Re: [bug-mailutils] mh strange behaviour

[Fichet]

Hello Sergey,

Thank you for your answer and all the work done on mailutils. It's a
great piece of software, which is, as far as I can judge, quite well
thought.

I have discovered two others little things on mu-mh.

1) When composing a message, 'whatnow' could (should?) offer the
possibility to call 'mhn' before sending the message. The book "Mh &
NMH..." indicates that depending of the implementation, whatnow
understands the commands 'edit mhn' or 'mime'. Cf:
http://rand-mh.sourceforge.net/book/mh/wormes.html#mhEdanau

2) I have a problem with the PAGER and moreproc variables.
As far as I understand, when the showproc is the default one ('mhn'),
and the message is not a mime one, all options given to the PAGER
environment variable are not understood as options, but as a file to
display.

Example:

  $ export PAGER='less -Rs'
  $ show 1
  show: unable to exec less -Rs: No such file or directory
  # 'less -Rs' works usually well.
  $ echo "moreproc: less" >> .mh_profile
  $ show 1
  show: unable to exec less -Rs: No such file or directory
  # apparently 'moreproc' is not taken into account here.
  $ export PAGER='less'
  $ show 1
  # this one works as expected.
  $ export PAGER='less -Rs'
  $ sed -i /moreproc/d .mh_profile
  # the above deletes the moreproc line in .mh_profile
  $ echo "showproc: mhl" >> .mh_profile
  $ show 1
  # this one works as expected too.

Here is my .mh_profile for the above problem:

Path: /home/me/Mail
Editor: /usr/bin/vim
Local-Mailbox: 'me <address@hidden>'
inc: -changecur -truncate -file imaps://me%40domain.org:address@hidden
mh-sequences: .mh_sequences
Unseen-Sequence: unseen
Previous-Sequence: pseq
Sequence-Neagation: not
pick: -list -sequence pick
repl: -format


> Yes, indeed.  The default -truncate and -changecur options are in effect
> only unless at least one -file option is supplied.  Once a -file option
> is given, the safer -notruncate -nochangecur options are assumed.

That makes sense, indeed.
        
> It would be desirable indeed, and in fact, the work is underway on
> implementing the corresponding checks (UIDL for POP3 and uidvalidity
> for IMAP4), for both movemail and inc.
        
> > 3) 'show' does not update the 'cur' sequence when used. If 'cur' is s       
> Thanks for pointing that out.  It's definitely a bug and I will fix it.

That is a nice thing to read, thank you so much for all that work !

Kind regards,
Pierre-Jean.