306984 - string "url(paper.png) scroll;" in page from W3C shows broken-image instead of typed text, works in IE, Netscape 8

Status: RESOLVED WORKSFORME

(Firefox :: General, defect)

Description

[rfb]

User-Agent:       Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.0.3705; .NET CLR 1.1.4322; Media Center PC 3.1)
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.10) Gecko/20050716 Firefox/1.0.6

I updated to Firefox 1.0.6, and was looking up some CSS specs at 
http://www.w3.org/TR/CSS21/colors.html#propdef-background-attachment, and I 
found where the source is:
------------------
<pre>
body {
  background: white url(paper.png) scroll; /* for all UAs */
  background: white url(ledger.png) fixed; /* for UAs that do fixed 
backgrounds */
}
</pre>
---------------- it displays in Firefox as:--------------------
body {
  [broken-image-symbol]) scroll; /* for all UAs */
  [broken-image-symbol]) fixed; /* for UAs that do fixed backgrounds */
}
-----------------
I tested it under IE 6.x, Netscape 8.0.3.3 and they work correctly.

apparently the statement is being interpreted as a style sheet instead of a 
text string to display, and since there is no local file of that name, it 
displays a broken-image or broken-link instead.  I've noticed it before, but 
never chased it down to see what it was, so it was probably doing it back in 
Firefox 1.0.4

Reproducible: Always

Steps to Reproduce:
1. Start Firefox.
2. type: http://www.w3.org/TR/CSS21/colors.html#propdef-background-attachment 
into location input area
3. type return. or press GO button

Actual Results:  
I see broken image or broken link images in place of the characters prior to 
the ')' in a 'background: <color> url("someurl") scroll; statement.

Expected Results:  
It should have simply displayed the text:

..... with the keyword 'fixed'. For example:

body {
  background: white url(paper.png) scroll; /* for all UAs */
  background: white url(ledger.png) fixed; /* for UAs that do fixed 
backgrounds */
}

See the section on .....


I have created a PNG image of trimmed screenshots from Firefox, IE, Netscape 
and the "show source" window of that URL.  You can download it from:
http://www.baptistanet.com/ff_tb/img/busted-image.png
It is pretty large so it will probably need to be displayed in Psp or some 
other graphix tool.  The image will remain at that location until you inform 
me that it is no longer needed there.

There needs to be another level for severity between minor and trivial.
It isn't something that is major, it's trivial but there isn't a workaround, 
and should be fixed if you find the problem is in the code and not just 
something that I am seeing locally here.

I only had one other Firefox install that I could compare against, and that is 
a 1.0rc2 on a Win2K platform, and that one works correctly.  I did search for 
an existing entry but didn't find any matches, so that might indicate it is 
just my install that is bad.

Comment 1

[Adam Guthrie]

WFM - Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a1) Gecko/20050902
Firefox/1.6a1 ID:2005090216

Comment 2

[Peter van der Woude [:Peter6]]

Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b4) Gecko/20050904
Firefox/1.0+ ID:2005090400

WFM

Comment 3

[rfb]

Peter, your agent id is 1.0+, and your date is today, are you using a nightly 
build of 1.0.7 beta?  Also, I changed my Netscape 8 to behave like Firefox (it 
was set to IE) and then it failed, too.  Previously I had gotten myself in 
trouble following nightlies (the last time I had to blow away my profile), so 
I'm not sure I want to try it until it is released.  I start a new job on Tues 
and I don't have the time to properly reinstall a fresh system to see if the 
problem goes away.  It could be an extension that's at fault, too.
rfb

Comment 4

[Adam Guthrie]

Please reopen if you can reproduce this in the latest builds with a clean profile.

Comment 5

[rfb]

Well, that "resolution" pretty much stinks. Every time I put in a possible bug
to be checked out, they immediately get closed by someone's WFM ****. Why bother
reporting possible problems if nothing ever gets investigated?  Did someone say
send me your <whatever> file so I can see if its corrupted. Saying "prove it"
isn't resolving it, from picture you can see that it is there, not a figment of
my imagination.  The idea of a "dirty" profile also is ridiculous.  How can it
get "dirty"?  What is dirty? A bad file pointer? A file that Firefox wrote
badly? What is it that I need to do to keep it "clean"?  Install SOAP? HA!  I
have 200 passwords, 52 extensions, specialized Tab software that I patched in
with my theme file, userChrome lines, auto fill, auto logins, weather,
customized buttons, external app buttons.... it literally will take 20 hrs of
pasting and merging and updating version numbers to get back where I started
from.  I even have to put "try { } catch() {}" around every "var p=" line in the
browser.jar:content\global\bindings\tabbrowser.xml file to keep it from filling
my javascript console with errors.

I'm not complaining about Firefox. As I said in the report, this is a ****-ant
problem.  I think the tool is great.  The premise of having the extension
capability, adding UI themes, etc. I have 5 horizontil bars and 2 vertical ones
filled with icons.
Before throwing around "WFM, reopen after clean install", the team should
develop a backup utility that extracts and saves all the installation updates I
put in, backs-up extensions, so a clean install just means running the
installer, then running a re-install-specializations function, followed by maybe
an hour of recustomizing by hand.  Or better yet, an installer that recognizes a
previous version and validates (un-dirties) control files as well as flagging
files/jars that were modified that may need special integration.
Perhaps there needs to be some efforts in finding out what **** up the profile
so that there is no need to do clean installs every month.  With a 4 hour
commute and 8 hr workday, who has the time to reinstall all the time.

Comment 6

[u88484]

http://kb.mozillazine.org/Profile_backup