221795 - Startup crash + error msg "INTERNAL ERROR on Browser End: Expected a version > 5! Version = 0"

Status: RESOLVED DUPLICATE of bug 209532

(Firefox :: General, defect)

Description

[Mickey Stein]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6a) Gecko/20031003 Firebird/0.7+ (Optimized for:Athlon XP-SSE by Yek)
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6a) Gecko/20031010 Firebird/0.7+ 

This is the exact message I got after pulling cvs at ~1500gmt 10/10/2003 and
building a vanilla (plain) firebird, running with the command ./MozillaFirebird

"INTERNAL ERROR on Browser End: Expected a version > 5! Version = 0

System error?:: File exists
"

I installed into a new directory and deleted the old profile. All builds up to
today have at least compiled and run, but the last 7 days have exibited some
sort of CVS "malaise". Today was a total breakdown. 



Reproducible: Always

Steps to Reproduce:
1. Pull cvs as of 1500gmt (maybe reproduce by using a -D '2003-10-10 15:00 GMT"
tag on the co.
2. build using this .mozconfig.

export MOZ_PHOENIX=1
mk_add_options MOZ_PHOENIX=1
ac_add_options --enable-crypto
ac_add_options --enable-default-toolkit=gtk2
ac_add_options --disable-tests
ac_add_options --disable-debug
ac_add_options --disable-mailnews
ac_add_options --disable-composer
ac_add_options --disable-ldap
ac_add_options --enable-optimize="-O3 -march=athlon-xp -msse"
ac_add_options --enable-extensions=cookie,xml-rpc,xmlextras,p3p,pref,transformii
x,universalchardet,typeaheadfind,webservices
ac_add_options --enable-jsd
ac_add_options --enable-xft
ac_add_options --disable-profilesharing


3. install into clear directory, move old profile, run using ./MozillaFirebird.

Actual Results:  
"INTERNAL ERROR on Browser End: Expected a version > 5! Version = 0

System error?:: File exists
"
is what happened and that's all that happened. 

Expected Results:  
run.

Unfortunately I'm on the way out of town in a few hours so don't have time for a
debug version of this source. That's all the info I can give today.

Comment 1

[Mike Connor [:mconnor]]

reporter, how is the added string on the UA being added for your build?  is this
using the vendorComment pref or something else?  The java plugin is looking for
a version, but is failing to get it.

Barring that, what java plugin are you using (suspect 1.4.2) and was this
working for the previous day's build?  What about if you pull a completely clean
tree and rebuild?

Comment 2

[Mickey Stein]

The added string (Optimized for .. ) came from the build I reported it
on(03-Oct), not the build that crashed (today's). I put it in
~./browser/**all.js as vendorComment for the 03 build. 

The java plugin is Java(TM) Plug-in 1.4.2_01 from Sun and ln -s'd to the correct
spot. That was a clean pull as of ~1500gmt or so. I moved my old tree aside,
moved .phoenix to .phoenix.sav  and did a cvs co. 

One thing that's possibly different about my environment (RH linux V9) is that
I've been running the 2.6.0-test6-mm4 kernel for a week, but no other builds
have had this particular problem. Perhaps it was just a bad moment on the CVS tree? 

Comment 3

[Mickey Stein]

Forgot one thing in my response: Yes, I've already kicked off a cvs co for
another try at this. 

Comment 4

[Mike Connor [:mconnor]]

Regression caused by the checkin for bug 209532, since that bug is reopened as
has a fix to patch the bustage, duping to that bug.

*** This bug has been marked as a duplicate of 209532 ***