219679 - Mozilla parses out the img src tags and makes bad HTTP requests for these

Status: RESOLVED DUPLICATE of bug 249508

(Core Graveyard :: File Handling, defect)

Description

[Paul Gregg]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030529
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.5b) Gecko/20030827

I have a customer 404 page at the above site which catches bad requests.  There
is nothing wrong with the page itself, but Mozilla is sent e.g. <img
src="/images/projects_oval_hi.gif"> and generates a request for:
/projects/radiusreport/index.php_files/projects_oval_hi.gif

This is plainly wrong - I've no idea if it manages to correct it on the client -
but this bug fills up server error_logs unnecessarily.
What I see on the server side is:
Remote Addr: 203.151.8.41
Referrer:
http://www.pgregg.com/projects/radiusreport/index.php?cookiesarenotok=1&q=001010171191LxhNPstCSfJopFF1Vpv3bBbwTEZj9nbUUtIKVcnOr6cQLRSHgoLs
User Agent: Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.5b) Gecko/20030827
File does not exist:
/web/www.pgregg.com/source/projects/radiusreport/index.php_files/projects_oval_hi.gif

Btw - I'm the server not the client - so can't reproduce it since I don't
haven't upgraded from Moz 1.4




Reproducible: Didn't try

Steps to Reproduce:
1.
2.
3.



Expected Results:  
Moz should have made a HTTP request for
http://www.pgregg.com/images/projects_oval_hi.gif


In the page source you'll see (for that particular image) javascript to change
it for a high/low version of the same image - but the same errors occur for
non-javascript driven images (just to dispel that as a possible source)

Comment 1

[Mats Palmgren (inactive)]

WFM, 2003-09-18-05 trunk Linux

Comment 2

[Olivier Cahagne]

this comes from someone who saved your page locally (index.php_files) and has
nothing to do with Mozilla.

Comment 3

[Bill Mason]

Per previous comment.

Comment 4

[Paul Gregg]

Ok, if it is as a result of locally saved pages (I presume via a Moz function) 
then Moz should be able to cleanly view such saved files without generating 
such bogus http requests ?

No matter what - this still highlights a bug in the browser.

Comment 5

[Paul Gregg]

Ok, I've found I can reproduce this:
1. Go to: http://www.pgregg.com/projects/radiusreport/index.php?cookienotok=1
2. Select File / Save Page As...
3. On the popup of where to save (with me its the desktop), click ok.
4. a file and a folder are created on the desktop called index.php_files and 
index.php.html  at the same time Mozilla makes a HTTP request for:
http://www.pgregg.com/projects/radiusreport/index.php_files/paulgregg_em.gif

This undoubtedly is a bug.

Hope this helps.

Comment 6

[Boris Zbarsky [:bzbarsky]]

Over to file handling; this is a persistence object problem...

Comment 7

[Boris Zbarsky [:bzbarsky]]

To Adam.

Comment 8

[OstGote!]

Clearly a dupe ... The bug is fixed now on trunk (Mozilla 1.8a6/1.8b and Firefox
trunk builds) since 01/10/2005. Branch builds (like Firefox 1.0 and Mozilla
1.7.x) still have this bug.

*** This bug has been marked as a duplicate of 249508 ***

Comment 9

[Paul Gregg]

(In reply to OstGote! from comment #8)
> Clearly a dupe ... The bug is fixed now on trunk (Mozilla 1.8a6/1.8b and
> Firefox
> trunk builds) since 01/10/2005. Branch builds (like Firefox 1.0 and Mozilla
> 1.7.x) still have this bug.
> 
> *** This bug has been marked as a duplicate of 249508 ***

Just pointing out that, yes, 249508 is Clearly a dupe of this - since I filed this bug 10 months before that one.  It's hardly my fault that you lot ignored it so long.