[PATCH] 1321 (partial): run Python pretxnchangegroup hooks before disk flush

Doug Philips dgou at mac.com
Fri Jan 16 03:10:11 CST 2009


On or about Friday, January 16, 2009, at 04:00AM, Gilles Moris indited:
>On Fri January 16 2009 09:08:48 Doug Philips wrote:
>> For the hook itself, which is what the environment variable semi-proposal was going to permit, the pending changeset visibility is the whole point, never meant to say otherwise.
>> 
>> By "outside visibility" I meant "outside of the context of the running hook."
>> In order for an external hook to see the changes-which-are-still-yet-pending-and-abortable, currently, the changes had to be visible to all other viewers/readers of the repo. Fixing this so that the changes are only visible to the hook itself is what I was suggesting.
>> And as such, the ability for the non-hook viewers to see the changes too is a bug. Fix the bug, not a big deal?
>> 
>
>As I understand, the problem has two folds, the visibility is one.
>But I am much more worried about the possibility for these changeset to spread through a pull, contaminating repos in a uncontrolled way. This should really be addressed in a first step, by locking outgoing traffic.

One of us is confused. If the changeset is not visible, how can 'pull' or 'incoming' or anything else see it?
That is the beauty of the way the repo is updated, with the beauty mark/blemish being that the changeset might be rejected, but is still visible for pulling, etc.

--Doug



More information about the Mercurial-devel mailing list