Hi Walter,

Walter Alejandro Iglesias wrote on Thu, Oct 12, 2023 at 09:42:33AM +0200:

> Given that I'm the OP of this thread

It seems you mean the thread "mail(1) MIME support".

> I feel entitled to officially closing it.

I have no idea what "closing" means.

> Those interested in my proposal will find my latest version of the
> patch in this new thread:
>   https://marc.info/?t=169616631000001&r=1&w=2
> For my part, what was said in this thread has already been assimilated
> and overcome.  As my patches show, I heard and took into account every
> correction and suggestion (even though many, if not most, regardless of
> the intention with which they were made, functioned more as a boycott
> than as a contribution.)

You misunderstand.

The reason that at some point i temporarily paused feedback is that
such changes cannot be committed closely before release.  Meanwhile,
the tree is open again for 7.4-current development, but even if we
had finished patches ready for commit, i would likely delay committing
them for another week or two until snapshot building for 7.4-current
resumes such that the committed patches get real-world testing as soon
as possible.  On top of that, several developers, including myself,
are still involved in work related to making the release happen, so
the time available for post-release work is not unlimited - and as you
can see from the current, intense flurry of commits, in particular
in ports, that much of the time available for post-release work has
already been allocated to stuff that developers had lying around in
their personal pipelines, which also limits the time available for
completely new stuff like multibyte mail(1).

It is well-known that developers exist who frequently use mail(1)
so *if* whatever we end up doing disrupts their processes, we have
to find out as quickly as possible and take corrective action
without further delay.  Delaying the work on patches until the
right time is not a problem.  Committing something in a hurry
and not immediately getting it tested in real-world use may be
a problem depending on what and how important it is.

On top of that, we still haven't achieved consensus among
developers exactly what is the best first step.  We are getting
closer, and i'm quite sure we will ultimately add some MIME
support, but some decisions about desired functionality still
need to be made, for example how much automation is desirable
and acceptable and what the defaults should be.

Changing the core functionality of a program that is 52 years old and
has been standardized for probably almost 30 years, or maybe even
for up to 35 years, is not a quick and easy matter.  It needs proper
consideration, caution, time, and patience.


Reply via email to