[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Why is IceCat/GuixSD slower than Firefox/Debian on headless servers?

From: myglc2
Subject: Re: Why is IceCat/GuixSD slower than Firefox/Debian on headless servers?
Date: Tue, 16 Jan 2018 19:23:57 -0500
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/25.3 (gnu/linux)

On 01/16/2018 at 11:56 Vladimir Sedach writes:

>> Running from a desktop to a headless server, and based on side-by-side
>> clicking, IceCat/GuixSD takes ~5x times as long as Mozilla/Debian to
>> respond to clicks. Further, Mozilla/Debian takes ~5x times as long as
>> Mozilla running on the desktop. The bottom line: IceCat/GuixSD is
>> unbearably slow.
>> Has anyone else seen this?
>> Is there a known IceCat/GuixSD performance problem?
> This sounds like an X11 protocol issue. Starting with Firefox 47,
> XRender support is disabled by default:
> Basically newer versions of Firefox stop using X11 correctly because
> Mozilla considers X11 "technical debt," and because X11 does not support
> some kinds of graphics acceleration.
> Back when Wayland development started the popular ideology became that
> no one needed or wanted remote graphics, if you want it use desktop
> forwarding like VNC, else tough luck. A lot of people bought into that
> and now we are seeing the results.

I'm not a VNC fan (sigh)...

> Check the value of gfx.xrender.enabled in about:config and try turning
> it on/off to see if that changes the performance. It should come enabled
> in Firefox 45, and is probably disabled in the newer IceCat.

Thank you, this was really helpful. gfx.xrender.enabled=false in
IceCat/GuixSD. Enabling it made the IceCat/GuixSD performance about the
same as Firefox/Debian (which comes w/ gfx.xrender.enabled=true).

> If it is lack of XRender causing the problems, the more interesting
> question is why running Firefox 45 remotely is so much slower than
> running it on the desktop.

OK, I tested some more. It look like I was exagerating about firefox
remotely. It looks more like 1/2 as fast. ISTM that ...


... could be 1/2 as fast as ...



- George

reply via email to

[Prev in Thread] Current Thread [Next in Thread]