Re: understanding 2025Q2 in aarch64

[Greg Troxel <gdt%lexort.com@localhost>]

Ramiro Aceves <ea1abz%gmail.com@localhost> writes:

> El 13 de agosto de 2025 12:10:35 CEST, Greg Troxel <gdt%lexort.com@localhost> escribió:
>>Ramiro Aceves <ea1abz%gmail.com@localhost> writes:
>>
>>> Just curious, what technically does trigger the 10.0 ->10.0_2025Q2
>>> link to be made?
>>
>>It's not technical.  It's a human running rm/ln -s.
>
> Ok, it is a human running a command, but what is the condition that must be satisfied to decide to issue the command? Just curious.

It's what I said earlier.  It's a judgement that the build is at least
mostly finished and that the package set is at least mostly ok, such
that users are better off with the new set, in some fuzzy way taking
into account an assumed ensemble of users with different concerns.  I
personally lean to making that decision earlier rather than later,
weighing fixing security problems over inconvenience.  But I'm not
involved in aarch64 package builds and lack understanding of the
details.

> Anyway, from now I think I will always better point to the 202*Q*/All
> to also prevent for undesired upgrades.

I think that's a perfectly fine thing to do.

> Just the last question. What happens if you upgrade to the for example
> , 2025Q3 release and some packages do not work, can I return back all
> the installed packages to 2025Q2 or backwards upgrade is not
> supported?

I would definitely listen to jperkin.  I would think that changing the
URL back, pkgin up, pkgin fug, would return you to the previous step,
and if not it's a bug.   But then you get to keep both halves, figure it
out, and submit changes to fix the bug.

Backups are wise, always, twice as many as you think you need, unless
everybody mocks you for too much in which case you might be right.

The other thing you can do if you have a problem is check out the 2025Q3
branch and fix/build the problematic package.  Staying back is not a
good long-terms strategy.