Tom Mueller wrote:

> IIRC, when this was first sent out, there was the intent that pkg(5)
> would still run on 2.4.  But based on your comments at the end of
> 10485, I see that after that bug is fixed, then pkg(5) is no longer
> expected to run on 2.4.

That's correct, though if there's a need, we should be able to do it.  The
only changes which aren't backwards compatible are the hashlib bits.

> If 2.4 isn't going to be supported, then maybe the Uuid25.py file
> should be removed and we should just use the uuid module from the
> Python library.

Good catch, thanks.

> About the python dependency analysis tests, I've been looking at
> those tests this week relative to getting them to run on other
> platforms.  Even the other Unix-like platforms do not have files
> such as /lib/svc/method/svc-pkg-depot and
> /usr/xpg4/lib/libcurses.so.1. I can file a bug for this, but I was
> wondering if you had more thoughts about what the bug should
> actually say.  Should tests only depend on files that are checked
> into our workspace, rather than from the file system?

Those tests are broken in a number of ways.  I think Bart was actually
looking at removing them entirely because they cause the image at / to be
modified in some situations, though perhaps he and Brock found a way to
solve the problem in a less extreme fashion.

Danek
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to