[acer] XML error after update
Categories
(Toolkit :: Startup and Profile System, defect)
Tracking
()
People
(Reporter: doug, Unassigned)
References
(Depends on 1 open bug)
Details
Attachments
(3 files)
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/90.0.4430.85 Safari/537.36 Edg/90.0.818.49
Steps to reproduce:
I allowed Firefox to update my Aspire V3-551-8887 yesterday. I got an error screen on restart. I started Bing (since i couldn't use Firefox) to go to https://www.mozilla.org/en-US/firefox/new/ about 07:30 EDT 4/30/2021 and clicked "Download Firefox". This downloaded "Firefox Installer.exe" of 333,048 bytes (Size on disk 335,872 bytes).
Actual results:
On starting Firefox gives a grey window with the text:
XML Parsing Error: undefined entity
Location: chrome://browser/content/browser.xhtml
Line Number 743, Column 15:
<toolbarbutton id="UITourTooltipClose" class="close-icon"
--------------^
[Last 2 lines are in red.
Expected results:
A Firefox screen should have come up. I have FF remember my windows & tabs, so the windows & tabs i had open before the update should have appeared.
I am running Microsoft Windows version 10.0.18363 Build 18363. I cannot get past this point to run Firefox.
Comment 2•4 years ago
|
||
The Bugbug bot thinks this bug should belong to the 'Firefox::Installer' component, and is moving the bug to that component. Please revert this change in case you think the bot is wrong.
Updated•4 years ago
|
Updated•4 years ago
|
Comment 3•4 years ago
|
||
Can you try creating a new profile following these instructions and see if that works?
https://support.mozilla.org/en-US/kb/profile-manager-create-remove-switch-firefox-profiles#w_starting-the-profile-manager
Do you know which version of Firefox you were using before the update?
Comment 4•4 years ago
|
||
Also, what language are you using Firefox in?
Comment 5•4 years ago
|
||
If it is remotely possible, would you be willing to zip up your profile and send it to us privately? We've been trying to figure out what is causing this problem and we're having trouble reproducing.
It's the C:\Users\USERNAME\AppData\Roaming\Mozilla
directory.
Comment 7•4 years ago
|
||
Would you be willing to help us debug this if the old profile still fails?
We've been trying to get a recreation scenario for this for a while, so it would be really helpful if we could have you turn on some logging and check things.
This appears to be the problematic file, being the only one with the string "UITourTooltipClose"
The file is:
C:\Users\Doug\AppData\Roaming\Mozilla\Firefox\Profiles\ulxgcc24.default-release\datareporting\archived\2021-04\1619782256664.e2d19360-ea83-4ef8-9ec0-d6d2253d100e.event.jsonlz4
I would like to be able to get back my old profile. Can you tell me how?
Comment 9•4 years ago
|
||
Sure.
From the new profile, you can type about:profiles and you can select your old profile.
If your old profile is still failing, we'd love to do some logging to figure out why. To do that, follow the instructions here:
https://firefox-source-docs.mozilla.org/networking/http/logging.html#windows
and add nsJarProtocol:5 to the module list so:
export MOZ_LOG=timestamp,rotate:200,nsHttp:5,cache2:5,nsSocketTransport:5,nsHostResolver:5,cookie:5,nsJarProtocol:5
this should give us log files.
You'll need to start Firefox from a command line to do it.
If you want me to walk you through it, I'm more than happy to do that via Zoom or something.
Reporter | ||
Comment 10•4 years ago
|
||
I did this. The old profile is still failing. Where do i find the log files?
Reporter | ||
Comment 11•4 years ago
|
||
The zipped profile is 103 MB. I am concerned that if i send it to you, i reveal passwords. I hope i have kept credit card numbers and bank account numbers out of my profile, but there's probably my IRA account #. Are there certain files you want?
Comment 12•4 years ago
|
||
The log files will be in your temp directory.
so if you open a command prompt, then do:
cd %TEMP%
all the files will begin with log.txt so they will look like this:
log.txt.child-1.moz_log.0
log.txt.child-2.moz_log.0
log.txt.child-3.moz_log.0
log.txt.child-4.moz_log.0
log.txt.child-5.moz_log.0
log.txt.child-6.moz_log.0
log.txt.child-8.moz_log.0
log.txt.moz_log.0
You can zip up all those files and send them to me mkaply at mozilla.com
And yeah, understand how you feel about the profile. Let's start with the logs and see if we can find the problem there.
Comment 13•3 years ago
|
||
Setting the need info request to the reporter to remind about the request above.
Comment 14•3 years ago
|
||
i mailed the logs to Mike at 12th May because we have the same issue for some users...
Comment 15•3 years ago
|
||
Here's the log that was just mentioned.
Updated•3 years ago
|
Comment 16•3 years ago
|
||
@Mike: i wasnt sure if i should upload it myself, you did it, fine.
Comment 17•3 years ago
|
||
Mike, can we consider this issue confirmed?
Also, can you provide it with a component that works for it?
Finally, please NI me if there is anything I could help with on the testing side.
Comment 18•3 years ago
|
||
Yes, I've confirmed it. Unfortunately the team doesn't have a lot of answers.
Comment 19•3 years ago
|
||
@Mike: any news here?
Comment 20•3 years ago
|
||
Unfortunately not.
zibi: are any of these logs helpful?
Comment 21•3 years ago
|
||
Not really. All the log tell us is that DTD parser got an empty string from Necko.
And from what I know Necko assumes that a customer of it should be responsible for not letting the document be constructed if the network error happens. If I'm correct it's an architectural disagreement on what API design leads to highest quality product and we're discussing it in context of bug 1703346.
@kershaw, is that accurate?
Comment 22•3 years ago
|
||
(In reply to Zibi Braniecki [:zbraniecki][:gandalf] from comment #21)
Not really. All the log tell us is that DTD parser got an empty string from Necko.
The attached log seems not complete and contains no information regarding how the xhtml files are loaded.
Is it possible to upload a full log?
And from what I know Necko assumes that a customer of it should be responsible for not letting the document be constructed if the network error happens. If I'm correct it's an architectural disagreement on what API design leads to highest quality product and we're discussing it in context of bug 1703346.
@kershaw, is that accurate?
I can only tell that necko follows the interfaces defined in nsIRequestObserver.idl and nsIStreamListener.idl to deliver data to the consumers of necko. Maybe this API design is out of date, but it requires huge efforts to change it. Given that this design is already existed for more than two decades.
Updated•3 years ago
|
Comment 23•2 years ago
|
||
A needinfo is requested from the reporter, however, the reporter is inactive on Bugzilla. Closing the bug as incomplete.
For more information, please visit auto_nag documentation.
Comment 24•2 years ago
|
||
Kershaw, should we close this? Is there anything we can do?
Comment 25•2 years ago
|
||
(In reply to Mike Kaply [:mkaply] from comment #24)
Kershaw, should we close this? Is there anything we can do?
Let's leave this open. I think :jesup will take care of ysod issues.
Updated•2 years ago
|
Description
•