I will just extract one thing:
On Tue, Jun 03, 2025 at 08:31:55AM -0500, Jacob Moody wrote:
>
> This is something we're actively working on improving, we have
> plans to replace ghostscript, we've continued to make progress towards
> removing ape reliance over the years. I would rather that continue to shrink
> instead of grow.
>
I will take the example of the TeX (kerTeX) distribution concerning
ape:
I don't use POSIX, but since I don't want to have to write different
scripts for different systems, I settled on POSIX.2 make(1) (that's
really minimalistic) and POSIX.2 sh(1) (not full POSIX: I use a subset
that works everywhere with more or less compliant Bourne like shells).
Even MS Windows can be used with MSYS64 this way.
Compilation is one problem, in fact the simplest (I could whether
generate Plan9 mk(1) compliant Makefiles from the stubs Makefile.ker,
but this would mean to review the shell commands...) but the other
problem is the packaging framework:
There are thousands of TeX related packages on CTAN, and LaTeX users
generally don't know what they use (because TeXlive provides
everything including the kitchen sink so that even for one letter
taken from a obsolete font, "it works").
My decision was then to not face the M:N problem (M CTAN packages to
be written in N different ways) but to keep M:1: I write a Bourne
shell recipe (because the Bourne shell can be available everywhere,
including on Plan9, and this is not ape, this is the Bourne shell utility
on Plan9), and only one, and this is kerTeX, hosted on the system, that
handles the packages wherever kerTeX is installed.
So concerning APE: the POSIX spec continues to evolve and it
is now absolutely crazy in size and it seems an impossible task to
maintain POSIX compatibility (even without options) on a not POSIX
system. Yes. But don't throw the child with the bath water. At least
keep an "enough" compliant Bourne shell and a POSIX.2
compliant make(1).
If APE would be dropped, I would have to provide the two utilities in
the compilation framework (risk_comp) in order to be able to compile
on Plan9 and, after, for kerTeX to be able to administrate the packages.
--
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
------------------------------------------
9fans: 9fans
Permalink:
https://9fans.topicbox.com/groups/9fans/T6b4ec01ec7f57dc8-M6dda7e61c0b8c37d1e2604b8
Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
