Hi Daniel,

I’m Shahrukh from [OSS Revival](https://github.com/Project-OSS-Revival). We specialize in revitalizing and maintaining open-source projects to ensure long-term stability.

As a team, we have a cumulative 15 years of experience in C/C++, and all our developers are programming polyglots with experience in maintaining OSS / commercial projects.

I am interested in the development ifupdown-ng-compat upstream issue  https://github.com/ifupdown-ng/ifupdown-ng/issues/247, ensuring it remains aligned classic Debian ifupdown.


Regards,

Shahrukh



On Wed, 11 Dec 2024 14:52:08 +0100 Daniel =?utf-8?Q?Gr=C3=B6ber?= <d...@darkboxed.org> wrote:

> Package: wnpp
> Severity: normal
> X-Debbugs-Cc: d...@darkboxed.org
> Tags: help
> Control: affects -1 ifupdown-ng-compat
>
> Hello internet,
>
> Debian is intending to replace classic ifupdown with ifupdown-ng, a
> clean-slate reimplementation intended from the beginning to be compatible
> with all existing flavors of ifupdown (Debian ifupdown, busybox,
> ifupdown2).
>
> Upon closer examination we found it is not yet fully compatible with
> classic Debian ifupdown to a degree we're satisfied with and boy are we
> ever sticklers for such things.
>
> Currently known incompatibilities are:
>
> - ifquery output is different
> https://github.com/ifupdown-ng/ifupdown-ng/issues/216
>
> - ifstate should be made compatible for a smooth upgrade path
> https://github.com/ifupdown-ng/ifupdown-ng/issues/246
>
> - interface renaming feature is missing
>
> - interface "pattern" matching is missing
> cf. /usr/share/doc/ifupdown/examples/pattern-matching or
> https://sources.debian.org/src/ifupdown/0.8.44/examples/pattern-matching/
>
> - old-school scripts (vlan, bridge-utils, ifenslave) need to be stubbed
>
> - `SKIP_DOWN_AT_SYSRESET=yes` is a bad default and a behavior change
>
> This upstream issue is tracking progress:
>
> https://github.com/ifupdown-ng/ifupdown-ng/issues/247
>
> The ifupdown-ng code is clean and simple C, making working on it a
> breeze. The challenge lies mostly in "reverse-engineering" and documenting
> the behaviour of the old implementation :-)
>
> Since ifupdown-ng would be installed on essentially all Debian systems
> (Priority: important) this is a low-effort but high-impact way to
> contribute to the project and consequently help with any of these would be
> greatly appreciated.
>
> For those hackers that need that extra bit of motivation: the alternative
> to ifupdown-ng as currently discussed seems to be further
> systemd(-networkd) monopolization -- make of that what you will ;-)
>
> Thanks,
> --Daniel

Reply via email to