monotone-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Monotone-devel] Re: random thought on peg restrictions


From: Derek Scherger
Subject: [Monotone-devel] Re: random thought on peg restrictions
Date: Fri, 24 Mar 2006 21:11:39 -0700
User-agent: Mozilla Thunderbird 1.0.7 (X11/20060205)

Nathaniel Smith wrote:
So along the lines of peg revisions:
  http://svnbook.red-bean.com/en/1.1/ch07s03.html
it might be useful to have an "any" peg, which means, "any file whose
name matches this given name, at whatever points in history where it
matches".  So you'd say 'log foo/address@hidden' or whatever, and get a
list of changes filtered by surface name, not logical node id.

Yeah, that sounds like it could certainly be useful.

And the sort of functionality required to code this -- keeping names
around, instead of just going straight to node ids -- seems
interestingly close to the wartish part of the node-id-based
restrictions code, where it needs to do something similar to deal with
unversioned items?

True enough. So are you suggesting that this makes that particular wart look a little less ugly or somehow justifies it or are you thinking that we don't want to use node id based restrictions at all?

I have no idea whether this goes anywhere, it was just a random
thought before bed, so I thought I'd jot it down for future reference,
and in case it sparked anything down the line :-).

Along the same lines, I was doing a perforce sync (aka update) to a particular revision of some file the other day. The syntax for this is something like:

        $ p4 sync //depot/path/to/file#version

and as I typed that I wondered whether I was about to get the version I hoped for or whether '#version' would be stripped off as a comment. Interestingly, it seems that #foo is not considered a comment (by bash at least) if some non-whitespace character immediately preceeds it. What I mean is:

        $ echo foo #bar
        foo
        $ echo foo#bar
        foo#bar

which struck me as a bit odd but possibly convenient. It does seems a bit nasty though and I don't know if this is specified or standard behaviour. Perforce has various "special" version numbers like #none, #head, #have among others which all seem to rely on this quirk to some degree.

Cheers,
Derek




reply via email to

[Prev in Thread] Current Thread [Next in Thread]