Re: vc-modify-change-comment for "modern" backend fix

[Alfred M. Szmidt]

   > What about something like this, I've tested it briefly with a
   > vc-fossil-get-log-entry, and with RCS and it seems to behave as it
   > does previously -- though slightly confused by what it does for RCS
   > (it inserts the header, but not the commit?).

   Maybe it's me, but it looks a bit inelegant: log-view calls the VC
   backend, which then turns around and calls back into log-view? a VC
   method that isn't implemented in any backend, but instead does TRT for
   each backend "by hand"?  Can't we come up with something cleaner, even
   if that requires to add a new function?

It is a bit of a kludge, yes.  If someone can come up with an idea how
this could look like, I can try implementing that.  But I won't be
able to test this for SCSS, CVS, SVN or Hg which seem to support this
-- the current patch should still keep those working as previously
though which is probobly the only thing it has going for it...

I think the modify-change-comment functionality hasn't seen much use,
and still quite confused how it is intended to work.  E.g., in RCS, I
have this as the latest commit (and in log view mode):

  ----------------------------
  revision 1.3
  date: 2022/05/18 06:32:36;  author: ams;  state: Exp;  lines: +2 -1
  Summary: this is a test message
  ----------------------------

At the beginning of the buffer, log-view-modify-change-comment, then
you get a new buffer with the following:

  revision 1.3
  date: 2022/05/18 06:32:36;  author: ams;  state: Exp;  lines: +2 -1

Note the lack of the actual log message.  And when you C-c C-c that
you get basically something that isn't very useful...

  ----------------------------
  revision 1.3
  date: 2022/05/18 06:32:36;  author: ams;  state: Exp;  lines: +2 -1
  revision 1.3
  date: 2022/05/18 06:32:36;  author: ams;  state: Exp;  lines: +2 -1
  ----------------------------