312482 - XMLHttpRequest data response arbitrarily truncated

Status: VERIFIED DUPLICATE of bug 293046

(Core :: XML, defect)

Description

[Shawn Neumann]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8b5) Gecko/20051006 Firefox/1.4.1
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8b5) Gecko/20051006 Firefox/1.4.1

I looked for a similar bug, but could not find one.  please let me know if this
is a dupe.

When getting the response from a XMLHttpRequest that is longer than a fwe
hundred chars, Firefox will arbitrarily truncate the list to a smaller size.  I
have tried this on all platforms with FFb2 and it the truncation length varies
from computer to computer.

This is a grave problem for our AJAX enabled web service in that the data is
being corrupted in some form interactions when it is getting incomplete data
sets.  With this present in the browser, we will have to ask people to not
upgrade to ff 1.5 or use other browsers until it is resolved.

Reproducible: Always

Steps to Reproduce:
1.Click on the Link Provided
2.Ajax is called to return a list of 4999 numbers
3.

Actual Results:  
The resulting data is truncated at variable lengths

Expected Results:  
all the ajax data should be displayed ending with the number 4999

We're using clean installs with new profiles and no extensions to test this.

Comment 1

[Eric Murphy]

I got a response length of 28901 with the last number being 4999. This seems to
be your expected results? I am using today's build on Windows 2000. Could this
be a problem with your network?

Comment 2

[Shawn Neumann]

We have tested this on different networks and on a internal network connection
(gig ethernet) with the same results.  The length of the response does vary
based on the computer and network but all out test on 1.5b2 have returned an
incomplete list

I will now test on a nightly build.

Comment 3

[Shawn Neumann]

I tested in Build Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a1)
Gecko/20051014 Firefox/1.6a1 and the problem is resolved.  However in a 1.5
nightly and it is not fixed - Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.8b5)
Gecko/20051014 Firefox/1.4.1

Obviously this needs to be verified, however given that is is, is there chance
to have this fixed in 1.5 prior to release?  The data coruption issue will mean
we will need to ask users to not upgrade until 1.6+ is available.

Comment 4

[Eric Murphy]

Darin, do you know about this? I have searched through Bugzilla and could not
find anything.

Comment 5

[Phil Ringnalda (:philor)]

I'd say from the timing when it started working on the trunk (between 20050813
and 20050820, you could narrow it down more by downloading the builds in between
from http://archive.mozilla.org), I'd say you're either a duplicate of, or also
fixed by, bug 293046, so if you need it, and actually need to both call
overrideMimeType and to return non-XML (if I understand it correctly, that was
the bug), then you need to mark this as a duplicate of bug 293046, and change
the blocking1.8rc1 flag on it to a "?", and add a comment explaining why you
need it, and get the patch author/reviewer/superreviewer to explain that it's
both very strongly needed, and at this point in the release cycle, the safest
thing on the face of the earth, which couldn't possibly make anything, anywhere
fail in any way.

Good luck ;)

Comment 6

[Darin Fisher]

I can reproduce the bug here as well.  Investigating...

Comment 7

[Christian :Biesinger (don't email me, ping me on IRC)]

fixing component, this is certainly not FE code... guessing at XML for now

Comment 8

[Shawn Neumann]

Given the impact this bug will have on the data integrity of our AJAX
interfaces, is there anything I can do in this process to encourage as
resolution prior to 1.5 final?

Comment 9

[Asa Dotzler [:asa]]

Shawn, can you verify that this is working on the trunk for you (fixed by the
checkin at bug 293046).  If so, we can investigate taking that change (or part
of it) on the branch. Thanks.

Comment 10

[Shawn Neumann]

Sure Asa. Yes, it is fixed in the latest nightly:

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a1) Gecko/20051017 Firefox/1.6a1

I just tried that now and the result is as should be expected. 

Comment 11

[Asa Dotzler [:asa]]

Ok, duping this against that. we'll nominate bug 293046 for evaluation for
inclusion in the 1.5 release.

*** This bug has been marked as a duplicate of 293046 ***