[Monotone-devel] renaming monotone executable (again)

[Nathaniel Smith]

So, this has been discussed before:
  http://thread.gmane.org/gmane.comp.version-control.monotone.devel/5481

To spoil the ending, I'm going to suggest we switch to the name "mtn"
for 0.26 and going forward :-).

Why switch?
  -- Real-world usage shows, many many users do not like the long
     name, and alias it to something shorter.  Even people who don't
     alias it, still habitually shorten it in docs, in emails, in IRC,
     etc.  This is just crazy -- it's not like the present situation
     is unbearable, but effort spent doing this, explaining to readers
     what one meant, etc., is just loss.  There's no point to this.
  -- Having a single, standard abbrev is very useful.  If, for
     instance, we switch the command name to "mtn", we should also (I
     suggest):
       -- make the bookkeeping directory MTN (instead of MT)
       -- make the ignore file .mtn-ignore (instead of .mt-ignore)
       -- switch to using the ".mtn" extension for db files
          consistently in documentation, etc.  (There is no
          competition for this extension, according to filext.com.)
       -- switch the namespace prefix on attrs we know about, except
          that by some lucky coincidence this is already "mtn" ;-).
     and it promotes consistency in discussion and such; no more
     some-people-say-mt, some-say-mtn, ...
  -- If we're ever going to do this, now is the time.

Arguments against switching (attempting to summarize the points of the
above thread):
  1) A real word is a refreshing change from the usual cryptic Unix
       command names...
     Response: Perhaps, but experience shows that the long name
       bothers lots of people enough to work around it.  OTOH, I don't
       think even subversion, with its huge and increasingly less
       sophisticated user base, has had any requests to provide a
       command line executable named "subversion" instead of "svn".
  2) A real word is more descriptive.
     Response: Err, well, except "monotone" is completely
       non-descriptive to most people already :-).  And 'man monotone'
       or 'info monotone' or even 'monotone --help' already provide
       plenty of clue what this tool is.

To symlink or not to symlink:
  It's been suggested that we should keep the current name, but
provide a standard shorter symlink, or switch to a new name but keep
the old name as a symlink.  I don't really see the advantages of this,
outside of a knee-jerk attempt at a compromise solution.  We still
have to pick one or the other to be the official name used in the
documentation, etc.; making packages set up symlinks is a pain and has
to be done for every packaging format; it doesn't work on windows
anyway; and the experience from every other system, where this is all
a complete non-issues, suggests that _no-one_ actually yearns to have
a longer name available.

What to switch to:
  -- "mt" is the most natural choice, but not available.  Annoying,
     since no-one uses the old mt command anymore, but there you go.
  -- I'm kind of fond of "m" (take _that_, you upstart 2-letter
     systems like hg!), but it got shouted down the last time I
     suggested it :-).
  -- "mmm" -- less boring than other suggestions, has appropriate
     associations ;-), and is, in fact, a mono-tone... but just
     perhaps a bit too cute.  Also, annoying to type.
  -- "moto" -- dunno, also has nice associations.  That it's 4 letters
     is a bit unfortunate... maybe "mto" would be a compromise,
     keeping the short pronounciation...
  -- "mtn" -- the boring, generic choice.  Of course, sometimes boring
     is good.

What next:
  This is a classic bike-shed issue, but we do have a deadline :-).
It's either going to happen or not for 0.26.  We pretty much need to
decide whether it will happen or not, and what the final name will be.
I lean towards "yes" and "mtn", so if no-one speaks up that will
probably be what happens :-), but I'm not attached to that or
anything.  The "moto" suggestion is clever too, and I had forgotten
about it until Derek mentioned it on IRC a few days ago, and then
reviewing the old thread.  So, what do people think?

-- Nathaniel

-- 
"If you can explain how you do something, then you're very very bad at it."
  -- John Hopfield