[Bf-committers] Removing auto registration

Martin Poirier theeth at yahoo.com
Sun Jan 16 18:01:09 CET 2011


Hi,

I think it's premature to take a decision until a registration order that fixes problematic situations has been determined.

I even had a proposal for a fixed order that seemed to solve most issues (that I could think of) but never received any comments on it (nor on anything on that page for that matter). http://wiki.blender.org/index.php/User:Theeth/API_Design

Such a fixed order is easy to implement with automated registration.

Even if we decide that people need to solve the issue themselves, we must tell them how to solve it, not just give them a pat on the back and a good luck.

I already raised those points months ago when this discussion started, if anybody had more discussions about that, they certainly didn't make it on this list (some people can't be on irc all the time).

Martin

--- On Sun, 1/16/11, Brecht Van Lommel <brechtvanlommel at pandora.be> wrote:

> From: Brecht Van Lommel <brechtvanlommel at pandora.be>
> Subject: [Bf-committers] Removing auto registration
> To: "bf-blender developers" <bf-committers at blender.org>
> Received: Sunday, January 16, 2011, 11:51 AM
> Hi all,
> 
> In november there was already a discussion here about
> python class
> auto registration, but there was no agreement on the issue.
> So I was
> asked to try to cut the knot here, and I'm proposing to
> revert to
> manual registration.
> 
> On first thought, I prefer auto registration, because it
> corresponds
> to how I envisioned the API working when doing RNA design.
> But I also
> think the advantages are quite minor, and I agree after
> reading the
> discussion that on larger projects it can add more
> confusion. Also,
> there's a number of bugs and issues with it, which haven't
> been
> resolved and don't seem like they will anytime soon.
> Solving them will
> also make the API code more complicated.
> 
> We could discuss the topic further, but it also seems that
> a certain
> point this is a matter of preference. Campbell can do the
> commit if
> there's no objection.
> 
> Thanks
> Brecht.
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers
> 




More information about the Bf-committers mailing list