Christian Ohler <[EMAIL PROTECTED]> writes:

>> -;; FIXME: I don't know what AGAINST is for.
>> +;; FIXME: use AGAINST
>>  ;;;###autoload
>> -(defun xmtn-dvc-status (&optional against)
>> +(defun xmtn-dvc-status (&optional against path)
>> +  "run monotone inventory on optional PATH (default current tree), display 
>> results.
>> +
>> +AGAINST must be a revision specifier (number, last:N,
>> +revid:foobar, ...) or nil, but is currently not used."
>
> What's a "revision specifier"?  It looks like a bzr thing.  "number" and 
> "last:N" don't make much sense for monotone.

The examples seem to be taken from bzr.

The idea is that you have different ways to identify a revision
(different, both in the sense "several, for a given back-end" and
"different, according to the back-end"). In bzr, last:3 means, "the
last but 3 revision in the current branch" for example. There are
several ways to identify the same revision.

In DVC, a revision is identified by a lisp structure that can be
converted into a string.

See docs/DVC-API. `dvc-revision-get-file-in-buffer' is an example of a
generic function using this structure.

> In what situations would it be useful to run dvc-status against a 
> revision other than the workspace base revision?

It depends on the back-end you use.

For baz, for example (I don't care much about baz/tla right now, but
it has a strong historical influence on DVC), "status" mostly means
"short diff". So, you can compare you local tree with any other
revision.

There are probably many back-ends for which this is meaningless (and
perhaps monotone is part of them, I don't know).

-- 
Matthieu

_______________________________________________
Dvc-dev mailing list
[email protected]
https://mail.gna.org/listinfo/dvc-dev

Reply via email to