Jochen Theodorou wrote:
> I did mean a patchfor Sql.java

Ho sorry, I'm very sorry I screwed up with recursive patch generation :/
Now they're right :
Patch for SQL
http://jira.codehaus.org/secure/attachment/36676/Sql.java.patch_2008_08_25.txt

Junit class (patch mode)
http://jira.codehaus.org/secure/attachment/36677/Junit_Sq_caching_2008_08_25.txt

/.../
>> Or I write it
>> by side, as a sort of groovy-fiction :) and we'll change it after
>> commit ?
>
> sure

>> For me, this last feature is a option to be use at your own risk. We
>> have to recommand closure-like feature, but if you want to something
>> with the second for your pattern, as a kind of general property, it's
>> at your own risk. Perhaps could we add a shutdowhook for closing
>> resources if the last feature is used ?
>
> no shutdown hook..

Ok. But I still thinking that we have to allow this feature and mark it as "at your own risk".
For Sql object passed to several method or shared by many objects it would be very usefull.
It's easy to implement and tricky to use with no attention, but it would be more difficult to use it the need be.
But if it's against general Groovy design, ok, let it be protected for developpers who want to use anyway.

And sorry for the overwork, I hate wasting my time and I hate to do other people wasting their time.


--
Marc DeXeT
CNRS - DSI
Pôle Expertises
<http://www.dsi.cnrs.fr>


---------------------------------------------------------------------
To unsubscribe from this list, please visit:

  http://xircles.codehaus.org/manage_email