Re: gnupg2: fail with libgpg-error 1.47

[tlaronde%kergis.com@localhost]

On Thu, Jan 09, 2025 at 09:13:04AM -0500, Greg Troxel wrote:
> 
> but a higher level point: You are seeming to want to run your system
> with old packages.  In theory this will all work.  In practice ~nobody
> who hacks on pkgsrc does this, so staying up to date is the recommended
> plan.  I don't understand why people want to have a mixed/old
> environment.  This leads to not getting bugfixes including security
> bugfixes.
> 

I don't "want" to run with old packages. But since there may be
packages with non default options, I need to compile.

I update pkgsrc regularily. Then I update some package(s). When the
installed versions are OK with what is installed, the new version of
the package builds. But some things hence drag behind.

I was hit once, and will never try the thing again, with a "pkgin"
that updated everything. Samba was installed and working (just sharing
files with Windows; nothing fancy; not big network of nodes, neither
complex "users"). But the version was old (and since it worked, I
didn't update it), but pkgin desinstalled it and was unable to install a
new one (I guess because in the mean time there was a samba4 and a
samba3, while the pkg was only "samba") and I noticed it was
desinstalled only when having to reboot (since the programs were in
memory, till reboot, it worked...). [and even installing a new version
would not have solved the problem: it doesn't work with some Windows
programs handling text files.]

The logic would be to not use the pkgsrc version for programs that are
crucial or that are picky (one version could work, the other one: no),
and to be able, for the rest,  to rebuild everything from scratch when
updating pkgsrc, but to be able to test before switching.

One may have nodes that are special and for which it is not possible to
test without running things on them. If these are crucial for a work to
be done, this is not always possible.

I don't blame pkgsrc---I'm happy to have a package for firefox and the
like; I wouldn't want to have to deal with the upstream source by
myself. But for the rest, I'm facing real world constraints too.
That's all.
-- 
        Thierry Laronde <tlaronde +AT+ kergis +dot+ com>
                     http://www.kergis.com/
                    http://kertex.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C