Performance of Firefox and Chromium

classic Classic list List threaded Threaded
17 messages Options
Reply | Threaded
Open this post in threaded view
|

Performance of Firefox and Chromium

Maximilian Pichler
Hi,

After upgrading to 5.9 (and thus Chromium 48 and Firefox 44) browser
performance seems degraded. Opening three different tabs with e.g.
newspaper websites results in a noticeable lag (up to several seconds)
when switching tabs or even just typing a URL into the address bar.

My question is: In the current state of things is this "normal" or
does it indicate that something is wrong with my system? I've read
http://www.tedunangst.com/flak/post/firefox-vs-rthreads, but am not
sure whether what is described there accounts for this sort of
behavior.

I'm aware that other browsers exist, but my question is specifically
about these two.

Thanks

Max

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Donald Allen
On Fri, Apr 29, 2016 at 8:11 AM, Maximilian Pichler <[hidden email]
> wrote:

> Hi,
>
> After upgrading to 5.9 (and thus Chromium 48 and Firefox 44) browser
> performance seems degraded. Opening three different tabs with e.g.
> newspaper websites results in a noticeable lag (up to several seconds)
> when switching tabs or even just typing a URL into the address bar.
>
> My question is: In the current state of things is this "normal" or
> does it indicate that something is wrong with my system? I've read
> http://www.tedunangst.com/flak/post/firefox-vs-rthreads, but am not
> sure whether what is described there accounts for this sort of
> behavior.
>
> I'm aware that other browsers exist, but my question is specifically
> about these two.
>

There has been a lot of discussion about browser performance on misc
recently. I run 5.9 and use a Firefox extension called 'Noscript' that
gives you fine-grained control over who gets to use javascript. It works
well, though it is a bit of a pain when you go to a new site and have to
tinker with it to allow the site to do its thing.

I have not tried 'current', but again, there is discussion in the archives
about internal changes (in the scheduler?) that have apparently improved
things significantly in this area. You should check the misc archives
yourself for the details; you might want to consider installing 'current'
if turning javascript off and on is unappealing to you.

/Don

>
> Thanks
>
> Max

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Maximilian Pichler
On Fri, Apr 29, 2016 at 8:45 AM, Donald Allen <[hidden email]> wrote:
> I have not tried 'current', but again, there is discussion in the archives
> about internal changes (in the scheduler?) that have apparently improved
> things significantly in this area.

Thanks for the hint. Updating to -current made both browsers usable
again, both for chromium and firefox. While there may still be room
for improvement, the difference is quite dramatic.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

alan01346
In reply to this post by Maximilian Pichler
Re: Performance of Firefox and Chromium

Several seconds?  Oh my.  Try 20 minutes or more on some of the most
bloated sites, with lots of reloads and watching iftop to see when
they're stuck like on my connection.  But thanks for the tip on
Noscript, I'm trying it out.
--
Credit is the root of all evil.  - AB1JX

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Solène Rapenne
Le 2016-04-30 14:23, Alan Corey a écrit :
> Re: Performance of Firefox and Chromium
>
> Several seconds?  Oh my.  Try 20 minutes or more on some of the most
> bloated sites, with lots of reloads and watching iftop to see when
> they're stuck like on my connection.  But thanks for the tip on
> Noscript, I'm trying it out.

In Firefox I also use FlashStopper to stop videos in html5 to autoplay
which is 1) boring and 2) resources consuming.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Alex Poslavsky
In reply to this post by Maximilian Pichler
On 04/29, Maximilian Pichler wrote:
>Hi,
>
>After upgrading to 5.9 (and thus Chromium 48 and Firefox 44) browser
>performance seems degraded. Opening three different tabs with e.g.
>newspaper websites results in a noticeable lag (up to several seconds)

Firefox saves its cache to ~/.cache, I mounted that as tmpfs and that
seemed to make it a bit faster as well. Off course you loose all your
cached stuff on reboot.

Alex

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Michael McConville-3
Alex Poslavsky wrote:
> Firefox saves its cache to ~/.cache, I mounted that as tmpfs and that
> seemed to make it a bit faster as well. Off course you loose all your
> cached stuff on reboot.

By many (most?) accounts, tmpfs is actually slower than an SSD. It's
presumably faster than a spinning disk, though.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

tinkr
In reply to this post by alan01346
On 2016-04-30 19:23, Alan Corey wrote:
> Re: Performance of Firefox and Chromium
>
> Several seconds?  Oh my.  Try 20 minutes or more on some of the most
> bloated sites,

Wait, what is the best guess now, did the recent scheduler patches that
were posted here recently, remedy the speed issue altogether?


Javascript may not be the fastest language around but also it's not That
bad, should be comparable with Python or shellscript I guess.

So then if JS-heavy sites specifically are the killer, it must be due to
some apparent irresponsible or redundant memory allocations, syscalls,
file IO or other bloat that the JS VM incurs?

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Mihai Popescu-3
In reply to this post by Maximilian Pichler
> Wait, what is the best guess now, did the recent scheduler patches that were posted >here recently, remedy the speed issue altogether?

Try and see for youself. I was doing that at almost each snapshot and
chromium was the winner all the time. Then firefox took over and I am
using it right now. It is because of the new modifications.

> Javascript may not be the fastest language around

A language has nothing to do with speed of execution!

If you want knowledgeable comments, please read tedu post and develop
further. The rest in this thread are assumptions.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Michael McConville-3
Mihai Popescu wrote:
> > Javascript may not be the fastest language around
>
> A language has nothing to do with speed of execution!

In many ways, it does. That discussion's off-topic for this list,
though. I'm happy to have it privately.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Donald Allen
In reply to this post by Mihai Popescu-3
On Sun, May 1, 2016 at 7:13 AM, Mihai Popescu <[hidden email]> wrote:

> > Wait, what is the best guess now, did the recent scheduler patches that
> were posted >here recently, remedy the speed issue altogether?
>
> Try and see for youself. I was doing that at almost each snapshot and
> chromium was the winner all the time. Then firefox took over and I am
> using it right now. It is because of the new modifications.
>
> > Javascript may not be the fastest language around
>
> A language has nothing to do with speed of execution!
>

That is simply not true in general. If your application is
processor-limited and written in an interpreted language, do you think it
would get faster if you re-wrote it in C (assuming we are discussing a good
programmer)? You're damned right it would!

What you say is only true if the application is inherently I/O-limited. It
is certainly not true of every application.

>
> If you want knowledgeable comments, please read tedu post and develop
> further. The rest in this thread are assumptions.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Stuart Henderson
In reply to this post by tinkr
On 2016-05-01, Tinker <[hidden email]> wrote:
> Wait, what is the best guess now, did the recent scheduler patches that
> were posted here recently, remedy the speed issue altogether?

Not altogether, but nonetheless a big improvement.

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Raul Miller
In reply to this post by Donald Allen
On Sun, May 1, 2016 at 3:49 PM, Donald Allen <[hidden email]> wrote:
> That is simply not true in general. If your application is
> processor-limited and written in an interpreted language, do you think it
> would get faster if you re-wrote it in C (assuming we are discussing a good
> programmer)? You're damned right it would!

I have seen significant counter examples (processor limited
application, written in an interpreted language, also written in C by
a good programmer - C version significantly slower).

That said, the interpreter was not javascript and was itself written
in C, so there's that, also.

(Point being once things reach a certain level of complexity, issues
like available developer time and architectural decisions and so on
can become rather significant. Also, I suppose, another point is that
even useful general statements can have counter examples.)

(Still, another issue is that web browsers fail, a lot: one person's
"better" is another person's "worse" and web browsers are caught
between billions of people and their conflicts.)

--
Raul

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Donald Allen
On Sun, May 1, 2016 at 9:54 PM, Raul Miller <[hidden email]> wrote:

> On Sun, May 1, 2016 at 3:49 PM, Donald Allen <[hidden email]>
> wrote:
> > That is simply not true in general. If your application is
> > processor-limited and written in an interpreted language, do you think it
> > would get faster if you re-wrote it in C (assuming we are discussing a
> good
> > programmer)? You're damned right it would!
>
> I have seen significant counter examples (processor limited
> application, written in an interpreted language, also written in C by
> a good programmer - C version significantly slower).
>

Only if the algorithm employed in C was worse. If all things are equal and
only the language was changed, obviously implicit in what I was saying,
then simply not possible (assuming a quality C compiler).

>
> That said, the interpreter was not javascript and was itself written
> in C, so there's that, also.
>
> (Point being once things reach a certain level of complexity, issues
> like available developer time and architectural decisions and so on
> can become rather significant. Also, I suppose, another point is that
> even useful general statements can have counter examples.)
>
> (Still, another issue is that web browsers fail, a lot: one person's
> "better" is another person's "worse" and web browsers are caught
> between billions of people and their conflicts.)
>

All of which is irrelevant to the (false) assertion I was discussing: "A
language has nothing to do with speed of execution!".

>
> --
> Raul

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

alan01346
In reply to this post by Maximilian Pichler
>> A language has nothing to do with speed of execution!

It seems like Javascript's gotten faster in the last 10 years or so.

I used to write little benchmarks to compare Turbo C and Turbo Pascal,
Pascal always won.

> (Point being once things reach a certain level of complexity, issues
> like available developer time and architectural decisions and so on
> can become rather significant.

For a one-time use program sure, but things like Python shouldn't be
unleashed on an unsuspecting public.  Gimp 2.8 is noticeably slower
than 2.6 I think it was in OpenBSD 5.2.  Move the cursor over the
image and it's like it's in la-la land.  Try to sign your name with
the mouse.  Of Inkscape and Libre Office Draw, surprisingly Libre
Office is the faster and works better for an SVG signature.  But not
as fast as this amazing little page:
http://mcc.id.au/2010/signature.html


--
Credit is the root of all evil.  - AB1JX

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Gregor Best-2
On Mon, May 02, 2016 at 11:55:34AM -0400, Alan Corey wrote:

> [...]
> For a one-time use program sure, but things like Python shouldn't be
> unleashed on an unsuspecting public.  Gimp 2.8 is noticeably slower
> than 2.6 I think it was in OpenBSD 5.2.  Move the cursor over the
> image and it's like it's in la-la land.  Try to sign your name with
> the mouse.  Of Inkscape and Libre Office Draw, surprisingly Libre
> Office is the faster and works better for an SVG signature.  But not
> as fast as this amazing little page:
> http://mcc.id.au/2010/signature.html
> [...]

I just tried it out on my Thinkpad T400 running a snapshots that's about
3 weeks  old and  I can't  reproduce Gimps  "la-la land".  I tried  on a
640x480 canvas,  with both the  pen and  the brush. Instant  painting in
both cases. The  Javascript thingie is almost as instant  but has a very
very tiny lag.

I'm not  sure what  hardware you guys  run OpenBSD on,  but on  my (old,
crusty,  crummy, shitty)  laptop,  it  and a  lot  of Gui-requiring  and
rumored to be "heavy" by whatever metric programs work nicely. That
includes Chromium by the way.

--
        Gregor

Reply | Threaded
Open this post in threaded view
|

Re: Performance of Firefox and Chromium

Kevin Chadwick-4
In reply to this post by alan01346
> >> A language has nothing to do with speed of execution!  
>
> It seems like Javascript's gotten faster in the last 10 years or so.
>

> I used to write little benchmarks to compare Turbo C and Turbo Pascal,
> Pascal always won.

Obviosuly C is generally faster than javascript and built in html5
routines should be, of course often javascript heavy sites send
javascript from many domains which is slow and insecure and probably
increases threading a lot by it's distributed nature.

--

KISSIS - Keep It Simple So It's Securable