Re: c-icap-modules not available anymore on pkgin command

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

Andries Annema <an3s.annema%gmail.com@localhost> writes:

> My apologies for using incorrect terminology. I'm still trying to wrap
> my head around the whole pkgsrc ecosystem.
> Your corrections give better insight already, thank you for that.

It's really not your fault; it seems things are not all clearly labeled
out there...

> Although your first statement does confuse me a bit: how can a pkgsrc
> fork (in this case the Joyent fork) have something that pkgsrc itself
> does not have?

All fork means is that it has a common history and some changes.  In
this case, the joyent fork tracks pkgsrc quite well, and has some bulk
build improvements, perhaps some other fixes, and also seems to have a
new "joyent" subdirectory with packages that are not in pkgsrc.

> I was under the presumption - maybe partly based on logical deduction
> - that any fork is likely to have only a subset of packages available
> on the pkgsrc base, but not/never the other way around.

I don't see that as likely at all.


There's a separate issue from what is in the source tree, which is what
packages  actually build on any particular platform.  People often
confuse these, saying things like "X is missing in pkgsrc-2019Q4 for
platform Y" when what they really mean is that a published binary
package set does not have an X package.   This is often due to it
necessarily failing to build because of how pkgsrc is, but it can also
be due to issues in the build envrionment.