annoying if you can’t control your servers’ settings. I think you might have a point here though, if I remember it correctly, the MIME issue I had was about sending files using a script. Because IE looks at the extension of the script I ended up implementing a module for Apache to serve those files properly. So, definitely that was an issue.

    "Once again: fix these bugs, adhere to them standards – well, and then bring the usability of Mozilla/Opera to IE *SCNR*"

    Treating Opera as a good browser is a direct insult to Mozilla. Opera is nowhere near mozilla. Mozilla as a browser sucks. Firefox made me to change to Mozilla, not the mozilla browser. You might be thinking that Opera is good either because you haven’t seen a problem yet, or that you simply hear others saying it. Usually slashdotters treat opera seriously, because it is a way for them to isolate Microsoft. They say like, IE sucks and other browsers rock, whereas really only mozilla is a great browser and the rest is crap. I am not sure about safari at this point though. It probably has lots of bugs still, and doesn’t support much of the javascript and dhtml stuff. I think the myth that Opera is a good browser come from this Microsoft bashing tactic to isolate IE as a target there. You see same tactic from Jim for example, he mentions that because Microsoft used a different term for Javascript (they had to because Javascript term is owned by Netscape) Microsoft doesnt support the stanards. Javascript is also a standard, called ECMA-262 and people refer to javascript as this standard, rather than the Netscape’s implementation of javascript. So they just try to bash Microsoft using different tactics, and as ignorance kicks in some myths become truth for some unsuspecting people.

    "And no, this was not malicious."

    I believe you.

  27. Anonymous says:

    Jim, you suprise me all the time, now you deny what you just said before. Anyway my point is not to have a flame war at all, but to inform people that instead of discussing whether Microsoft does support Javascript or not, which they do, we should be discussing really important issues. Mime type issue is an important issue, but nevertheless because I have seen you trying to confuse people intentionally, I wanted to warn people about the fact that some people make requests with bad faith. I think your request is another such malicious claim. I think the proper way of solving this problem is not to do what you want, but to look at real world cases and adjust the behavior based on that. For example, respect the HTTP headers when scripts are sending them, for example .php .asp etc….

    Regarding malicious requests, one example is Wise Guy who requested that his web site has access to all programs in the user’s computer, the ability to change the icons of the browser and so on. You seem to be in the same line with those people, so it is very normal for people to know where you stand on the issues. Remember you are the one who attacked me when I wanted to correct the claim that Microsoft does support Javacript standard properly.

    I don’t care about you in particular, since there are many more people like you who bash Microsoft and try to disrupt the process. You may want to be seen as a normal person trying to argue in good faith, but again there are significant number of people who simply come here to bash IE. Anyway let’s stay on topic and be positive. Watch out those slashdotters who make stupid false claims and bash Microsoft.

  28. Anonymous says:

    Jim: Shining Arcanine in the Channel9 Forum asked about the header to IE dev Dave Massy:

    http://channel9.msdn.com/ShowPost.aspx?PostID=13027#13027

  29. Anonymous says:

    Alex, but wouldn’t JScript support ECMA-262 instead? Not Javascript? I though things like document.all and such weren’t in Javascript.

    </not a web dev but I know how to close tags!>

  30. Anonymous says:

    Alex I never accused you of malice, I pointed out that you had accused Jim of it. You’re right that my initial blog post about this site was negative. That’s because up until that point no-one at MS had given me reason to believe that this was anything but a marketing excercise. All we had seen was comments about IE being great. When Dean comented on my blog I was quick to accept that he, at least–and by extension I assume the team as whole–is genuinely trying to make IE better. I fully support that attempt.

    As for this: "You disrespect Microsoft developers, you bash Windows users in your blog, you bash Microsoft, you praise Apple as if they are the only truth, you show your own zealocy and illogical way of thinking and then demand that Microsoft change IE the way you want, but not the way the real IE users and web developers want?"

    I didn’t disrespect anyone. I don’t bash Windows users, I am a Windows user. I see nothing wrong with pointing out MS’s flaws, or even making them the butt of some jokes. I have never used an Apple and I recently posted about the appalling state of some of their Windows software. I challenge you to point out zeal or lack of logic in what I’ve written. I have made no demands, simply suggestions and laments. And finally users could never be the victims of any of the changes I support.

  31. Anonymous says:

    Rory Parle, I really liked your positive post. I hope you keep that positive view so that we can come to a point where we can pressure Microsoft to update IE earlier than Longhorn. At this point Microsoft doesn’t want to release IE earlier, but I think they can do that. If we, as users, really focus on the real issues, rather than politics, I believe we will have a much better chance of making a real difference before Longhorn. Though, I am not sure if that would be relevant because Microsoft releasing a new browser is one thing, installing it to all the machines out there is another thing.

    Your initial comment about Microsoft developers was negative, and in my view unacceptable. It is one thing to make cheap shots at Microsoft, it is another thing to personally attack developers. I have seen Dave being extremely positive about these issues. However, people seem to be more interested in malice and negative rather than positive constructive communication.

    Regarding other stuff, I have read your blog, in few places that’s what I have seen, but never mind about that. No need to discuss them, let’s keep it positive as it is now.

  32. Anonymous says:

    Ok, here is your top 10 list of priorities for IE development:

    1. Security

    2. Standard Compliance

    3. Security

    4 .Standard Compliance

    5. Security

    I think you got the idea. Get everyone in the team to read and learn the relevant RFCs by heart, and follow them. Word by word! Forget about adding any "cool" extensions, any non-complient stuff, any blinkenlights. Get your basics right first. And this includes standard compliance.

    Have EXTENSIVE security audits by security experts (and not the junior programmers who just got hired)! Whenever in doubt if a feature might pose a security risk, even if extremely unlikely, DON’T DO THAT FEATURE. Everytime the beast is supposed to read or even write data to the local file system make tripple sure that this access is neccessary, and make ten times sure that it can’t hurt.

    Get extra training for your coders, even for the senior ones. Teach them that maintainability and code readbility is valued much higher than using a "cool" language feature, and that using a slightly slower piece of code (e.g. one which has an extra test) is still preferable over some clever optimized piece of code which might become a security risk if "that what never can happen" happens.

    Invest in first-class code analysis (lint-like) and runtime-checking tools (e.g. for array out of bounds checking and pointers running wild). Use them. Investigate every report or warning, and fix the root cause.

    Everytime the beast is supposed to execute some local code, don’t do it at all, every time it is supposed to execute some code loaded from remote (e.g. JavaScript) let it just run in an extremely restricted environment (Yes, S*u got it right with their sandbox approach). Do not grant any local access.

    Get rid of ALL ActiveX access – all, period. No discussions. The distingtion between secure and non secure ActiveX controls is a joke. History has shown that Microsoft is just not capable of even remotely guaranteeing the security of an ActiveX control for whatever reasons. So please give up on this. Be honest to yourself. You can’t do it, admit to it and be done with it.

    When installing, the default configuration should be the most secure one possible. I don’t want to use strong language, but IMHO it is appropriate to say that IE currently more behaves like a street wh*re, granting access to private parts for all.

    Build the browser around the security mechanisms. They should be the heart of the application. Don’t "add" security as a wrapper around core functionality. It is exactly this "ok, let’s just slap some additional security onto that code" attitude which made IE the number one security nightmare on windows for end users (OE comes close). Security is not an add-on, not something you bolt onto the code on the side. Security needs to be designed in the application right from the beginning.

    Do not tie the browser into the operating system (no, I don’t want to discuss your legal/political/monopoly issues here, just the technical ones). The deeper you tie the browser into the OS, the more the browser gets vulnerable to security issues in the OS. And also the OS gets more vulnerable to security issues in the browser. Whichever of your system architects, lead developers, or whoever you call them had that "bright" idea in the past, please don’t let him/her come near the new browser. He/she is apparently not capable of understanding the concept of system boundaries. Security means to have unpenetratable walls, not intimite tie-ins.

    Do NOT let markedroids or managers make security decissions. They only know one type of "security" decissions: lowering security. Establish some kind of penalty for even attempting for someone to lower security.

    If you ran out of time when inplementing a feature, don’t skip the security audits or take any other shortcuts. Skip the feature. Again, establish some penalty for everyone who tries to rush a feature or who tries to circumvent QA.

    You have to make some substential changes, as well in your development processes, but also in your (your team’s) attitude, if you want to build a better, compatitive browser. The first step might be very hard. Can you get the team together and let them in a corous loudly shout "IE is a FPOS. We will do better!" every morning? Only when the development team has reached this attitude they can do better. As long as there is a "well IE isn’t too bad" attitude, you will fail. (Ok, your markedroids in the propaganda department will fix it for you, but you will fail in technical terms).

    Ok, if you would have reall balls, you would take the firefox code as your basic code and develop it further 🙂 Of course open-sourcing your extensions :-)))

  33. Anonymous says:

    Ok, as a suggestion, drop the bloody flame war, take it to another site, email, etc. As someone who’s just now reading, it’s annoying to find on what should be a purely technical forum.

    Now, thought I’d post a few comments as a web developer. I originally wrote my own blog system, and have managed several websites, plus support a LOT of people with web development. I thought I’d make a few comments with regards to IE, compatibility, etc.

    First, I find on occasion that things that work in IE6 don’t work in IE5 – that’s understandable. BUT, things that worked in IE5 may not always work in IE6. There’s not too many things that happen this way, but I’ve found some of the more esoteric css tags tend to act this way.

    With regards to IE6 and some rendering issues, here’s an example of a site: http://www.furay.com/julia/ IE6 still has problems with the thing. The image in the top should be a link to the main site. It works fine under KHTML, Mozilla, etc. And based upon the source code, I don’t see any problems with the html/css. Stuff like this makes me want to strangle IE. I realize other browsers have issues, but at least they get updated a little more regularly, and it’s easy to tell people "Oh, you’re using Firefox 0.8? Upgrade to 0.9.2 – that fixes that bug". Right now, all I can say is switch to Firefox.

    Another set of examples has to deal with the refresh tags and caching. I worked on some code for a friend to handle image loading. From what we could tell of the server logs, IE was reloading the images regardless of having already loaded them. Then, we hit another fun issue. http://beanbagcentral.com/codicody/ – on this page, every refresh of the page should load a different set of text on top of the images. It’s all rendered correctly, but here’s the annoyance. Clicking on the link which references the page itself (it seems to be fixed by changing the link to "./" instead of the full URL) didn’t work. You can see this behavior by typing the same link in the address bar and hitting enter. However, the reload button works perfectly well.

    Here’s a last fun one – try the float css tag. It sorta works, but sorta not. I’ve hit issues where the image disappears (I can replicate it if need be) instead of floating properly. The text is wrapped around the image properly, the image just doesn’t show up. Other fun oddities – sometimes resizing the frame fixes it so it does show up.

    It’s a lot of this stuff that I find on a regular basis which have caused me to tell most of my friends to stop using IE. I guess the biggest issue is getting an actual fix for these bugs in a timely manner. That’s the biggest thing I hope that comes up – regular fixes for all the CSS bugs that crop up. In my experience, no matter how well you test a product before release, you’ll find bugs afterwards. It’s whether you fix those bugs or ignore them that defines whether you’ve got a good product. Thus far, IE isn’t a good product. I just hope that it becomes a good product….

  34. Anonymous says:

    I regret to say that I haven’t read all of the comments here, so I hope this isn’t redundant; however, I think it’s rather important. I’m viewing this page with Opera and I had quite a difficult time getting this message in, since the title, name and URL input boxes are each one character long and the comment box is rendered as about 10 columns wide. I’m not an IE basher by any means (I just don’t find it as functional as Opera), but I do find a little bitter irony in having trouble posting in a blog for IE obviously designed to /render/ in IE.

  35. Anonymous says:

    I’m curious as to whether you will be at liberty to discuss improved support for CSS 2 and 3? XAML and the other new proprietary features will be useless to me because I have to support Mac and Linux boxes in my webapps. Besides, the uptake of Longhorn will likely be pretty gradual, so I don’t expect that IE 7 will be even at 50% market share until late 2007 or beyond.

    However, if IE 7 supports CSS and XFORMS, we can code a cross-browser site, and let IE 6 users know that they can upgrade to Longhorn for full functionality as opposed to just recommending Firefox or Opera.

  36. Anonymous says:

    I use Firefox and everything looks fine here. Opera users complain that it doesn’t show properly. Which one is following the standards? Some people claimed that opera is as good as Mozilla and even better than IE. Where are those who claimed that Opera is standard complaint and better than IE. Forget about Opera competing with IE, it is nowhere near Mozilla. I had the same exact experience, yet alistapart and others claim that opera’s support is better. Clearly we can’t trust such sites. They are lying about these facts.

  37. Anonymous says:

    Alex, perhaps you are unable to see that compliance is not always /the/ most important thing. I prefer Opera for two very simple reasons over Firefox, Mozilla and IE:

    1.) Opera has state saving. When I must reboot or close Opera for any reason, it always gives me the option to open in the exact state (all tabbed websites) as when I closed it. I am a /heavy/ tab browser, so this is very important to me.

    2.) Opera is more stable than Firefox. Firefox would quickly crash once I pushed it up to 30 or 40 tabs, which I regularly do. Opera does not. It has yet to crash or hang on me, something I cannot say from using Firefox.

    My point, Alex, is that the little quirks about displaying, whether they’re compatibility related or otherwise, are relatively trivially if they aren’t on a broad scale (like pre-Opera 6). Right now, Opera as a piece of /software/ suites my needs, whereas Firefox or IE might give a little more liberty to different uses of code.

  38. Anonymous says:

    Would you look at that, I recognize the writing style, arguments and narrow view of Alex!

    He sometimes uses the name Daniel and tends to troll around blogs like the one at designbyfire.com a lot. Just ignore him, he’s not to be taken serious. I repeat: ignore him.

    This will be my first and last post on this matter, so Alex/Daniel/whatever, go throw your crap at me if you want to, I’m already gone.

  39. Anonymous says:

    Love that these threads keep disintegrating into shouting matches between 2-4 participants. You know, you guys would be better served just giving out your email addresses to each other and taking this on in private…

    The "he said, she said" thing doesnt’ help the IE team or the readers here very much.

    While I applaud the IE Team for not removing these comments, I’d encourage everyone to remember what this blog is for: communication between individuals and Microsoft, internal ramblings from the IE Team, etc.

    It’s not a PR machine, it’s not a place for you to (once again) say "support this!". They know what they need to do. If you read long enough you’ll find that out I’m sure. I know several of the team members to varying degrees and I assure you there’ll be value here.

    Enjoy. I’ll be leaving my encouragements for the team in private (and my thoughts) so they don’t get lost in the sea of arguments and "do this now or I won’t believe you forever!" posts.

    Have a great weekend everyone, hope everyone enjoys Bourne!

  40. Anonymous says:

    Man oh man.. I was just reading up SP2.. this is going to be a real kick in the "junk" for me. I am primarily an Intranet developer and we have several unsigned ActiveX controls – but are allowed because of the "site restrictions" for Intranet..

    It’s funny to see each side of developers. The Internet people want less functionality and stricter security and as an Intranet developer, I want just the opposite. Where I work, people want web apps to work like windows apps, but don’t want to install anything. That’s a tall order, and some pretty n-a-s-t-y javascript has been written in my days to accomodate that.

    If the IE team is anything like the VS.NET team, I will be quite excited to see what comes out of that group in the next year or two in terms of CSS and scripting support.

  41. Anonymous says:

    Another blog that could be aspired to: The JobsBlog by the Gretchen and Zoe (blogs.msdn.com/jobsblog). They’ve found their voice and silenced the critics purely because they’ve done what they set out to do with the blog: tell the world what’s happening in MS recruiting, why it’s happening and how they hope things’ll change (different from making forward looking statements).

  42. Anonymous says:

    Jeremy, to be fair those "do this now!" posts are less then 1/8 of the content here.

    However–with IE only being released in service packs until Longhorn, doesn’t this mean IE 7 (6.5?) will be left untill Longhorn? Will another service pack be coming for XP?

  43. Anonymous says:

    > I use Firefox and everything looks fine here. Opera users complain that it doesn’t show properly. Which one is following the standards?

    It’s possible that they both are. input elements are defined as being 300px wide at 10px font size. If somebody has set a minimum font-size, this will bump the size of the text up, while keeping the input element the same width.

    Having said that, I believe older versions of Opera exhibit this behaviour when you attempt to set a width on them using CSS. I’m not sure if that bug still exists, since I stopped this practice. Try a user stylesheet with input[type=text] { width: auto; } to correct this problem.

    Needless to say, there are many people who would say that the font size shouldn’t be set in pixels, and even if it was, that it shouldn’t be set so low, but that’s an argument that I don’t want to get into here, and the relevent points are easy to find in any web authoring forum. I suggest anybody currently using pixels to size text investigate the issues when they get the chance.

    > I think the proper way of solving this problem is not to do what you want, but to look at real world cases and adjust the behavior based on that. For example, respect the HTTP headers when scripts are sending them, for example .php .asp etc….

    That isn’t feasible, I’m afraid, as file extensions aren’t a reliable indicator of whether something is generated by a script or not. I have, in the past, maintained websites with dynamic pages that have .html extensions, and websites with static pages that have .asp extensions. It’s quite a common practice, and you can’t infer the correct behaviour from whether something is script generated or not in any case.

    Also, could you clarify whether you are talking about the Content-Type header or the X-Internet-Explorer… header? Right now, what we are both saying applies to both topics.

    Thanks for the pointer, JP. They appear to be talking about doctype switching again, so no wonder they are concerned about regressions. Do you know if they have considered an approach like the one I suggest, where authors explicitly mark documents that should be rendered according to the specifications? It seems to me that there is no possibility for regressions here as authors have to actively "switch on" support for the specifications.

    Once more, I’m interested in hearing any criticism of the X-Internet-Explorer… header solution. Do people think that it’s an unfair burden on people that want to author to the specifications? Do people think that it won’t work? Do people think that it will cause regressions? As far as I can tell, it’s fine in all these respects, so I’d be interested in hearing one way or the other if people are thinking along the same lines.

  44. Anonymous says:

    Alex Almeroth:

    I’m a Firefox user. But I would say Opera is very excellent as well. Perhaps you have not checked out Opera’s 7.5.x versions. The (probable) reason why Opera does not render this page properly and Firefox does is that Firefox has a better "almost standards" mode. The DOCTYPE of this page is HTML 4 Transitional, which triggers Firefox’s "almost standards" mode rendering; it’s an attempt to make sure sites render properly. (More information here: http://www.mozilla.org/docs/web-developer/quirks/doctypes.html) Additionally, as people have mentioned, this page doesn’t validate; there is no actual correct way of rendering it, since I don’t think HTML 4 specifies rigorous error handling procedure. When I code XHTML 1.0 strict, 99% of the time it displays exactly the same in both Gecko and Opera.

    The reason why some people state Opera has better standards support is that I believe Opera supported more CSS earlier than the Gecko engine for Mozilla. I don’t know if that part is true any longer.

    To the IE team:

    An easy way (at least, it seems to me) to ensure backwards compatibility is to do DOCTYPE sniffing, which you have already done for the box model. (more information on IE’s box model switch: http://www.digital-web.com/articles/toward_a_more_standards_compliant_ie/) People coding with a strict DOCTYPE should know what they are doing. You might even ignore HTML strict and conservatively only render XHTML 1.0 strict and later correctly. The number of sites coded in XHTML strict that depend upon incorrect rendering is probably extremely low.

  45. Anonymous says:

    Jim, I honestly DON’T know. I’m don’t understand anything other than the most cursory knowledge someone could have of the web shtuff and how it all works. (honestly I’m here because I’m interested in the internet….stayed because I like you guys!). Just remembered seeing this on C9 and voila the search tool worked pretty fine! Anyway, would "content load" be something like mimetypes? <–prepared to look very ignorant here.

    Also, would this X-explorer-type stay even after more pages switch over and it is (hopefully) a majority standard? And if it ‘s so simple I wonder hwy the IE team hasn’t implemented this…like you Jim I’m inteirested in all of your posts!

    (Ignore the above if it seems really ignorant, I’m about to Google this stuff)

    Looks like caiuschen has a far better answer than me though. You web designers…! 😉

  46. Anonymous says:

    If the site looks different on Opera vs. IE, the first thing you should do is look at the HTML (view source) and see if .NET pumped out bad HTML. The .NET framework is made by default to send out horrible HTML to every browser other than IE if you use attributes for width, height, etc in the asp:textbox and other webform tags. Good developers will get around this by changing the browser caps section of the machine.config or adding the updated section to the web.config (because MS relegated the upkeep of that file to a company who has NEVER done anything other than put up a page that says "we’re working on a new version"). Firefox/bird/moz is usually covered by the developers’ custom browser caps section but Opera is often overlooked (I admit my guilt there).

  47. Anonymous says:

    > Anyway, would "content load" be something like mimetypes?

    I don’t recall seeing anything about the Content-Type/mime-type issue on the link you provided.

    The crux of the matter is that servers usually tell browsers what type of file they are sending (referred to as a media type or mime type) by using a Content-Type header. Browsers are only supposed to guess if the server didn’t tell them. The HTTP 1.1 specification makes that very clear.

    Internet Explorer, under some circumstances, goes against the HTTP 1.1 specification and ignores what the server is telling it. The first link I provided gives details on how this mechanism works.

    The third link I gave mentions that they’ve made some changes, so I am hopeful that this issue is fixed in XP SP 2. It is very nice to see the Internet Explorer team make progress on this issue (thanks guys!). However, it still talks about file extensions, which is confusing as the file extension plays no part if servers are sending Content-Type headers. I suspect that there still may be special cases where the Content-Type header is ignored, and I’d just like it clarified.

    I’m also wondering if non-XP users will also get this fix at some point.

    [Note that doctype switching is a seperate issue to the Content-Type bug.]

    > Also, would this X-explorer-type stay even after more pages switch over and it is (hopefully) a majority standard?

    That’s a possible downside. We don’t want to be lumbered with having to send this header out for eternity. What I would expect to happen, and what is not entirely unreasonable, is that a new version of Internet Explorer would come out that honours this header and the relevent specifications. The version after that would drop "quirks mode". The version after that would drop what is currently labelled "standards mode" (i.e. the best Internet Explorer 6.0 does at the moment).

    Doctype switching allows for the possibility of widespread regressions. For instance, when Internet Explorer 6.0 was released, documents expecting "quirks mode" that were now rendered in "standards mode" had trouble with centred, nested tables not doing what they did previously. That’s a regression, and that’s the kind of thing that I’m sure Microsoft are keen to avoid.

    The fundamental problem is that the doctype line in HTML documents is there to declare what type of document it is. Microsoft attempted to make generalisations about what was and wasn’t intended to be rendered by browsers that complied with the specifications. This is something many people disagree with. I disagree with the fine details, but not with the general approach – it’s not logical from a purely theoretical perspective, but it’s pragmatic and solves the problem at hand.

    However, faced with bolting another type of rendering on top – let’s call them "quirks mode", "standards mode", and "perfect mode", it’s becoming less and less pragmatic to make this kind of generalisation, as people are already using certain document types to trigger certain modes, and expecting certain behaviour. In a way, current documents are "polluted" with the expectation of certain modes, especially as other browsers also do doctype switching as well. That is why I feel it would be far safer to keep rendering exactly as it is, but allow web authors to trigger "perfect mode" by indicating such in HTTP headers or equivelent meta elements.

  48. Anonymous says:

    Far above, Rory Parle asked for preview functionality for comments. From what I can tell, Scott Watermasysk is responsible for the functionality on this site (see http://blogs.msdn.com/, and http://scottwater.com/Blog/about/contact.aspx). His response:

    From: Scott Watermasysk

    Subject: RE: preview functionality in comments?

    It does not yet exist. It will eventually…

    Currently no HTML is allow in comments. Links will be transformed into full

    href’s. Line breaks = <br><br>.

    HTH,

    Scott

    So… sounds like great functionality. Something beyond my control. Scott Watermasysk’s aware and sounds excited to do it.

  49. Anonymous says:

    Hello IE devs,

    I want to say that I think it’s great that this blog has been kicked off, and I hope that you guys maintain this amount of dedication to interacting with the community.

    I am a web developer and a Firefox user who used to be a pretty avid IE user. Open-mindedness caused me to switch browsers, and it could cause me to switch again. (Btw, on behalf of the Firefox community, sincerest apologies for all the flaming…how embarrassing…).

    Anyway, as I said, I am a web developer, and know the standards extensively. I think people are ignoring the fact that, although IE 6 does not support CSS 2 on up, its support for CSS 1 is unarguably pretty decent when the browser is running in Standards compliance mode. I have made a number of XHTML/CSS layouts that run very well in IE 6 with little to no tweaking of the code…providing I can get IE into the standards compliance mode.

    Anyway, I know that your team is focusing mostly on security issues at the moment, but I was wondering if there is any possibility of the following two (relatively simple) problems being resolved in the up-and-coming patch.

    1. The presence of the XML prologue at the top of XHTML documents seems to throw IE 6 back into quirks mode…Without knowing anything about the IE 6 code, I might guess that this wouldn’t take too much work to correct.

    2. XHTML documents served with the MIME-type application/xhtml+xml seem to cause a file download in IE rather than the page be loaded by the browser. Instead it would be better to parse the document in standards compliance mode.

    Anyway, I just wanted to throw that out there and to commend the IE team for this new blog. I’ll be looking forward to reading what the team has to say in the future.

    Cheers,

    Darren

  50. Anonymous says:

    > 1. The presence of the XML prologue at the top of XHTML documents seems to throw IE 6 back into quirks mode…Without knowing anything about the IE 6 code, I might guess that this wouldn’t take too much work to correct.

    I’m sure it wouldn’t take them long to correct it. The question is whether or not they should. Doctype switching has been around for a while now, and making this change would cause a lot of documents to go from being rendered in "quirks mode" to being rendered in "standards mode". If the documents are relying on quirks, this will introduce regressions. If there’s a single message that Microsoft are being quite clear on wrt. improvements in Internet Explorer, it is that they want to avoid regressions.

    > 2. XHTML documents served with the MIME-type application/xhtml+xml seem to cause a file download in IE rather than the page be loaded by the browser. Instead it would be better to parse the document in standards compliance mode.

    They can’t do that without violating the XML 1.0 and CSS 2.0 specifications. The XML specification is very clear in stating that if a document isn’t well-formed, the parser should throw a fatal error instead of continuing. "Standards mode" doesn’t do this. Furthermore, CSS makes certain concessions to HTML documents that it doesn’t make to XHTML. "Standards mode" doesn’t take this into account. I believe there’s also an issue relating to the DOM and namespaces, but I’m not familiar enough with that subject to comment upon it.

    The recently published XHTML FAQ mentions the CSS issue:

    http://www.w3.org/MarkUp/2004/xhtml-faq#css

    Implementing application/xhtml+xml properly isn’t simply a case of getting the code through an existing HTML parser. That is almost certainly a matter of a single line of code (it is in Konqueror, anyway). There are very specific things it must do to support XHTML properly, and it may very well be a lot of work to implement it robustly in Internet Explorer.

  51. Anonymous says:

    This is off topic, but perhaps will help keep the flaming to a minimum here. This bookmarklet will disappear any comment from our buddy Alex.

    javascript:(function(){var l=document.getElementsByTagName(‘a’);for (var x=l.length-1;x–;){if (l[x].innerHTML.indexOf(‘Almeroth’)!=-1) l[x].parentNode.parentNode.style.display=’none’}})();

    Enjoy

  52. Anonymous says:

    Thanks for your insight, Jim.

    I am aware of the repercussions of proper handling of HTML as an XML document.

    I didn’t express this very clearly, but I am not asking the IE team to rewrite their XML or HTML parser, but only to make some small alterations.

    IE already contains logic for rendering mode switching via DOCTYPEs. I am just requesting that the XML prologue not break the switch to standards compliance mode.

    Furthermore, I don’t agree that allowing for an XML prologue would break sites that rely on quirks. I simply cannot imagine a site that implements quirky/nonstandard markup or code, and then contains an XML prologue at the top, or a DOCTYPE declaration for XHTML for that matter. Why would anyone do that?

    And as for the issue of the application/xhtml+xml MIME-type, you are correct about what you said, and of course in an ideal world Microsoft’s XML parser would be equipped to receive the application/xhtml+xml type and do a very strict parsing. However, since I realize there is little chance of the IE team being able to rewrite large portions of the parser(s), for the mean time I feel that it would be a better trade-off for IE to accept the document in standards compliance mode (provided an appropriate DOCTYPE) than to limit the MIME-type options available to developers.

    Then again, maybe you’re right and I should just shut up and continue having my server test HTTP user agents, in order to protect IE from application/xhtml+xml… 😉

    Cheers,

    Darren

  53. Anonymous says:

    > Furthermore, I don’t agree that allowing for an XML prologue would break sites that rely on quirks. I simply cannot imagine a site that implements quirky/nonstandard markup or code, and then contains an XML prologue at the top, or a DOCTYPE declaration for XHTML for that matter. Why would anyone do that?

    Who knows, but I’ve seen a damn sight weirder code than that!

    I’m not saying it’s impossible that they’ll fix that bug, but I expect that regression is their primary concern as it really does alter rendering in a large way for a large number of documents.

    > for the mean time I feel that it would be a better trade-off for IE to accept the document in standards compliance mode (provided an appropriate DOCTYPE) than to limit the MIME-type options available to developers.

    I disagree here. Once that happens, you’ll see websites with malformed XML springing up left right and centre that work in Internet Explorer and fail completely in browsers that implement XML 1.0 properly. You have to consider the network effect and not just the immediate effect on functionality.

    > Then again, maybe you’re right and I should just shut up and continue having my server test HTTP user agents, in order to protect IE from application/xhtml+xml… 😉

    Actually, I’ve recently become aware of an XSLT trick you can use to get around this issue, as long as you don’t mind using application/xml instead of application/xhtml+xml

    http://www.w3.org/MarkUp/2004/xhtml-faq#ie

    I haven’t experimented with this yet.

    Personally, I try and avoid content negotiation and just stick with text/html – it’s much more cache-friendly.

    PS: It’s not immediately apparent, but are you using the User-Agent header to determine whether to send application/xhtml+xml or not? That’s prone to failure as the User-Agent header isn’t dependable. The Accept header is the proper way to determine which media type to send (and don’t forget to send an appropriate Vary header so that caches don’t screw up).

  54. Anonymous says:

    Jim,

    Thanks for the response… I think I stand adequately corrected on a few issues. I was too naive to consider what would happen if people tried to serve XML but didn’t know what they were doing…I think, as you say, the results could be disastrous, and failure to render the pages by XML supporting browsers such as Mozilla would then be perceived as a weakness….*shudder*

    I am especially thankful for your tip about the Accept header… Looking forward to your insight in the future as well.

    Cheers,

    Darren

  55. Anonymous says:

    Jim, I appreciate that post of yours re: mime/type/content load, etc. Learned a bit here, learned a bit here and just all jumbles together!

    Anyway, google is good, as I have learned 😉

  56. Anonymous says:

    You’re welcome guys 🙂

  57. Anonymous says:

    Jim is talking nonesense again. He is talking to Slashdot folks here. He is probably talking about a name difference here again, as he claimed that JavaScript and JScript are different, because the first has more characters. He is just being malice for the sake of it. There is no problem in what he says. IE definitely supports HTTP, that’s how internet works anyway. I guess what he means is that IE uses the term http whereas he is using the term HTTP or something like that. It is just a name difference. Jim should put more content out there to convince us that he is not talking about some name difference but actually something that matters for IE, because he seems to be keen on naming differences rather than real core issues. He should post his comments related to name differences to Slashdot not here although there are enough of them are here, majority are still on slashdot.

  58. Anonymous says:

    > IE definitely supports HTTP

    I said that Internet Explorer violates RFC 2616. Directly from RFC 2616:

    "If and only if the media type is not given by a Content-Type field, the recipient MAY attempt to guess the media type via inspection of its content and/or the name extension(s) of the URI used to identify the resource."

    Directly from Microsoft’s description of Internet Explorer behaviour:

    "A MIME type is ambiguous if it is ‘text/plain’, ‘application/octet-stream’, an empty string, or null (that is, the server failed to provide it)."

    So, in the case where the Content-Type header is provided as text/plain, the HTTP 1.1 specification says that Internet Explorer is not permitted to guess. From Microsoft’s description, Internet Explorer labels text/plain documents as "ambiguous". What does that mean? Well there’s a list of steps it goes through:

    "1. If the "suggested" (server-provided) MIME type is unknown (not known and not ambiguous)…"

    Okay, we can skip that step, as text/plain is ambiguous.

    "2. If the server-provided MIME type is either known or ambiguous, the buffer is scanned in an attempt to verify or obtain a MIME type from the actual content."

    There we go. The server provided the text/plain media type, the HTTP 1.1 specification states that, under these circumstances, Internet Explorer is not allowed to guess by content inspection, and Microsoft themselves say that Internet Explorer guesses the media type by inspecting the content in these circumstances.

    Alex has a grudge against me. Alex, in future, please substantiate your claims against me. This one was clearly an unsubstantiated personal attack and not a valid criticism of the point I was making. Let’s try and keep this constructive, okay?

  59. Anonymous says:

    Nice to see some of the Microsoft people start IEblog. If I might make a suggestion, how about posting news about security updates for IE, 2000 and XP when they are released. It would be cool to hear about the development and issues with IE and also be notified when it needs to be updated.

    Keep up the good work.

  60. Anonymous says:

    Jim, please stop whining, either discuss the topic or stop pretending to be innocent. You have lied numerous times before, you didn’t make your point clear.

    The guideline you specify is not working in real life, thus IE helped us to correct the problem. Many people do not setup their servers correctly, thus they send pdf files etc… as text/plain and so on. IE team correctly addresses the issue and it is fully complaint with the standard. What you are saying is that, you want people to waste their time with server configuration. Remember today web means IE, majority of people choose to use IE because it works. Other browsers couldn’t solve as many problems as IE solved, thus people choose not to use them very often. The only thing IE team should do here is to make that guess more intelligent, like don’t guess when scripts are sending. People love such features, thats why 90% of web users use IE. You just can’t break things by citing an arbitrary statement in a document. This is similar to your claim that IE doesn’t support JavaScript because its name is JScript instead of JavaScript. You really should work to convince people to go to other browsers, and they will end up going right back to IE, because other browsers do not work properly. For example, Firefox in many cases refuse to render html pages and instead show the content of the html page. That’s just unacceptable to people. Firefox misinterpreted the standard here and refused to fix the problem. Please, stop whining like slashdotters, and discuss the real topic.

  61. Anonymous says:

    Alex, I posted links to the relevent documents. I quoted sections of them that directly conflict with each other. I have more than adequately shown that Internet Explorer is violating RFC 2616 to anybody capable of understanding the issues. I offer Microsoft’s own words, and the revelant specification as evidence to support my claims.

    If you want to stick your fingers in your ears and chant "liar, liar, liar", there’s nothing I can do about that. But you haven’t offered a shred of evidence to disprove my statements about Internet Explorer and RFC 2616. The only thing you have offered are insults and lies.

    Either participate in a constructive fashion, or be quiet so that the rest of us can.

  62. Anonymous says:

    "I have more than adequately shown that"

    No all you have shown us a piece of document. It doesn’t mean IE is doing the wrong thing. I have on the other hand showed you that thanks to IE’s that feature web works today flawleslly. Firefox doesnt work properly in most of the cases for example. Second you haven’t showed us that, anywhere in that document it says you have to treat such headers as it is otherwise you will be non-complaint.

    "Either participate in a constructive fashion, or be quiet so that the rest of us can. "

    Quit whining and carry on so we can have some entertainment.

  63. Anonymous says:

    from Alex Almeroth:

    > No all you have shown us a piece of document. It doesn’t mean IE is doing the wrong thing.

    It’s just RFC 2616 so when you put:

    [object data=’/somevirus.exe’ type=’image/jpeg’]A nice picture[/object]

    (since html is not allowed I use brackets) IE is going to try to execute somevirus.exe and Firefox don’t work properly because it will not display any image? It’s simply ridiculous and don’t use the server parameters as an excuse for bad web developpers and a bad implementation of RFC 2616.

    To keep on the initial subject:

    There is now a lot of communication from the new IE team but as it was stated it’s right now more marketing than technical and there’s no roadmap at all. Your situation (IE developpers) is not easy at this time, but as a web developper the only thing I care is : do you want to implement web standards as it was done by Tantek Çelik and other people working on the IE-Mac Team?

    I mean do you really want? If so please tell us, a lot of web developpers are waiting for a strong signal *now*. I don’t want to stress you but the easy way at this time is just to help users to switch to Firefox or Opera and forget about IE ; we will not wait for Longhorn.

    I think you have now all the resources, ideas, hints on channel 9’s wiki and forums to make a decision but let us know [soon] please.

    (please forgive about my bad english, I’m french 😉 )

  64. Anonymous says:

    Francois, your English was perfect, and I think a lot of people are interested in some type of roadmap too. Hoping the pointy heads will accede if that is the bottleneck 😉

  65. Anonymous says:

    If you are looking for a better blogging tool have you considered installing WordPress at http://www.wordpress.org. It has a lot of the functionality you could use, it generates valid xhtml and uses separated CSS. Being open source and released under the GPL it is free, will run on any platform, and it is open to the end user to hack like there is no tomorrow. Ideally though you would need to run it on a linux / apache combo both of which are available for MS use with free downloads, no licensing required, and unlimited tech support from an informed and involved community. It is unlikely to bring down your whole network, it is easy to install and does not require *upgrades* , Active X, or proprietary plugins of any kind. You guys would have a lot of fun using it, and I would not have to keep scrolling to the right to read this thing. What is this asp net weblogs stuff anyway?

  66. Anonymous says:

    Root,

    When you say "any platform" I don’t think you were really thinking. I’ve been looking for a Blog tool that can use a Oracle backend for a while now, and they don’t seem to exist.

  67. Anonymous says:

    Security aside, W3C Standards are what it’s all about…until IE (developed by Microsoft, which has the most resources on Earth and, I believe, is a member of that standards-development body) will accept that developers want standards for valid, everyday life reasons (and not just to bash Microsoft), then developers will indeed bash MS.

    Adobe and Macromedia are just two examples of companies that have excellent relations with their customers: Photoshop artists are respected, Dreamweaver developers are courted, etc. – none of us outside of Microsoft need Bill Gates to come do our dishes for us (we’re not asking MS to do *everything*), but the year-after-year, arrogant disregard for what *actual* CUSTOMERS want is beginning to reap its proper rewards.

    A Wish-List for IE?:

    1 – FULL native support for CSS 1, 2, and as much of 3 as possible

    2 – FULL native support for MathML (yes, real people actually do want, and *need*, this) – think scientists, mathematicians, researchers, statisticians, etc.

    3 – FULL native support for PNG

    4 – FULL native support for SVG

    5 – FULL native support for XML *and* XHTML (delivered either as html or xml)

    6 – Tabbed Browsing

    7 – pop-up blocking (customizeable by site)

    8 – NO ActiveX, Smart Tags integration or dependencies

    9 – NO integration with the OS

    10 – NO integration with other Office products

    11 – SMALL Core Footprint, with the ability for plug-in/extension development to provide for more specific features

    12 – DROP Alexa: it’s spyware

    13 – FIX IE’s printing bug, where it prints something like "file://C:DOCUME~1(name)LOCALS`1TempNNL02KNO.htm" instead of simply printing a web page’s actual URL

    Until such time as IE sports some of these *basic*, *necessary* features, here are a few resources that are more useful than the way-out-of-date marketing…opps, sorry, technical, info provided by Microsoft:

    http://www.google.com/search?q=IE%27s+shortcomings&sourceid=firefox&start=0&start=0&ie=utf-8&oe=utf-8

    http://css-discuss.incutio.com/?page=CategoryBrowserBug

    http://www.positioniseverything.net/index.php

    http://www.blooberry.com/indexdot/css/supportkey/syntax.htm

    http://www.quirksmode.org/css/contents.html

    http://westciv.com/style_master/academy/browser_support/index.html

    http://nemesis1.f2o.org/bugs

    http://www.stopdesign.com/log/2004/01/26/ie-factor.html

    http://www.galaxygoo.org/blogs/archives/000528.html

    http://geocities.com/csssite/index.xml

    http://www.mozilla.org/start/1.0/guide/product.html#standards

    http://www.mozilla.org/products/firefox/why/

    http://texturizer.net/firefox/extensions/

    Sorry, but you’ll need Mozilla, Firefox, Netscape 7.1, or other Gecko-based browser to see some of these sites properly, since IE doesn’t fully support standards such as MathML, SVG, PNG, CSS, etc.

    PLEASE, listen to us…some of us actually DO like IE…but it has caused more trouble than any other single piece of software is many of our lives…and I’m not saying that to be mean or make you feel bad.

    I would LOVE to see an IE with incredible standards support and tough-as-nails security…wouldn’t happy developers constitute a small (or maybe not-so-small) army of marketers eager to tout the benefits of such an IE to all *their* customers, if they had a reliable browser from you?

    That just seems like good business.

  68. Anonymous says:

    The truth is that Microsoft has a habit of producing bad software – probably due to being closed sourced and the limited minds can’t produce anything else. It "looks" pretty on the outside, but, that’s as far as the beauty goes.

    Linux is a wonderful product, which is winning the OS war, by the way, despite what you want to believe, because practically the entire world is working on it in some way – it’s open source. The source code for Microsoft products is kept secret and hidden, thus, severly limiting the amount of people who can work on it and, therefore, limiting its quality.

    My suggestions is thus. Open ALL the source code for Microsoft products, thereby allowing more folks to improve it. Or keep it closed and continue to wane until such time as Microsoft products are no longer wanted by the public at all.

    Microsoft Corp. is going to lose the revenue generated by it’s Microsoft Windows OS’s. Either by opening the source code to the public, or by the public demand for the products dying out. Revenue generated by MS Windows OS’s is going to be lost, one way or another.

  69. Anonymous says:

    Hiya!

    First, thanks for offering some webspace where web developers can ramble about IE. We really are desperate to see some improvement soon!

    Therefore, my recomendations after reading _all_ the posts are:

    1. Pay attention to the suggestions by fedUP (he summarises almost all of the others).

    2. Talk to CSS Gurus!! Talk to Tantek Celik, get E. Meyer Books, and try to make ALL his examples work in Strict Mode. That would assure you almost all CSS1, 2 (and that means pure CSS navigation menues, fixed backgrounds, etc!!!). And yeah, as much as CSS3 possible would be nice too 🙂

    3. Accessibility could be a good marketing kick. Check what "Designing with Web Standards" have to say about all the CSS and accessibility.

    and, fix that whitespace annoying bug/behaviour in anchors, li’s, etc!! When doing navbars and web menues with CSS, we use a lot of A and UL/LI, so it gets out of hand when you cannot put a breakline between them to write cleaner code…

    Anyways, thanks for the big effort in fixing IE. I for one will cry loudly when I see IE behaving "almost perfectly". And I am sure you guys will manage.

    Thanks again!

  70. Anonymous says:

    > Talk to CSS Gurus!! Talk to Tantek Celik

    Tantek used to work for Microsoft, on the Mac Internet Explorer team, but no longer works at Microsoft.

    > And yeah, as much as CSS3 possible would be nice too 🙂

    Actually, as much of the CSS 3 specifications _that have made it to Recommendation or Proposed Recommendation status_ as possible would be nice.

    > Accessibility could be a good marketing kick.

    Internet Explorer already allows you to ignore fonts and colours, etc. It also lets you use a user stylesheet. Mozilla derivatives have an *extremely* user-unfriendly way of using user stylesheets.

    > fix that whitespace annoying bug/behaviour in anchors, li’s, etc!

    Please, yes.

  71. Anonymous says:

    Installing Branded IE7 on Windows XP Service Pack 3

  72. Anonymous says:

    Internet Explorer 8 Beta 2 Coming in August

  73. Anonymous says:

    Internet Explorer 8 Beta 2 Coming in August