Re: git: 2b48bc59c42b - main - ports-mgmt/poudriere-dsh2dsh: added fork of poudriere 3.4.99.20240812

From: Nuno Teixeira <eduardo_at_freebsd.org>
Date: Tue, 22 Oct 2024 08:08:02 UTC
Hello,

Awesome features like those presented in poudriere fork should be included
in main port as development options so people could test it properly.
The problem of a fork is that it will eventualy diverge from main project
and it can get into future problems.

Choosing between poudriere and poudriere-devel is already a difficult
decision to make related to reliability/stability/new features...

This is only a vague opinion on what I feel about a poudriere fork port.
I vote for port OPTIONS giving an oportunity of big testing features A, B,
C, etc.

Cheers,

Matthias Fechner <mfechner@freebsd.org> escreveu (terça, 22/10/2024 à(s)
05:12):

> Am 22.10.2024 um 00:33 schrieb Daniel Engberg:
> > While we do not have any written guidelines about the selection of
> > projects to be added to the tree this seem like a counterproductive
> > move given it's maintained within the FreeBSD umbrella already and
> > also potentially be perceived as poor management. Can you please
> > elaborate about the reasoning of adding this to the tree?
>
> I do not understand why this is a problem.
> The new version bring some signification performance improvements which
> is very helpful to users.
> For me, it drops the compile time of go based ports from 20-40 minutes
> to 2 minutes which is a massive improvement for poudriere.
> And if the developers do not include this modification, this is fine.
> But I think it is only fair to give also the user a change to decide and
> therefor are forks existing.
>
> Using a fork enable us to battle test new features without have an
> effect to stability of the core poudriere.
>
> If the port get many installations that would also give an idea to the
> developers how popular these modifications are and they can be included
> upstream.
>
> Sry, but I do not get the point, why that would be counterproductive, I
> see it exactly the opposite, it will help the poject to better get an
> idea if these modification are requested by the community and if they work.
>
> Gruß
> Matthias
>
> --
>
> "Programming today is a race between software engineers striving to
> build bigger and better idiot-proof programs, and the universe trying to
> produce bigger and better idiots. So far, the universe is winning." --
> Rich Cook
>
>

-- 
Nuno Teixeira
FreeBSD UNIX:  <eduardo@FreeBSD.org>   Web:  https://FreeBSD.org