[SOLVED]Firefox and updates

[ticojohn]

I just downloaded Firefox from Mozilla, but did not install it, and am running it from a folder in my /home directory. It appears to work as expected and I no longer having the Yahoo email crashes issue I had with Firefox-ESR. What I am wondering is this: will Firefox update when in that folder or does it actually have to be installed in order to update? I suppose, when updates are available, I can always download the latest version and extract it to the same folder, but an update might be less hassle. Anybody have experience doing this?

And as a side note, I wanted to put Firefox on the XFCE launch bar but could not figure out how to do that, so what I did was put Firefox-ESR on the launch bar and changed the command line in the properties so it launches Firefox instead of Firefox-ESR. Just need the correct, and complete, path to the Firefox executable. In my case that is /home/john/Firefox/firefox/firefox

[stevepusser]

You should be able to use its updater as the standard user, since you have it in a local directory.

[ticojohn]

You should be able to use its updater as the standard user, since you have it in a local directory.

Thanks Steve. Hopefully, when the time comes, it will work that way.

[deborah-and-ian]

Isn't using the Debian Mozilla Team Repo to get the latest Firefox more convenient?

[ticojohn]

Isn't using the Debian Mozilla Team Repo to get the latest Firefox more convenient?

I don't know if it is more,or less, convenient. Does the Debian Mozilla Team Repo load the latest Firefox directly from Mozilla?
Downloading directly from Mozilla ensures the latest version, and if it can be easily updated then no problem. It's not difficult. But, yes, doing it the way you suggest would be more in line with the Debian way of doing things. Actually, I was not even aware of the Debian Mozilla Team Repo method.

Thanks for the tip. And as a side note, I have found that the latest version of Firefox, downloaded directly from Mozilla, is WAAAYYYY faster than the latest Firefox-ESR from the Debian repos and doesn't crash with Yahoo email. It even loads Gmail faster than does Chromium.

[Head_on_a_Stick]

And as a side note, I have found that the latest version of Firefox, downloaded directly from Mozilla, is WAAAYYYY faster than the latest Firefox-ESR from the Debian repos

I call shenanigans!

I use FF version 50.1.0 in Arch (bspwm/xfce4-panel desktop) and firefox-esr in my BunsenLabs systems (Openbox/tint2) and I can tell no significant difference between the two in respect of performance.

Would it be possible to post some objective benchmarks?

See also https://en.wikipedia.org/wiki/Confirmation_bias

[ticojohn]

And as a side note, I have found that the latest version of Firefox, downloaded directly from Mozilla, is WAAAYYYY faster than the latest Firefox-ESR from the Debian repos

I call shenanigans!

Got me, didn't ya? I wouldn't even know how to go about taking performance measurements. I can only say, at this point, that when I open Firefox (50.1.0) it seems to open faster than Firefox-ESR and that when I browse to different sites they seem to open faster with Firefox than the same sites with Firefox-ESR.

So, how do I take measurements? Is there some kind of performance tool in Firefox? Perhaps I ought to take a look.

UPDATE: Okay. So I found how to measure the performance on Firefox. No problems with Firefox 50.1.0. However, with Firefox-ESR when I try to record the performance I get a message that says something like

real time measurements disabled in non multi-process Firefox

. HMMM! Not sure how to fix that.

[ticojohn]

And as a side note, I have found that the latest version of Firefox, downloaded directly from Mozilla, is WAAAYYYY faster than the latest Firefox-ESR from the Debian repos

I call shenanigans!

Got me, didn't ya? I wouldn't even know how to go about taking performance measurements. I can only say, at this point, that when I open Firefox (50.1.0) it seems to open faster than Firefox-ESR and that when I browse to different sites they seem to open faster with Firefox than the same sites with Firefox-ESR.

So, how do I take measurements? Is there some kind of performance tool in Firefox? Perhaps I ought to take a look.

UPDATE: Okay. So I found how to measure the performance on Firefox. No problems with Firefox 50.1.0. However, with Firefox-ESR when I try to record the performance I get a message that says something like

real time measurements disabled in non multi-process Firefox

. HMMM! Not sure how to fix that.

Here is the exact quote from Firefox-ESR when I try to measure performance

Realtime recording data disabled on non-multiprocess Firefox

[ticojohn]

And as a side note, I have found that the latest version of Firefox, downloaded directly from Mozilla, is WAAAYYYY faster than the latest Firefox-ESR from the Debian repos

I call shenanigans!

Okay, I was finally able to enable Firefox-ESR for multiprocess. When I run the performance test, the one thing that stands out is that it is recording an average of 31.7 fps. With Firefox 50.1.0 I see around 42.5 fps . Doesn't that kind of imply a fairly substantial performance increase (like around 35%)? Oh, and I had both versions reload a web page (Google News) and Firefox reload the page about 30% faster.

Now I know you're kind of skeptical, without seeing the actual data, so I am going to try and get you actual screenshots of the performance tests.

[deborah-and-ian]

Okay, I was finally able to enable Firefox-ESR for multiprocess. When I run the performance test, the one thing that stands out is that it is recording an average of 31.7 fps. With Firefox 50.1.0 I see around 42.5 fps . Doesn't that kind of imply a fairly substantial performance increase (like around 35%)? Oh, and I had both versions reload a web page (Google News) and Firefox reload the page about 30% faster.

Now I know you're kind of skeptical, without seeing the actual data, so I am going to try and get you actual screenshots of the performance tests.

You're comparing 2 different versions, but you claimed there were performance differences between the same version of Firefox from the Mozilla repos and upstream. Please make performance tests of Firefox 50.x from the repos and the latest Firefox from upstream. Thanks.

[Head_on_a_Stick]

http://browserbench.org/, perhaps?

EDIT: also, disable e10s on firefox-esr, it is not stable on that version.

[ticojohn]

You're comparing 2 different versions, but you claimed there were performance differences between the same version of Firefox from the Mozilla repos and upstream. Please make performance tests of Firefox 50.x from the repos and the latest Firefox from upstream. Thanks.

You're correct. That's all I had to go by as I have not installed 50.x from the repos. So we'll just let this go and mark it as a fubar by the OP.

[Head_on_a_Stick]

OK, I am really not at all happy about this but here goes...



From left to right: firefox-esr (jessie version), Firefox 50.1.0 (direct from Mozilla) & Google Chrome Stable (direct from the big G).

Poor old firefox-esr had crashes for pretty much all of the scripts and wouldn't play all of them fully

One final note: the Mozilla FF is the Windows version running under WINE.

I'm off to have a little cry now...



EDIT: hold on, I've just remembered that I'm running firefox-esr under firejail and that is probably skewing the test results down.

I will update later...

EDIT2: the score was the same but the tests didn't crash

[deborah-and-ian]

OK, I am really not at all happy about this but here goes...



From left to right: firefox-esr (jessie version), Firefox 50.1.0 (direct from Mozilla) & Google Chrome Stable (direct from the big G).

Poor old firefox-esr had crashes for pretty much all of the scripts and wouldn't play all of them fully

One final note: the Mozilla FF is the Windows version running under WINE.

I'm off to have a little cry now...



EDIT: hold on, I've just remembered that I'm running firefox-esr under firejail and that is probably skewing the test results down.

I will update later...

EDIT2: the score was the same but the tests didn't crash

These results are logical, as Firefox ESR is, what, like version 45? That was the beginning of multithreading. Version 50 has proper multithreading. I'm still waiting for Firefox 50 from the mozilla repos and from upstream.

[fireExit]

Firefox improved with version 53 (nightly channel) but still miles away from chrome stable.

[deborah-and-ian]

Chrome/Chromium has had multithreading since a couple of years now, so it makes sense that it's better than Firefox in this regard.