Firefox 106.0.3 fixes a crash and hangs on Windows
Mozilla plans to release Firefox 106.0.3 Stable later today, if you are reading this on October 31, 2022. The new release for Firefox addresses a crash and a hang issue that may affect Windows users.
Select Menu > Help > About Mozilla Firefox to display the version that is installed currently. Firefox runs a check for updates when the window opens and will download and install a new version automatically. Note that Firefox 106.0.3 will be released later today and that the browser won't pick up the update if you run the check early on October 30, 2022.
Firefox 106.0.3
Firefox 106.0.3 is an update that is relevant to Windows users only. It addresses two issues on Windows. The first fix addresses a startup crash for some users on Windows machines. The linked bug report, bug 1797464, reveals that Firefox is throwing an access violation exception on affected machines.
Enable/disable window occlusion only when compositor session exists on Windows
When compositor session exists, gfxConfig is already initialized.
If first AppWindow is destroyed in nsAppShellService::JustCreateTopWindow() because of error, the first window could be destroyed before calling gfxConfig::Init()
The bug affected Firefox 106.0, Firefox 106.0.1 and Firefox 106.0.2, and newer development builds of the browser. Firefox ESR 102 is not affected by it. The update to Firefox 106.0.3 addresses the issue, so that Windows users should not experience startup crashes anymore after updating.
The second fix addresses an issue with Windows 11 version 22H2 new Suggested Actions feature. The feature displays hover actions when users copy a date, the time, or a phone number on the operating system. The feature is controlled in the Settings of the operating system under System > Clipboard.
Mozilla noticed an incompatibility with Suggested Actions on the new Windows 11 version that could result in a hang in Firefox when copying text on webpages. The linked bug report, bug 1774285, reveals that Firefox could freeze for over 20 seconds when copying text, e.g., the URL of the address bar.
The browser freezes for a period of time (over 20 seconds), during which the entire window (chrome, viewport) is unresponsive to user input. Video and animation appears to continue playing. After a while, the browser becomes responsive again. The amount of time is variable, and seems to depend on the number of extensions loaded - my default profile can freeze for minutes.
The next major version of Firefox, Firefox 107, will be released on November 15, 2022.
Now You: did you run into any bugs lately when using Firefox?
Firefox View is a bit of an odd duck. it’s a tab, but you can’t close it on that page.. have to resort to doing it via the taskbar.
The Win11-bug was reported 5 months ago, and has been radio-silent since 3 months ago.
Now they had to do a point-update after the public release.
Worrisome.
Has Firefox brought back the manual update option yet?
How about I use Google Chrome? :D
A Reddit-style would be good, where comments with more points are put to the top.
I feel that in their rush to stick to their upgrade schedule that Mozilla is not thoroughly testing their released stable versions of Firefox. But yes, I know also that they are not going to uncover each and every bug.
I agree! However the same can be said about most software these days where it’s more about change for the sake of change and then using the general public as beta testers.
+1
I haven’t ran into any of the 106.[0-2] official bugs but I have encountered an odd issue with Firefox 106.2 (or was it 106.1 and 106.2?) which persists with this latest 106.0.3 update :
Checking for addons updates takes over a minute.
I unblocked all Firefox startup connections to Mozilla servers which allowed [firefox.settings.services.mozilla.com] to be established, which then allowed two other connections to Mozilla servers : nothing blocked. Issue unresolved.
I’ve spent hours trying to find the culprit. Seems I’ve found a work-around :
Exit Firefox – Firefox profile / startupCache / DELETE webext.sc.lz4
Restart Firefox, check for Addons Updates : no problem
Holds for 24 hours, so I do it every morning.
Something has changed in the management of Firefox addons, concerns apparently the relation between
[FF-PROFILE]\storage\permanent\chrome\*] and [FF-PROFILE]\startupCache\webext.sc.lz4
No idea if anyone else encountered such an odd issue which can very well be related to my own OS/Firefox environment and settings. The work-around works for me.
having same issues on 106.0.3 that did NOT have 106.0.2 or lower
although could be issue on their end.
@dmacleo, the culprit could indeed be on their end :
Extensions slow at updating : firefox
[ https://www.reddit.com/r/firefox/comments/yjzonw/extensions_slow_at_updating/ ]
“There are currently some issues checking for updates if you have too many addons installed:
[ https://github.com/mozilla/addons-server/issues/19837 ]”
I have over 50 extensions myself so that could explain it all …
@dmacleo, why is it that we feel relieved when others share a problem when a beautiful soul should in fact feel sadder? I shouldn’t rejoice to know that you encounter the same issue, should I? Unless maybe to consider that the culprit could be on their side as you state it, But why are we so few to mention this add-on update problem? Maybe because we share common Firefox settings and tweaks? No idea. Meanwhile this issue is bothering. And should this issue vanish (culprit on their side) remains the fact Firefox connects to Google servers which is quite a disappointment to be frank.
What happens is that up to FF106 I’ve always been able to update extensions (manually) with all connections to various Mozilla servers blocked at startup:
1 [ firefox.settings.services.mozilla.com ]
If not blocked calls :
2 [ content-signature-2.cdn.mozilla.net ]
If not blocked calls :
3 [ prod.content-signature-chains.prod.webservices.mozgcp.net ]
So I have to let all three connect at startup to be able to update my add-ons! What the heck?!
This was never required before Firefox 106. Never!
Moreover,
[ firefox.settings.services.mozilla.com ] : ISP = GOOGLE-CLOUD-PLATFORM
[ content-signature-2.cdn.mozilla.net ] : ISP = GOOGLE
[ prod.content-signature-chains.prod.webservices.mozgcp.net ] : ISP = GOOGLE-CLOUD-PLATFORM
Even the server required to check for add-ons’ update availability is:
[ versioncheck.addons.mozilla.org ] : ISP = GOOGLE-CLOUD-PLATFORM (which I *never* block of course)
For someone as myself who avoids Google, its services, its servers to the maximum extent of my possibilities, such a mix between Firefox and Google servers is annoying, to put it mildly.
At this time the alternative is between allowing all three mentioned above OR keeping them all blocked together with [ versioncheck.addons.mozilla.org ] given the latter won’t update if the former three are blocked at startup.
Not sure I won’t block them all and check/install updates from AMO only. Fed up of Google cancer. Can’t Mozilla proceed differently than with Google servers? Should this be part of the deal between Mozilla and Google? Why does Firefox have to connect to Google servers to work properly? WHY?
I don’t have that problem.
My sister has FF 106.0.3 and I can confirm that cheking for addons update is slower with FF 102.x ESR that is what I use. Time for FF 106 was 6.67 seconds, time for FF 102 was 0.29 seconds, the only addon installed is Ublock Origin. Just my two cents here.
* FF 106.0.3 is slower than FF 102 ESR while cheking for addon updates.
I didn’t notice any bugs. I really don’t know what they’ve been fixing lately.
Q: “did you run into any bugs lately when using Firefox?”
A: No, never in fact. If you don’t really care for having all the latest features with Firefox right away, then by far the most worry-free route is just sticking to the ESR versions (now at 102.4.0 esr).
^^ What Henk says. I’ve been using FF’s ESR versions since they became available.
Happy to let others do the de-bugging on new features that I rarely need or use.
Today is October 31, 2022 (hint hint)
Trick or treat! Awesome day! :D
Trick or treats! brings me back to 60 years ago: now I only treat, though here in France, in my area anyway, kids don’t ring at your door as they would (and still do?) in the States. Looks like we’ve imported the business but that the tradition hasn’t made its way to the kids’ fun and fascination I experienced within my childhood in NY then. Maybe in the States has kids behavior, that of enchantment, vanished as well? I do hope it hasn’t.
Here in NYC the kids do still ring bells, but not as much as when I was young. It now seems that in more affluent neighborhoods they only go to stores and ring bells of houses with decorations. Last year I sat on the stoop with candy and kids walked right past me, as I do not decorate. We never did that as children! But the “enchantment”, as you call it, does still exist here. I can’t say much about anywhere else in the States, though I suspect the “spirit” is still there. ;-)
@just an Ed, “[…] the “enchantment”, as you call it, does still exist here[…].
You don’t know how happy I am to hear that. Not surprised even if I had doubts given our modern era. Not surprised at least when it comes to NY, NYC, Big Apple! What I experienced when in Queens, NY (sixties) — and that I believe is typical of the USA and not only NY, is that adults “play the game” with kids, they don’t frown as in Europe (or France) when it comes to children’s dreams considered as childish when in fact it has the importance of kids’ enchantments, that US media announces officially latest news about Santa (ABC, CBS, NBC were those we knew in the early sixties, still broadcasting!), they pretend to be frightened on Halloween… they just adjust to honor kids’ dreams! that is something I really love in America, whilst some Europeans dare say that a society which has landed a man on the moon is childish … Visit America, live there for some time, and you’ll start to understand the complexity, diversity of American culture.
Sorry for this long off-topic, looks like I’ve kept all of my enthusiasm.. and enchantment!
Thanks for the article.
PS:
@Martin, could be possible to add some kind of score to the comments? I meant to be able to give a +1 or -1 to them?
Because voting system is toxic, just look at Reddit and how people treat voting system, they use bots to boost vote and downvote people. With enough downvote people can completely hide ones post.
Yes, ability to upvote but more importantly downvote brings a new level of integrity. In my knowledge Reddit is the largest online platform with downvote option and it is probably the best social media platform out there. Sure it is not always updated like Twitter and in some topics opinions are of less expertise. But in political, historical and scientifical sense with the added bonus of human experience it sure is the best. And that would’ve never been possible without downvotes. I think I went a bit off-topic but still point remains – downvotes really helps. However if it doesn’t ever see the day here, not a deal breaker. Upvote/downvote will help but things will be just fine without them. Cherry on cake that option. Ghacks would still be a tech cake without it.
@Yash Reddit’s downvote feature is exactly why the place is a huge echo chamber. On whatever subject.
Also, it just fuels hivemind thing and groupthink. This is why nobody with more than 2 brain cells uses Reddit for actual information.
And turn this into Reddit? Please no.
Why do people want to spread their disease everywhere?
I know John G. wants this because from previous comments, you can tell he’s an avid redditor.
All this “voting” system is going to do is what goes on on Reddit: On paper, the voting system is to let everyone know which comments are helpful and which ones are not.. but it just turned into “I agree” or “don’t agree” with your opinion.
Again, please no.
@John G., before Martin’s reply, I’d just like to add that I fully endorse your suggestion :=) Even if I remain aware that the downside of scoring is that some may simply give a score without any argumentation, as on AMO for instance, which occurs most often when providing a bad score. IMO scoring is nevertheless worthy provided readers give no attention to a whatever score which hasn’t been explained, motivated.
Agreed on that point. I have been testing out “I still don’t care about cookies” and testing things a lot just trying to help and for a fair while they had a clean sheet of 5 star ratings on mozilla and then someone came in with a one star rating with absolutely no comment so there is virtually no chance of fixing any issues at all.
I accept that there are going to be people that are likely affiliated with competitor products or cookie trackers/advertising and analytics companies but lets face it sometimes you just win.
There are downsides to this as some people might be controversial here but there comments often add to the conversation and often they also add much to the website even if I am not always inclined to agree.
The simple +1 or -1 might rob us of the chance to read some discussions but its whatever. I feel like many of the people here are passionate enough to upvote/downvote and still add their 5c or in my case 100 paragraph reply that’s usually too late and the article has had its time in the sun. lol
@Mystique, I’m running as well the ‘I still don’t care about cookies’ Firefox extension and indeed, 76 reviewers score 5 stars and 1 -one- reviewer scored 1 star. Perhaps he doesn’t even know why himself. Anonymity allows, or exacerbates, the most exotic behaviors.