emacs-devel
[Top][All Lists]
Advanced

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

Re: Should native compilation be enabled by default?


From: Eli Zaretskii
Subject: Re: Should native compilation be enabled by default?
Date: Fri, 08 Mar 2024 10:28:08 +0200

> From: Richard Stallman <rms@gnu.org>
> Cc: emacs-devel@gnu.org
> Date: Thu, 07 Mar 2024 21:29:54 -0500
> 
>   > This is true, but the slowdown is IME insignificant on a reasonably
>   > modern system.
> 
> Could you explain what "reasonably modern" refer to OS releases, to CPUs, or 
> what?

CPU and memory.

> The only CPUs we can morally recommend to people are the ones before
> the hardware backdoors became impossible to disable completely.

This is not about recommending CPU to people, this is about the level
of CPU power and memory size that most Emacs users have nowadays.
Given the popularity of native compilation, I conclude that very few
of them stick to old CPUs and small memory sizes.

> They are around 15 years old.  Does installing a current GNU/Linux distro
> release on one of those computers add up to a "reasonably modern system"?

I don't know enough about those systems to answer that question.  If
it turns out that those systems are too slow for building Emacs with
native compilation, there's the --without-native-compilation
configure-time switch, which is very easy to use, and needs only to be
specified once for every source tree.

IOW, the default configuration of the Emacs build is supposed to cater
to the average system which Emacs users have, not to the lowest
possible ones.  For example, we build by default with image libraries
if those are installed, even if the user always invokes "emacs -nw"
because the system cannot endure running an X server.  For systems
that are far from the average, we offer configure-time options to
disable features that get in the way for some reason.

>   > I disagree, based on my experience.  I see a significant speedup when
>   > using Rmail for just reading email, something that I cannot
>   > characterize as "power use" of Emacs.
> 
> Could you tell me more about this experience?  Which operations become
> faster?

Reading new messages and displaying messages, especially if the
message is in HTML (as opposed to plain text).

> How big is your Rmail file?  How many messages are in it?

Currently about 75 MB and more than 2K messages.  I purge it weekly,
and after a purge it becomes about 1/3rd that size.

>   > And there are other significant speed advantages.
> 
> Would you please tell me more?

This was all discussed during development of Emacs 28, two years ago.
I don't remember all the details, and cannot afford looking through
the archives to find the discussions with benchmarks.  Maybe someone
else (Andrea?) could look them up.  I do remember the conclusion:
natively-compiled code was on average faster by a factor of 1.5 to 2.
And that is good enough for me, and I think should be enough for
anyone, to support the decision of making this the default, as soon as
the initial issues with it (and there were issues, believe me) are
resolved.



reply via email to

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