Fwd: [Monotone-devel] Re: Removing a branch from my local database

[Andy Jones]

---------- Forwarded message ----------
From: Andy Jones <address@hidden>
Date: 23-Aug-2006 21:35
Subject: Re: [Monotone-devel] Re: Removing a branch from my local database
To: Bruce Stephens <address@hidden>

True.  In fact, I'd rather sync with a new local database than hack $CVSROOT.  Never again.

I guess my "eek" really translates to: "Where is this written down"? 

Which is why I've just started a new thread to talk about best practice.  If there are some docs somewhere, I want to read them.  If I'm the only one that thinks there needs to be something like that, fair enough (I think...).  OTOH if I can help Monotone by asking stupid questions and then writing about it, brilliant.

On 23/08/06, Bruce Stephens < address@hidden> wrote:

"Andy Jones" <address@hidden> writes:

[...]

> In order to get rid of the extra file size you'd have to create a new
>> DB and locally sync the branches you're interested in into the new DB,
>> and then drop the old one.
>
> Eeek.   That isn't good.

Why is that so bad?  syncing databases is a normal part of working
with monotone.

> I would have thought that this was a pretty likely use case, if you
> compare with how CVS or Subversion is used.

OK, with CVS you can probably get away with removing things from the
repository.  I'm not entirely sure---maybe CVSROOT/history has stuff,
and you'd might want to edit CVSROOT/modules.

How would you do this with subversion?  Surely you need to do svnadmin
dump, then filter, then svnadmin load into a new repository?  That's
*much* more of a problem than using sync in monotone to a new
database, isn't it?


_______________________________________________
Monotone-devel mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/monotone-devel