I'm going to attach a patch that reports the failure to load a stylesheet due to a MIME type mismatch to the error console. We've discussed this a few times, and I finally broke down and did it. (Do we have another bug on this somewhere? Oh well.) I'll probably want to move the main added function somewhere else when I work on localizability of CSS parser error messages. However, for now, I'll just put it where it's needed. I probably need to do more magic than I just did to add a new jar.mn. It works on Linux, though. (Does Mac need babysitting?)
Status: NEW → ASSIGNED
Priority: -- → P2
Target Milestone: --- → mozilla1.1beta
Created attachment 89666 [details] [diff] [review] patch, v. 1.1
Attachment #89665 - Attachment is obsolete: true
Comment on attachment 89666 [details] [diff] [review] patch, v. 1.1 >+ NS_ConvertASCIItoUCS2("").get(), /* file name */ >+ NS_LITERAL_STRING("").get(), /* source line */ They're both NS_LITERAL_STRING now in my tree.
bug 113399 is the orig; dup whichever way is more convenient. Do we want to use nsIRequest.name or get the URI from the nsIChannel and GetSpec on that? Also, is nsIRequest.name actually ASCII? Or UTF8?
GetName seems to be equivalent for all the ones that actually are also nsIChannel. You're right that it's probably UTF8, though. I'll fix that...
Created attachment 89667 [details] [diff] [review] patch, v. 1.2
Attachment #89666 - Attachment is obsolete: true
Created attachment 89669 [details] [diff] [review] patch, v. 1.3 Add changes to MozillaBuildList.pm so the Mac will stay happy too.
Attachment #89667 - Attachment is obsolete: true
Created attachment 89670 [details] [diff] [review] patch, v. 1.4 Refactor a bit. Use URI rather than Name. Change status to error, and add warning for case where we do load. (I'd rather use warning and info, but nsIScriptError doesn't have info -- we only seem to do that for generic console messages, perhaps.)
> + * @param aErrorName The name of a string in css.properties. > + * @param aParams The parameters for that string in css.properties. > + * @param aParamsLength The length of aParams. Fix this to match the actual params the method takes and r=bzbarsky
Comment on attachment 89670 [details] [diff] [review] patch, v. 1.4 sr=waterson
Attachment #89670 - Flags: superreview+
Fix checked in 2002-07-01 20:11 PDT.
Status: ASSIGNED → RESOLVED
Last Resolved: 16 years ago
Resolution: --- → FIXED
Madhur, if you're going to plaster "[whitebox]" over bugs, please at least don't do it for bugs that are easily verifiable.
You need to log in before you can comment on or make changes to this bug.