291455 - xmlhttp.responseText is cut off with application/rdf+xml

Status: RESOLVED WORKSFORME

(Core :: DOM: Core & HTML, defect)

Description

[Jim Ley]

User-Agent:       Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; .NET CLR 1.1.4322)
Build Identifier: 

When retrieving a URI with mime type application/rdf+xml xmlhttp.responseText 
is the incorrect length, appears to be cut off at random points, I mostly see 
4096 bytes but have seen less.

This has also been tested on platforms other than Windows XP

Reproducible: Always

Steps to Reproduce:
1. Visit http://jibbering.com/2004/5/mozillaBug.html
2. See the 2 lengths, they should be the same.


Actual Results:  
The lengths were different 6054 for the xml resource 4096 (generally) for the 
application/rdf+xml resource.

Expected Results:  
Both should return 6054

This test page sometimes crashes, especially on reload, I believe that is 
related to bug #289156 though.

Comment 1

[Xanthor Sccas]

WorksforMe with Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8b2) Gecko/20050405

Both return 6054, even reloading many times..

Comment 2

[Boris Zbarsky [:bzbarsky]]

Worksforme with a current trunk build as well.  Marking so; Jim, please reopen
if this is reproducible for you in a current build.

Comment 3

[Jim Ley]

Yes, current nightly doesn't exhibit the behaviour any more for me either, so 
looks like it's been fixed as part of another issue.

Cheers!

Comment 4

[Joey Guerra]

I'm getting an error message similar to this one but in a Mac OS X environment.

Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.8) Gecko/20050511
Firefox/1.0.4

Error: [Exception... "Component returned failure code: 0x80040111
(NS_ERROR_NOT_AVAILABLE) [nsIXMLHttpRequest.status]"  nsresult: "0x80040111
(NS_ERROR_NOT_AVAILABLE)"  location: "JS frame :: http://localhost/ajax.js ::
AjaxRequestCallBackFunction :: line 48"  data: no]

How do I reopen this ticket?

Comment 5

[Boris Zbarsky [:bzbarsky]]

You don't reopen until you've tested with a current trunk build (say the Deer
Park alpha).  You're testing with a build that has a Gecko implementation from
April 2004, when this bug did exist.  Something has changed since then to fix it.

Comment 6

[Joey Guerra]

But it looks like the build date is 05/11/2005?

Comment 7

[Boris Zbarsky [:bzbarsky]]

Sure.  The source is from April 2004, but some security fixes were made to it,
and the slightly modified source with those security fixes was compiled on May
11 2005.  The build date is the date when the source was compiled; it has
nothing to do with when the source was actually checked out from CVS, for
instance...