Blog homepage

Content from old blog

www.marcustucker.com

Bookmark this page



previous month  JULY 2010  next month
s m t w t f s
1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31


BLOG ARCHIVE
 
RSS ATOM  Full archive
 
current month



CATEGORIES
 
alert [ 1 ]  RSS ATOM
blog [ 8 ]  RSS ATOM
development [ 9 ]  RSS ATOM
    ASP / scripting  [ 13RSS ATOM
        snippets  [ 1RSS ATOM
    database [ 0 ] RSS ATOM
    web standards  [ 1RSS ATOM
    XML [ 0 ] RSS ATOM
general [ 24 ]  RSS ATOM
    humour [ 0 ] RSS ATOM
personal [ 6 ]  RSS ATOM
photography [ 1 ]  RSS ATOM
politics [ 1 ]  RSS ATOM
rant [ 8 ]  RSS ATOM
reading [ 18 ]  RSS ATOM


SEARCH
 


BLOGROLL
 
 


ARTICLES
 
Access Stored Queries

VBScript String Concatenation
 


23 Feb 2005
Printable version  |  Email to a friend
IE's "friendly" error messages
While hacking URLs at www.sonyericsson.com to make it possible to be automatically informed when a new firmware is released for my P900 by using the excellent free service WatchThatPage to monitor the URL that I had hacked together (no thanks to the abominally unfriendly URL scheme used by their CMS), I came across the following message hidden in a comment in a redirector page:
Unfortunately, Microsoft has added a clever new "feature" to Internet Explorer. If the text in an error's message is "too small", specifically less than 512 bytes, Internet Explorer returns its own error message. Yes, you can turn that off, but *surprise* it's pretty tricky to find buried as a switch called "smart error messages" That means, of course, that many of Resin's error messages are censored by default. And, of course, you'll be shocked to learn that IIS always returns error messages that are long enough to make Internet Explorer happy. The workaround is pretty simple: pad the error message with a big comment to push it over the five hundred and twelve byte minimum. Of course, that's exactly what you're reading right now.
NOTE: "Resin" is presumably the name of their CMS "Resin" is a JSP/servlet load balancing system... (thanks Rick, should have Google'd!)

So, it would appear that 512 is the magic number of bytes required to force IE to display your error message instead of your own. Personally I think it's is damn stupid of IE to attempt to intercept an error and display its own error message instead of the server's, and such a feature certainly shouldn't be on by default. It's impossible to fathom the total loss of productivity that this single implementation "feature" has caused to puzzled developers across the world, but here's proof (if proof be needed) that it is a recurrent problem.

It's a matter of habit that I always switch off "Show friendly HTTP error messages" (the correct name for the switch mentioned in the above text), and "Show friendly URLs" in the preferences section of Internet Explorer, since I find them irritating and they often interfere with debugging web apps during development. But clearly one should take precautions (as described in the quoted text above) to ensure that IE doesn't interfere with error messages displayed to end users.

Another thumbs-down to MS for that not-so-bright idea. [:(]

Incidentally, I also switch on "Display a notification about every script error", which is one of the main reasons why there's so much bad JavaScript floating around - developers who don't switch this on aren't told about errors in their code. Naturally, this also causes huge headaches during development for those who aren't aware of this.
 
development , rant
posted by  Marcus at  12:28 | comments [3] | trackbacks [69]


BLOG COMMENT


posted by   rick hurst  [ http://hypothecate.co.uk ] 23 Feb 2005 at 13:43
Resin is in fact a JSP/ Servlet engine :-)

http://www.caucho.com/resin/
 




posted by   Nick Fitzsimons  [ http://www.nickfitz.co.uk/ ] 5 Apr 2005 at 13:38
"Display a notification about every script error", .. is one of the main reasons why there's so much bad JavaScript floating around


I've even encountered web app code which seemed to work OK until the error notification was enabled. I never had time to work out what was going on (it was quicker to rewrite from scratch than try to unravel the spaghetti), but it seemed to be something to do with an error in an iframe causing the calling code to terminate if the error notification was enabled, but allowing the calling code to continue otherwise.

When I did have time to investigate, judicious googling allowed me to identify client and ASP code from four separate tutorial or open source sites which had been cobbled together to form a CMS. Naturally the developer hadn't told his emplyer where the code came from; they thought he must be brilliant, getting something like that working.. then I enabled script error notification on their machines :-)
 




posted by   James 7 Aug 2006 at 19:55
I saw that you have a page that discusses patent-related resources at http://blog.marcustucker.com/. I wanted to suggest adding www.freepatentsonline.com to the page. This web site has free PDF downloading (instead of having to page through TIFFs like at the US PTO). It is by far the best free patent searching site.
 




ADD COMMENT
(*) indicates required fields
author (*) :
email address :
url :
 
  bold italic underline add hyperlink add email hyperlink centre unorder list order list add image quote emoticon smiles
 
comment (*) :

html code

ubb code

max characters : 1000

+
- +
-
characters remaining :
remember me :