[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: C-x v v no longer works the way it used to
From: |
Miles Bader |
Subject: |
Re: C-x v v no longer works the way it used to |
Date: |
Mon, 15 Oct 2007 07:51:24 +0900 |
Eric Hanchrow <address@hidden> writes:
> Jim> if I were visiting a non-version-controlled file and hit C-x
> Jim> v v, it would effectively add that single file to the version
> Jim> control system. i.e., cvs add, git add, etc.
>
> Jim> Now, all it does is print "No fileset is available here."
> Jim> Is this intentional?
>
> I assume so.
If so, it's arguably wrong. "C-x v v" is "do the right thing for
version control", not "commit this file".
Even if new behavior _is_ desirable, the error message is very vague and
confusing; something like "The file NAME is not under revision control;
use <binding> to add it" would be much more useful.
I think maybe some good guidelines might be:
1) Don't use unusual terms like "fileset" in error messages unless
it's impossible to do better.
2) Try to refer to something concrete that the user can associate with
the command he gave (one could use a helper function for signalling
vc errors that would use alternative phrasing depending on the
situation, e.g., "file FOO" for a single file, and "files FOO, BAR,
BAZ" for multiple files, etc).
3) Don't use vague phrasing like "not available here". What does
"here" refer to -- the machine, the current directory, some
internal code state...? What does "not available" mean? It sounds
like it could be a temporary error...
-Miles
--
((lambda (x) (list x x)) (lambda (x) (list x x)))