218908 - Very bad performance when using document.load or XMLHttpRequest with compressed (gzip) xml stream

Status: RESOLVED FIXED

(Core :: XML, defect)

Description

[Lionel MUSY]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.5b) Gecko/20030827
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.5b) Gecko/20030827

When I load compressed xml file (content type : gzip), it takes a very long time
for Mozilla to treat this xml file. It seems that at this step, Mozilla runs
slowly and sometimes it can even stop. Once the file is loaded and when Mozilla
is not stopped, the process runs correctly.

When I load the same xml file but not compressed it works fine. In my case the
time factor is approximately 20. While it takes 1 minute to load and parse
uncompressed xml stream, it takes more than 20 minutes to load and parse gzipped
xml stream.

Sample :
 var domDocument = document.implementation.createDocument("", "", null);
 domDocument.async = false;
 domDocument.load("http://server.com/servlet");


Reproducible: Always

Steps to Reproduce:

Comment 1

[David Pérez]

The same happens to me.
For example if I try to compress external CSS files or JavaScript files, Mozilla
gets the result correctly, but very very slow.
On the other hand, IE works ok.

Operating system: XP
Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.5) Gecko/20030925
MultiZilla/1.6.0.0e

Comment 2

[Jeffrey Baker]

Confirmed on WinXP firefox PR1.

Comment 3

[matiyam]

I have reported a new bug about the XMLHttpRequest object taking 100% of CPU
when waiting to fetch a page with async=false.

I suppose that is what is happening to you:
- You make a request with XMLHttpRequestObject and async=false
- Mozilla takes 100% of your CPU
- Because of that, your CPU takes longer to decompress your file
- Because of that, your file takes longer to be fetched and Mozilla keeps using
100% of your CPU.
- And so on...

Comment 4

[matiyam]

Ups! Sorry, i forgot to link to the other bug: Bug 273578 

Comment 5

[Darin Fisher]

This bug may go away or at least be significantly reduced in severity once bug 326273 is fixed.

Comment 6

[Wayne Mery (:wsmwk)]

do you see the problem in :

* FF 2.0 ?    ...  Bug 273578 was fixed in FF 1.5
* FF trunk* ? ...  bug 326723 is fixed on trunk builds

*http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/latest-trunk/ 

Comment 7

[Ben Bucksch (:BenB)]

Lionel Musy or David Pérez, do you still see the bug (last 2 comments)?

Comment 8

[Wayne Mery (:wsmwk)]

I emailed them both direct on 11/28 but no replies (read receipt from Musy on 12/10)

"should" be FIXED per comment 6 bugs
please reopen if you see otherwise.