USA Today loading fix – Vivaldi Browser snapshot 2009.3
-
@Ayespy Thanks, I'll give it a try. How can the Login Data file become corrupt? So far, no such case has occurred.
-
@npro Thank you for testing & advising me.
Fwiw, here's an extract from my bug report of last night, from which you can see the extent of the problem.
The latest Snapshot 3.3.2009.3-1 has a 100% failure rate [fails to launch at all] in my various Linux installations, real & VM, Arch, Arch-based, Debian-based. The failure is identical across distributions [Arch, EndeavourOS, ArchLabs, Anarchy, SolydK, SparkyLinux] & desktop-environments [KDE, Cinnamon & MATE]. The failure is identical regardless of whether the Linux system had an existing Snapshot that was upgraded, or was a new/virgin installation clean-installed specifically for this Snapshot 3.3.2009.3-1.
When this Snapshot is launched from the system application menu, nothing happens. When it is launched from the terminal, the resultant error message isTrace/breakpoint trap (core dumped)
.Hence, with respect, tbh i can't see how this might be a local [non-Vivaldi] package fault at my end, given the sheer number of distros & DE's with which this new Snapshot achieves a 100% fails-to-launch result [across two separate physical PCs].
That said, here's my comparative Manjaro VM results this morning, after first fully updating & rebooting them.
[steffie@ManjaroStableKDEVM ~]$ vivaldi-snapshot Trace/breakpoint trap (core dumped) [steffie@ManjaroStableKDEVM ~]$
[Manjaro-UnstableVM ~]$ vivaldi-snapshot Trace/breakpoint trap (core dumped) [Manjaro-UnstableVM ~]$
Ergo, it remains that fabled ex-parrot pushing up the daisies, having joined the choir invisible... for me. Still a 100% fail rate.
-
@Gwen-Dragon
bug report VB-70703 -
WoW!
This update acted for me like a fork bomb!
I just decided to download small file, 11 megabytes in size.
Like usual, at the end of downloading, browser did not showed that download is finished.
Some time is usually needed for that (maybe because of antivirus, who knows).
I decided to let it running for a minute, but there was no reaction even after a minute of waiting.
Then strange things starts to happen: mouse movements starts to lag, no programs can be opened.
When, after ~20 minutes, system became responsive, I have opened task manager.
And here is what I saw:
Within several seconds, Vivaldi created 200+ process instances, rising from ~80 to ~280.
Next my decision was to close browser window to kill all those processes: it worked: after another ~5 minutes, processes disappeared.
Of course, that OS hangs was because of excessive RAM usage and RAM contents starting to go into swap file.
But WHY? Why so many RAM was needed to download such small file and why such enormous amount of processes was created?
This is how system-wise RAM consumption chart looks like for this event:
File download was started at 10:20. Only at ~11:00 I was able to gain control over the system again. -
Good grief that is scary.
Is it reproducible with differing small download files.?.Maybe just that 1 download but is worrying none the less. -
@Gwen-Dragon said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
I do not see such crash on Ubuntu 20.04 LTS
I do not have Ubuntu, but fwiw, in addition to all the many Arch & Arch-based distros i've tested with this bad Snapshot [100% fail-to-launch], the same bad result happens in two of the Debian-based distros i've tried [SolydK & SparkyLinux; also 100% fail-to-launch]. Here's the latest just now [was a brand new installation of Snapshot into this existing VM]:
SparkyMATE-VM[~] 19:01:22 Sun Aug 16 $> vivaldi-snapshot Trace/breakpoint trap
-
@Steffie Snapshot works on Fedora 32. It would be interesting to figure out what all your failing installs have in common, considering kernel version, hardware, setup and whatnot.
-
@Priest72 said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
Is it reproducible with differing small download files.?.Maybe just that 1 download but is worrying none the less.
I think that such problem happened because of combination of different factors.
Browser was running for a long time, there was ~15 tabs opened (most likely, I was opening them during 10:15 .. 10:20 period), it was the first time I downloaded this file (antivirus saw it first time), maybe some other factors too.
When I try to reproduce it again with fresh browser process and 2 tabs open, no glitch happens (except for ~10 seconds delay before download finish message appears - I expect that it may be a sign of a problem). -
Hi @Gwen-Dragon,
Just tested in my Xubuntu 20.04 install and see same error.glen@XuFocal:~$ inxi -S System: Host: XuFocal Kernel: 5.4.0-42-generic x86_64 bits: 64 Desktop: Xfce 4.14.2 Distro: Ubuntu 20.04.1 LTS (Focal Fossa) glen@XuFocal:~$ vivaldi-snapshot Trace/breakpoint trap (core dumped)
-
Here is how download on my PC looks like:
https://www.youtube.com/watch?v=nQu9TMk_d54
6s: download starts.
12s: download is finished.
13s..25s: Vivaldi is doing.. what?
(I have stopped my antivirus, so it should not interfere) -
@luetage said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
all your failing installs
Iirc from other posts, i am not the only Nixer having failure-to-launch outcomes with this Snapshot. I wonder at what point the focus stops being on assumed local errors at the individual user level, vs accepting that this is indeed a bad update... allegedly.
Different users, different distros, different DEs, different PCs here ... amazing coincidence if all these are individually bad Nixes.
-
if it is somehow relevant,I downloaded the 32bit debian file which installs as a petfile in my puppy distro and i noticed i had to click 3 times to get vivaldi to launch.this is the 3.2 stable version.
Once vivaldi has launched and i have performed configs etc then it launches normally. -
@luetage said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
Snapshot works on Fedora 32
...and yet...
[steffie@localhost ~]$ cat /etc/fedora-release Fedora release 32 (Thirty Two) [steffie@localhost ~]$ cat /etc/os-release NAME=Fedora VERSION="32 (Workstation Edition)" ID=fedora VERSION_ID=32 VERSION_CODENAME="" PLATFORM_ID="platform:f32" PRETTY_NAME="Fedora 32 (Workstation Edition)" ANSI_COLOR="0;34" LOGO=fedora-logo-icon CPE_NAME="cpe:/o:fedoraproject:fedora:32" HOME_URL="https://fedoraproject.org/" DOCUMENTATION_URL="https://docs.fedoraproject.org/en-US/fedora/f32/system-administrators-guide/" SUPPORT_URL="https://fedoraproject.org/wiki/Communicating_and_getting_help" BUG_REPORT_URL="https://bugzilla.redhat.com/" REDHAT_BUGZILLA_PRODUCT="Fedora" REDHAT_BUGZILLA_PRODUCT_VERSION=32 REDHAT_SUPPORT_PRODUCT="Fedora" REDHAT_SUPPORT_PRODUCT_VERSION=32 PRIVACY_POLICY_URL="https://fedoraproject.org/wiki/Legal:PrivacyPolicy" VARIANT="Workstation Edition" VARIANT_ID=workstation [steffie@localhost ~]$ [steffie@localhost ~]$ vivaldi-snapshot Trace/breakpoint trap (core dumped) [steffie@localhost ~]$
Maybe my friend's joke was actually not a joke -- this Snapshot needs the
Southern Hemisphere Patch
installed before it can work, down here. Was the package compiled this time without including the Coriolis hook? -
@Steffie Yet again, no problem for me (Linux Mint)!?
As I get the updates from Vivaldi repo I just had a (admittedly weird) thought - are all your attempts using herecura?
-
@vort: Create another video with opened Task Manager to see this bug "memory leak" in action. Try to do the same on another system to compare results.
-
@ugly said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
It really only seems like the Vivaldi UI has the graphical corruption. I don't see any issues on webpages.
Scratch that, I noticed the graphical corruption on some website content. I've noticed it on some headers on a few websites. It looks like headers that use the CSS attribute position: fixed.
In particular, I noticed it on https://www.nhl.com/scores/ and https://www.gamingonlinux.com/
And for some reason it only appears on the top left of the screen.
-
@iPristy memory leak is not reproducing.
Only ~10 seconds lag is reproducing (which was ~10 minutes lag when memory leak happened).By the way, I have 2 crash dumps, from 10:35: (DropMeFiles link).
Near that time task manager screenshot was taken.Skype also crashed somewhere between 10:32 and 10:35.
-
@Steffie Well I can't say what could be wrong (or it totally couldn't) with your Arch as it's much more "user dependent" (dependencies-wise and if everything is set right by the user at any time, your friend's "V-SS running fine" example) but considering you like to tinker in general and have many distros installed for that reason in VMs, and all your VM distros report the same thing which can't be a coincidence, maybe some parameter in your VM's general setup is responsible/uncompatible for the trouble across all distros? What VM do you use? Are there any similar reports in the Arch forums? But as long as you have reported it I guess we will soon know
-
@Steffie said in USA Today loading fix – Vivaldi Browser snapshot 2009.3:
this Snapshot needs the Southern Hemisphere Patch installed before it can work, down here. Was the package compiled this time without including the Coriolis hook?
You mean they failed to compensate for Coriolis force in the patch???
http://www.bbspot.com/News/2004/09/hard_drive_spin.html(sorry)
-
I had a really weird problem after installing this version – upon starting, Vivaldi attempted to run W:\Program Files\Vivaldi\Application\2.11.1784.3\Installer\setup.exe, which did not exist (and which resulted in error from Windows informing me that the program is missing); I solved the problem by copying the Installer directory from 3.3.2009.3 - Vivaldi started afterwards.