Re: [Monotone-devel] ikiwiki monotone support

[William Uther]

On 23/02/2008, at 10:56 AM, Brian May wrote:

> > > > > > "William" == William Uther <address@hidden>  
> > > > > > writes:
>
>    William> Um, syncing to a database is not going to call a hook in
>    William> _MTN/monotonerc either...
>
> I suspect we might be making different assumptions here. Or some other
> confusion exists.


Yeah - this is a little tricky.  And I don't remember exactly what I  
was doing
when I put that stuff together.

> If I  type in
>
> cd $path
> mtn sync
>
> Will it call the hooks in $path/_MTN/montonerc?

I believe so, yes.

> Nor is it possible to use a monotone server and ikiwiki from the same
> database, I believe the monotone server will lock the database so
> ikiwiki can't access it. So, on my setup, I have use two databases,
> and supplied my own script for the server database, which means that
> the supplied note_netsync_revision_received hook is still redundant.

My original thought was that the note_netsync_revision_received hook
would go in the monotonerc for the server mtn process.  That process
has nothing to do with the workspace that contains $path/_MTN/montonerc.

The rev-received hook was still being played with, when I got busy and
stopped having time to play.

> Also, if this code is used, you probably only want to call the ikiwiki
> wrapper script on the last revision received, not for every revision
> received.

Good point :)

>    William> er, no.  I believe that rcs_update() is called by ikiwiki
>    William> when it wants ikiwiki to update the workspace.  Look at
>    William> the comments in the Stub.pm example code in ikiwiki:
>
> I am still confused when rcs_update() is called. If it only gets
> called in response to the note_netsync_revision_received() hook, then
> the new revisions have already been received, haven't they?
>
> Or maybe ikiwiki calls rcs_update() on other occasions too?

My understanding was that it could call it at other occasions too.

I had a small test setup where I had one database and two local
workspaces.  There was no netsyncing going on at all.
I vaguely remember that working, although I've since
blown away that test setup, so I can't easily double check exactly
what I did.

Maybe we should check the ikiwiki source? ...  :)

It is called after an ikiwiki commit... and nowhere else.  Which is
weird.  I don't know how my setup could have worked.  Maybe something
has changed there in Ikiwiki?

Cheers,

Will         :-}