cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Leszek Gawron <lgaw...@mobilebox.pl>
Subject Re: Restructuring directory structure[was [Vote] Block artifact directory structure]
Date Mon, 06 Nov 2006 20:12:06 GMT
Giacomo Pati wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> 
> 
> Carsten Ziegeler wrote:
>> Leszek Gawron wrote:
>>> Carsten Ziegeler wrote:
>>>> Sorry, I got totally lost in all these mails. As far as I understand we
>>>> are currently discussing when/where it's possible to set the running mode.
>>>> There are two places: applicationContext.xml and system property. The
>>>> system property takes precedence over the applicationContext.xml. I'm
>>>> not sure which role cocoon.xconf should play in this?
>>> I think everyone means applicationContext.xml. I happend to put the 
>>> wrong name in the first place.
>>>
>> :) OK, fine, so is there still a problem?
> 
> Actually, yes, the problem is still there. If I understand Leszek
> correctly, we do have to places in code where we examine the running
> mode with the same algorithm.

I actually have two problems:

1. IMO cocoon components should either use Settings.getRunningMode() to 
get the current running mode or have the mode injected. Allowing 
components to determine the running mode using any algorithm (even the 
easiest one) leads to inconsistencies.

2. Some components (like CocoonOverridePropertyConfigurer, which has 
already been fixed) use the Settings object but happily fallback to 
default mode when settings object is not available. The result of such 
situation is that a bean that has been incorrectly dependence injected 
uses other running mode than you might think. Beans using running mode 
should simply throw if they cannot determine one.

I am probably picky about this but hey - you're the ones I've learnt my 
values from :)

-- 
Leszek Gawron                                    CTO at MobileBox Ltd.


Mime
View raw message