Re: libthreads provenance

Roland McGrath <[email protected]> writes:

> If you do a diff of hurd/libthreads vs mk83/user/threads it's obvious which
> changes are ours and which are old CMU vs newer CMU.  Most or all of our
> changes have comments, most of which contain "GNU".  There are lots of
> other nit diffs you can see, like K&R -> prototype and 64-bit cleanups that
> we obviously never would have manually reverted.  There are some other dead
> giveaways like CMU copyrights with year 1993 present in MK83 vs missing in
> our version.  And to really prove my case that the MK82 code has changes
> that postdate your snarfing the code into the hurd:

Ah, yes indeed, the CMU logs do establish it.  Anyway, I think the
only way to get the exact answer is to look at old MK releases.  There
aren't too many to check, MK79 was the first that supported i386.

Why are we bothering at all, since this is soon-to-be-dead code
anyway?