From Newsgroup: comp.os.linux.advocacy

Program optimization is essential, but yet it is difficult
to arrive at a best method.

For example, unrolling all loops can either improve or
degrade performance.

The user therefore, to get the best optimization, will
have to experiment through profiling or trial and error
to arrive at the best means. This can be prohibitive
for many users.

I propose that GNU/Linux programmers should determine
the best options and then publish these recommendations
in the source tree to guide the interested user.

When I create programs I always determine the best optimization
but these programs are only for my own use. They are
never published.
--
Gentoo: the only road to GNU/Linux perfection.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

Lester Thorpe wrote this post while blinking in Morse code:

Program optimization is essential, but yet it is difficult
to arrive at a best method.

For example, unrolling all loops can either improve or
degrade performance.

The user therefore, to get the best optimization, will
have to experiment through profiling or trial and error
to arrive at the best means. This can be prohibitive
for many users.

I propose that GNU/Linux programmers should determine
the best options and then publish these recommendations
in the source tree to guide the interested user.

When I create programs I always determine the best optimization
but these programs are only for my own use. They are
never published.

"Premature optimization"
--
... so long as the people do not care to exercise their freedom, those
who wish to tyrranize will do so; for tyrants are active and ardent,
and will devote themselves in the name of any number of gods, religious
and otherwise, to put shackles upon sleeping men.
-- Voltarine de Cleyre
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

Follow-ups trimmed to comp.os.linux.misc

In comp.os.linux.misc Lester Thorpe <lt@gnu.rocks> wrote:

Program optimization is essential, but yet it is difficult
to arrive at a best method.

<snip>

I propose that GNU/Linux programmers should determine
the best options and then publish these recommendations
in the source tree to guide the interested user.

I find O1 is good enough for all programs I create.

To, me, testing and retesting different optimizations is a
huge waste of time and at most you might save 1 second :)

For programs created by others, I keep whatever setting
the use since they know much better than me.
--
csh(1) - "An elegant shell, for a more... civilized age."
- Paraphrasing Star Wars
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

Le 10-09-2025, Lester Thorpe <lt@gnu.rocks> a écrit :

Program optimization is essential, but yet it is difficult
to arrive at a best method.

Yep. You proved it in your last video: you have no clue about the way to optimize your kernel.

For example, unrolling all loops can either improve or
degrade performance.

What does that mean? Do you want to exchange RAM for CPU? Or the other
way around?

The user therefore, to get the best optimization, will
have to experiment through profiling or trial and error
to arrive at the best means. This can be prohibitive
for many users.

Like you when in your previous video you chose only half of the options
by accident and avoid the other half?

I propose that GNU/Linux programmers should determine
the best options and then publish these recommendations
in the source tree to guide the interested user.

You mean: becoming a distro lackey?

When I create programs

I don't believe that.

I always determine the best optimization but these programs are only
for my own use. They are never published.

How convenient. The few times you provide some code you claimed to have created, two things were obvious:
- they never compile without some change,
- someone found the website which "inspired" you.

Another way to state things: your only "ability" is to find some code on Internet and to change it to render it impossible to compile.

Great job.
--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

On 13 Sep 2025 13:22:09 GMT, Stéphane CARPENTIER wrote:

Yep. You proved it in your last video: you have no clue about the way to optimize your kernel.

Please show us how you optimize YOUR kernel.

But you cannot. You don't build your own kernel. You are an ignorant
distro lackey.

Therefore, you have no basis for criticism.

Therefore, shut the fuck up.
--
Gentoo/LFS: the only road to GNU/Linux perfection.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

Le 13-09-2025, Farley Flud <ff@linux.rocks> a écrit :

On 13 Sep 2025 13:22:09 GMT, Stéphane CARPENTIER wrote:

Yep. You proved it in your last video: you have no clue about the way to
optimize your kernel.

Please show us how you optimize YOUR kernel.

Why? I did it decades ago when it was mandatory. I don't do it any more
as always say it: it's a waste of time.

But you cannot.

Of course, I have better things to do than loosing hours of optimisation
to win milliseconds of usage.

You don't build your own kernel.

Nope. It's obvious: it would be stupid and I've better things to do with
my spare time.

You are an ignorant distro lackey.

Nope. I know why I don't build my own kernel, it's unrelated with
ignorance, it's related to usefulness. My distro provide me with a very
good optimization of my kernel, if I ever see the need to compile my own
kernel again, I'll do it. But by now, it would only be a lost of time.

Therefore, you have no basis for criticism.

I have your video. Who tells by itself. It's the ultimate proof you
don't know what you do when you chose your options. Nobody needs to know anything about compiling the kernel to realize the only reason you
managed to compile your kernel rely only on luck and a well designed
kernel. And the distro maintainers know well more than you about their
reason behind the options, so they do a better job than you.It's
obvious.

Therefore, shut the fuck up.

I do what I want. You have no way to make me comply with your will. You pretended you sent me your killer boyfriends months ago but it looks
like they take their time. I'm still there.
--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

On 14 Sep 2025 12:10:21 GMT, Stéphane CARPENTIER wrote:

And the distro maintainers know well more than you about their
reason behind the options, so they do a better job than you.It's
obvious.

Ha, ha, ha, ha, ha, ha, ha, ha!

I *am* a distro maintainer. I have created my own distro that
is tailored to standalone, high performance, desktop workstations.
I just have not published it yet, but I could very easily, within
hours, post it to GitLab.

My distro does not use that garbage known as systemd nor even
the traditional SysV init. I have written my own special init
system.

However, I am too busy with other things to have to concentrate
on a publicly available distro.

What about YOU? Do YOU have your own distro?

Ha, ha, ha, ha, ha, ha, ha, ha!

No you do not and you never will. You simply do not have the
talent. You will be a lackey forever.

Therefore, as I already indicated, you have no basis for criticism.

Therefore, shut the fuck up.
--
Gentoo/LFS: anything else is a mere pipe dream.

--- Synchronet 3.21a-Linux NewsLink 1.2