Once again, the client-side code highlighter doesn’t work properly. This question stops my browser from responding completely, until after a few seconds I am asked whether I want to continue the script or stop its execution. I can press “continue” as often as I like, the script never finishes and the browser never responds.

I’m assuming that the script is caught in an infinite loop somewhere since, although the source code in the question is quite long, and it’s in VB (which the formatter still doesn’t support), it should never take that long (on all other codes, highlighting is instantaneous).

The browser is Firefox 3.5.4 (OS X 10.5.8).

(Notice that I am only assuming that the error lies in the formatter rather than somewhere else.)

EDIT Uhm … can’t repro any more, although there’s still a very noticeable delay just at the end of loading the page. According to Firebug, this is due to http://sstatic.net/so/Img/search.png, which is waiting 3.9 seconds (!!!) for a response (but then loads instantaneously). I’ve therefore taken the liberty to change the title of this bug report.

And just to be clear, this problem only appears on the question linked above, nowhere else.

EDIT 2 Apparently the problem isn’t sstatic.com after all, but a script. I don’t know exactly how to repro but now I’ve once again got the timeout dialog (at least most of the time) which tells me that, in fact, the script http://sstatic.net/so/js/master.js?v=5005:1 isn’t responding. Well, that’s probably not all that helpful.

Anyway, I’ve attached links to four images (quite large, so not inline) of the firebug network diagnostics:

  1. After first loading, no script timeout, just noticeable delay
  2. Reload of page, this time during script’s timeout
  3. … and after the script has finally finished loading
  4. After going forward and backwards again in the browser history – notice that no content gets loaded but scripts get executed (hence Google analytics and pixe.… is loaded).

Notice that clearly at some point the whole network traffic is put on hold to wait for some script, and also notice the red line which indicates that loading has completed. To repeat: this only happens on that one question, and it coincides with the time it takes for the code to be highlighted: during the execution of the script, the code appears without syntax highlighting. If I halt execution of the script, the code remains unhighlighted.

  • Works perfectly with FireFox 3.5.4 on Windows. Get a OS ;) – Ladybug Killer Nov 3 '09 at 9:56
  • 1
    I agree that it's a new bug, so you really ought to reopen this as a new bug; contorting the subject so wildly is ... not a good fit. – Jeff Atwood Nov 4 '09 at 12:37
  • Jeff: yes, sorry for that. I wasn’t sure on the best way to do that, since I had run off into the completely wrong direction with my diagnosis. I’ll just make a new posting. – Konrad Rudolph Nov 4 '09 at 17:04

Well, I don't see any problems in the cacti graphs for web3 which is where sstatic.net is hosted on bare metal (no VM).

  • memory
  • cpu usage
  • network bandwidth
  • http connections

all look normal through the last few days.

  • It wasn’t due to sstatic.com after all, see my update to the question. The syntax highlighting really has troubles with this code. – Konrad Rudolph Nov 4 '09 at 12:30

I don't have any figures, but indeed - when I'm using SO on the train (limited bandwidth), the main "waiting for..." that I see is always sstatic. Fine from my main connection(s) though. It is quite a regular thing to see the main text loaded, but the "other bits" taking forever to dribble in.

My first thought was some kind of routing / caching difference between SO and sstatic, but I'm not a network guy.

  • simple explanation: the largest assets are from sstatic.net, including images – Jeff Atwood Nov 4 '09 at 8:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .