Public SourceMage Discussion List

[Eric Sandall <eric AT sandall.us>]

Flavien Bridault wrote:
> Le mercredi 08 novembre 2006 à 16:01 +0100, Jaka Kranjc a écrit :
>> On Wednesday 08 November 2006 14:59, Flavien Bridault wrote:
>>> Ok it seems my methode catches everything. So before committing the
>>> stuff, I would like to be sure that I can remove all the TRIGGERS that
>>> contains "on_cast python cast_self" ? The UP_TRIGGERS research does find
>>> these spells, but I just wanted to know if those TRIGGERS were only
>>> intended for minor updates (2.4->2.5) or was it also for maintenance
>>> releases (2.5->2.5.1) or anything else ?
>> I wouldn't remove them, they will always trigger, not just now. 
>> UP_TRIGGERS 
>> would catch them only in this version-change case. Unless you made it more 
>> general, making more work for yourself. The stuff there now works, albeit 
>> all 
>> the if checks you posted should be changed to use is_depends_enabled.
> 
> Well after discussion on IRC, it appears that no one really knows why
> those TRIGGERS are here, except that *sometimes* they are useful. I have
> the feeling that they have been added to handle minor versions upgrade,
> as I am trying to do here. So in that case we have a duplication.
> However the UP_TRIGGERS is more efficient as it manages to trigger all
> needed spells and it explicits the case where this is needed, so here
> the minor version upgrade.
> 
> To verify this, I :
>   - removed all TRIGGERS on_cast python in the grimoire, 
>   - downgraded python to 2.4.3, and so triggered all python dependent
> spells with the UP_TRIGGERS,
>   - verified that most of the previous spells work,
>   - upgraded python to 2.4.4, so in that case the UP_TRIGGERS did
> nothing,
>   - verified that most of the previous spells still work, and they do,
>   - upgraded python 2.5, and so triggered all python dependent spells
> with the UP_TRIGGERS,
>   - verified that most of the previous spells still work, and they do,
> 
> So my proposal would be to remove all these TRIGGERS and to replace them
> by a single UP_TRIGGERS in the python spell, unless someone gives me a
> good reason. ;-) For instance if the TRIGGERS were there for an another
> purpose. I will wait until tomorrow evening before applying this.

AFAIK the TRIGGERS were there before we had UP_TRIGGERS trying to fix
this very problem (similar with many of the other TRIGGERS).

-sandalle

-- 
Eric Sandall                     |  Source Mage GNU/Linux Developer
eric AT sandall.us                  |  http://www.sourcemage.org/
http://eric.sandall.us/          |  SysAdmin @ Shock Physics @ WSU
http://counter.li.org/  #196285  |  http://www.shock.wsu.edu/