Follow-up Comment #2, bug #67390 (group groff): At 2025-08-03T08:19:31-0400, Deri James wrote: > Follow-up Comment #1: > > Your patch has been committed:-
Cool, thanks!
> I have left this report open requiring further information, because it
> sounds like there is a further serious problem. Since the total
> run-time is about a second, I don't understand at what point you were
> hammering ctrl-c, was afmtodit looping?
No. I was doing a build on a relatively slow ARM-based tablet.
To reproduce the problem on my x86-64 daily driver, I added a
`sleep(1);` at the top of the `while (<AFM>)` loop in "afmtodit.pl".
> Did BuildFoundries not exit 2 if afmtodit exited with a non-zero
> status?
It did. But on that tablet, it took a while (several seconds) to get
there.
> Because you have not given sufficient information I have not been able
> to test your patch.
It's working fine for me. I've now found that there are two sources of
duplication warning. The second, I'm not inclined to expose to
`$opt_q`, because I think the real cause is the character mapping table
internal to _groff_ and which we inline into the generated `afmtodit`
script.
Specifically, we get duplicates between the following pairs of
PostScript glyph names:
* "Delta" and "uni0394" (capital delta)
* "mu" and "uni03BC" (lowercase mu)
* "uni03A9" and "uni2126" (capital ohm)
...but all of these are matters for bug #67244. On which I'm waiting
for feedback from you, incidentally, but I think I can proceed with
revising _libgroff_'s "uniglyph" table and similar in the meantime.
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?67390>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
signature.asc
Description: PGP signature
