Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Tigerwolf]

I'd let Firefox check for and install some updates and this was working up until version 2.0.0.7.   These came from Mozilla.

At that point, I noticed a VL release 2.0.0.7 flagged in the repository.   In order to  let the package database get up to date, I let glslapt re-install the same version.

Since then, it seems to find no updates at Mozilla.  Both 2.0.0.8 and now 2.0.0.9 have not been seen, and even a manually initiated check for updates says none available.

I've looked at update-related stuff in about:config, and everything there seems ok.

Is there anything in the VL-packaged Firefox that would prevent updates from being seen?

Is anyone else noting this behavior?

[exeterdad]

2.0.0.9 is compiling at this very moment for VL5.9, VL5.8 should be done Monday. 
Using the official Firefox update could be bad.  Our Firefox is compiled with our toolchain and flags.  Mixing official binaries with ours, might leave a bad taste in your mouth.

[Colonel Panic]

Mine updated automatically to 2.0.0.9 whilst I was using it (in Vector 5.8 ), I don't know whether that's a problem or a good thing.

[nightflier]

Default is to automatically update, but can only happen when you're running as root (or administrator in other OS).

[Tigerwolf]

Default is to automatically update, but can only happen when you're running as root (or administrator in other OS).

It was running under root since I'm doing upgrades to the master machine for an upcoming convention.

I even forced a Help/Check for updates.   It always (including just now) said there were none.

Our Firefox is set to check, but ask before installing anything, but I never saw any indication updates were available.  In the past, nag popups happened fairly quickly after a new release.

I can always do a manual update, or wait for the VL packaged releases, but was more curious as to why the VL version was ignoring checks for new releases in case the install of 2.0.0.7 changed something, or left some cruft behind someplace that borked it that I need to clean up.

I will stick the one Exeterdad's working on in once it's available and see how it does.

[exeterdad]

As far as I know, I haven't disabled anything to suppress the checking. If you check out

about:buildconfig

in your browser, you can see the build options I used.

[Tigerwolf]

As far as I know, I haven't disabled anything to suppress the checking. If you check out

about:buildconfig

in your browser, you can see the build options I used.

Oh!  Cool!  I didn't know about that one!   

Now if information like this was included as part of every VL package, life would get lots simpler!

[GrannyGeek]

My Firefox 2.0.0.7 won't update, either. It tells me there are no updates available (running Firefox as root in order to check). This is with 5.9 beta 2.4.
--GrannyGeek

[exeterdad]

A little digging, I didn't disable the features, but I didn't enable them either.

--enable-update-channel=release --enable-update-packaging

I'm can't be sure this is a good idea though.  Not when Firefox is compiled by us.  I may build a test package with these flags enabled to test to see it it blows up when updating.  But wouldn't think about releasing until it's been tested.

[GrannyGeek]

I've used Firefox's Update feature for a very long time and never had a problem. I start out with whatever version comes with VL and just update through Firefox from there. So if bad things can happen by updating this way, I haven't encountered them. Maybe I've just been lucky.
--GrannyGeek

[Tigerwolf]

I've used Firefox's Update feature for a very long time and never had a problem. I start out with whatever version comes with VL and just update through Firefox from there. So if bad things can happen by updating this way, I haven't encountered them. Maybe I've just been lucky.

I've not seen any problems using Firefox's own  updates  either.   However, I usually install the official VL packaged version later to keep the package database up to date.

The problem is, with many machines in public use, closing any holes or crasher bugs via FF updates gets the fixes in place ASAP, since it's the most used application.

[exeterdad]

So if bad things can happen by updating this way, I haven't encountered them. Maybe I've just been lucky.
--GrannyGeek

Until Firefox-2.0.0.6, VL Firefox packages were merely repacks of the Firefox you download right off of the website.  Since then we have been compiling our own from source.  So since we are doing that, updates through the Firefox interface would mix binaries compiled with different tool chains and flags.  That's why I fear something could go wrong.

[GrannyGeek]

Until Firefox-2.0.0.6, VL Firefox packages were merely repacks of the Firefox you download right off of the website.  Since then we have been compiling our own from source.  So since we are doing that, updates through the Firefox interface would mix binaries compiled with different tool chains and flags.  That's why I fear something could go wrong.

I would prefer the old way (repacks of the Firefox you download off the Web). Then we're not forced to wait for a VL package when a new Firefox is issued. Most of those updates are for security fixes and I'd prefer to have them right away.
--GrannyGeek

[exeterdad]

I would prefer the old way (repacks of the Firefox you download off the Web). Then we're not forced to wait for a VL package when a new Firefox is issued. Most of those updates are for security fixes and I'd prefer to have them right away.
--GrannyGeek

This is a good thread. Some valid thoughts and concerns here. Thought I'd report back to you that we are looking into this.

[nightflier]

Do we gain any speed or other advantages by compiling our own version?