Re: [Nmh-workers] unapplied patches

[Jon Steinhart]

> I've had a look through for any pending patches since the last release.
> Mostly, this is Meillo's recent patches. Does anyone know of any further
> patches that we might apply or rework.
> 
> Eric Gillespie: scan message numbers from stdin
>       It looks like people were generally in favour in principle, at
>       least until the details were discussed: Should it allow space
>       (or IFS chars) as a separator and sorting/uniqing of the list.
>       Perhaps it might be better to add -form and -format options to
>       pick. inc already has those options.
> 
> Savannah bug #29716: some slocal error handling
>       I can't see a problem with this but it might be good if someone
>       who actually uses slocal takes a look.
> 
> Exchange of MD5 implementation due to license
>       I don't really see a problem with this if a bunch of pedantic
>       Debian legal people are happier with the cut down BSD to the
>       current "no distribution of derivatives" code. And the newer code
>       is supposedly faster. It'd certainly be better for us to apply
>       it than for the Debian package to be carrying the patch around.
> 
> undo of install-mh
>       Nobody would ever run an uninstall-mh script and I doubt that the
>       person raising the bug was a user looking for it. Most software
>       creates a pile of dot files without even asking these days. If
>       any fix is necessary, I'd vote for a tiny bit more detail in the
>       man page - less than the original patch included.
> 
> gcc warnings: possible use of uninitialized data
>       I'm not so keen on initialising variables that don't need
>       initialising. That said, it'd be good to kill a few compile 
>       warnings.
> 
> I would have thought the following would be uncontroversial.
> 
> improvement of folder man page
> -link instead of -unlink in rmm.man
> small fix for configure.in (nmhman expansion)
> slocal.man .TP
> 
> Any objections if I apply those? Doesn't git have some ways to preserve
> details of the actual author of a patch? Or better still, could someone
> give Meillo permissions on the git repository.
> 
> If I understand correctly, the MIME handling improvments are not yet
> finished?
> 
> Oliver
> 
> PS. Sorry for the fact that I'm mixing what was in separate threads in
> a single message here. If replying at length on a particular patch such
> as MD5 or uninstall-mh, consider changing the subject back.

This all seems good although I'm not sure of the final details of meillo's
patches and haven't had time to look.  There were numerous messages about
backward compatibility and I'm not sure of the outcome.  I'm only in favor
of this if backward compatibility is preserved.

I think that David patched a bug in attach that did dumb things if no arguments
were specified.

The MIME handling improvements are coming along but are not even close to
ready yet.

Jon