Wow, seems everyone is going back and forth on side issues, not the technical WTF at the core of the story. Is there any legitimacy to an arbitrary stack limit, particularly such a low one, and especially in a way that's inconsistent depending on the type of object the function is attached to?
This is only exacerbated by the uselessness of the available debugging tools and Microsoft's general unhelpfulness.
vsync - Thanks for the feedback; We have made massive improvements to the developer tools we ship with IE. Please take a minute to download & install the latest Platform Preview from here - http://ie.microsoft.com/testdrive/ - and give it a whirl. If you hit the F12 key, you should see the developer tools show up, and you will be able to see the tons of improvements we have made since IE8.
If you have any suggestions or comments, please reply back. There are many open channels of communication, e.g. the IEBlog, Connect, etc. but as an FYI, I will also be looking at responses in here, within news.yc
Ravi, even though Microsoft browsers make my life as a web developer a misery and I hate your employer's tactics with a passion, I appreciate you putting your head on the block here in a humble and courteous manner.
Now all your team has to do is to ship a great IE9.
We are doing all we can; no, I'm serious - this isn't a cookie cutter, "Here's $nice_boilerplate_response, now go away" thing. We are severely limited by law (both US and international) about shoving updates down people's throats when they don't want them. We are limited by law about what we can call a "Security Update", and what we cannot ... We have absolutely no control on what IT Admins decide for their corporation's intranet as well ... etc.
While it would be awesome to silently push down IE9 onto everyone, I unfortunately doubt that it is going to happen (I'm not a lawyer ...)
I'm no lawyer either, but can't you change IE9's EULA? In Google Analytics I never see a great range of Firefox and Chrome versions - they seem to update fine. 80 - 90% are on the latest version.
People who use Firefox and Chrome care about having a good browser. People who use IE are content to use what came default with their computer and might not even care about running automatic updates.
What? I'm serious--you can't compare Firefox and Chrome users to IE users on how many of them keep their browsers up to date, because FF and Chrome users by definition are people who take an active role in downloading and installing web browsers. It's a selection bias, not a slam against IE or IE users.
Chrome automagically updates itself all the time without any end-user intervention at all. IMHO, that's the surest way to keep everyone current. The problem is that corporate IT shops would never go for that because they have too many brittle 3rd party applications that could break.
Google's Chrome is currently a fringe browser and one that was never the subject of antitrust lawsuits. There's an unfortunate disparity between the two situations; understandable, but still a pain in the ass for webdevs.
Communication is always an issue with larger companies. This is one of the awesome benefits of 0th/1st stage start-ups: the right hand always knows what the left hand is doing.
Communication is always an issue with larger companies
No, not always. A lot of the Google Chrome team hangs out in #chromium. If you come in to track down a bug showing genuine intelligence, it's likely a solution will start getting pushed down to user machines in the dev channel by the end of the month.
I've heard passing reference to this bug before, in the form of "you can't attach unlimited window.onload events", but I've never seen anyone run into it in practice.
Glad to see its fixed though. Wish I had a neighbor who worked for MS. ;)
I've had a number of really negative experiences with MS evangelist - I generally see them as a sales tool, especially on the dev-side of things (I can't think of a single good developer that I know who would want to become an evangelist).
Anyways, in my ever-lasting cynicism, this is what I got from the blog post:
-There's a bug in IE
-IE bugs are hard to debug
-There's a bug with MSDN downloads
-It takes "a lot of help" for an insider to figure out how to report an issue and identify the people to report it to
-It took 20+ days, and numerous attempts to get help from Microsoft
-When the IE team says its "tracking it", you shouldn't expect any resolution
(Disclaimer: I'm the "Ravi Rao" mentioned in the post, and I work full time on the IE Team at Microsoft)
That's extremely unkind, if you ask me. We have a support line that puts you in direct contact with Product Support Services people who are best equipped to debug your issues, and I know that every support call generates a ticket # and gets resolved; in cases where the PSS Engineer isn't able to debug/sort out the issue themselves, they get escalated to us, on the Product Team - I've done tons of investigations myself, and we funnel the information back to the caller via the PSS, so they're not dealing with 10 different people but rather, maintain a single point of contact.
In this case, I not only exchanged email with Joel, but also with Ross Boucher of 280 North behind the scenes, and within the same day, narrowed down the bug to our JScript engine. A trouble ticket was generated the same day I learnt of the problem, and that's how we were tracking it. You could get the exact same service by calling our 1-800 support desk albeit the support folks will have to go through their script to make sure the mistake isn't human error at your end, to begin with ... sure, it'd be great if all your issues got resolved within 30 minutes of us hearing about it, but the reality is that it may sometimes take longer than that. You're spinning it absolutely the wrong way if you imply that having "insider connections" is essential to getting a bug report investigated.
Joel and I (and others internal to MS) worked hard to get a startup's problem resolved as soon as we could; I can't believe you're actually complaining about it.
I want to give another thank you to all the folks at Microsoft who helped us track down the problem and provided us with a simple test case that reproduced the problem. They were quite helpful, and we (along with all of our users) appreciate it.
"IE bugs are hard to debug":
Well, this particular one was hard to debug. People work around all sorts of other problems in IE without needing to contact Microsoft.
"It took 20+ days, and numerous attempts to get help from Microsoft":
It was 10 days, and most of that was my fault. I was new to Microsoft at that point, and still learning how to get bugs filed.
"It takes 'a lot of help' for an insider to figure out how to report an issue and identify the people to report it to": When that insider is new, yes :(
- The author or his startup friend didn't google "report bugs to microsoft" and click on the first link http://connect.microsoft.com/ where you can report bugs for programs including IE.
- Dealing with big companies takes more time than dealing with one small teams.
- When the IE team says it's tracking a different issue, and you don't mention it again in your blog, people who hate MS will jump to negative conclusions.
This is only exacerbated by the uselessness of the available debugging tools and Microsoft's general unhelpfulness.