groff's pre-grohtml and posix_spawn() (was: groff 1.23.0.rc4 on mingw)

[G. Branden Robinson]

Hi Bruno,

At 2023-04-16T02:24:01+0200, Bruno Haible wrote:
> G. Branden Robinson wrote:
> > > This can be simplified to a single implementation for both, based
> > > on posix_spawn. Gnulib implements posix_spawn also for native
> > > Windows; see https://savannah.gnu.org/news/?id=10219 .
> > 
> > To make sure I understand clearly, do you expect gnulib's
> > posix_spawn to resolve _both_ of the above issues (notes 1 and 2)?
> 
> No, only note 2.
> 
> posix_spawn is faster than fork()+exec() on Linux/glibc, because some
> effort has been made to optimize that. I'm not aware of a similar
> effort for Cygwin, and won't do it myself either.
> 
> > I'll bet Gauis Mulley would have been thrilled to have gnulib's
> > posix_spawn way back then.  :)
> 
> Me too :) posix_spawn was invented in 2001. But it took a long time:
> Implementing posix_spawn for native Windows was like climbing a steep
> hill.

As nice at it sounds, migrating to posix_spawn() doesn't sound necessary
right now to resolve build or runtime support for Windows-based
platforms, at least not relative to curing regressions from the level of
support we had in groff 1.22.4.

But it's a good idea, if not obviated by the tbl(1) work I mentioned, so
I have filed https://savannah.gnu.org/bugs/?64055 about it.

Regards,
Branden

signature.asc
Description: PGP signature