Sure, having the hooks with arrays is definitely better than not having them at all. They won't be easy to change later, as people will have begun to rely on them the way they are, which is why its risky to rush into them.
If you have the system redone as arrays, it would be really minimal to refactor that into named arrays. I would certainly be willing to contribute to the effort. Either way.
Wouldn't life be dandy if there were units tests written to reassure us that everything was still working the same before and after the change. That's what I should really be working on.
Anyway, as long as the references work, I'll drop the issue. I think having the hooks be part of the release is important, as it will help the mediawiki software grow in different directions. Let me know how I can help.
/Jonah
----- Original Message ----- From: "Evan Prodromou" evan@bad.dynu.ca To: "Wikimedia developers" wikitech-l@wikimedia.org Sent: Monday, March 14, 2005 11:22 AM Subject: Re: [Wikitech-l] Late checkin on 1.4 branch
On Sun, Mar 13, 2005 at 05:16:58PM -0500, Jonah Bossewitch wrote:
It's a good idea, but I don't think it'll improve clarity. I think the current format is easier to write and read than having named params.
I guess its partially a matter of opinion,
Let me say that I understand that there could be some value in using a named-arg array. I understand that it would make adding new params less invasive, although I think that that's a less-than-frequent problem that can be handled other ways (another event, or adding params at the end).
We disagree which calling method would be better. Can we agree that named args are not _sufficiently_better_ that we should change this part of the software just before a release? Or to keep event hooks out of the release altogether? Because that's where this conversation is going, and I really don't like that direction.
~Evan
Wikitech-l mailing list Wikitech-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/wikitech-l