geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeremy Boynes <>
Subject Re: TranQL Addition
Date Sat, 02 Jul 2005 16:24:50 GMT
Aaron Mulder wrote:
> 	You were the one who argued that key generators might be used by 
> web applications.  If we agree that key generators won't be used outside 
> of the EJB container, then the code can go in OpenEJB.
> 	However, I think the license story is actually better if we put
> this in TranQL.  I'm planning to move the existing key generator code into
> the pkgen module, and I'd like to move toward eliminating the
> PrimaryKeyGeneratorWrapper, which would mean the pkgen module would not
> dependent on Geronimo, would not have OpenEJB licensed code, and would
> include a bunch of TranQL ASL-licensed code.

Can we get to that point first then please to avoid the IP ambiguities.

> 	We could also put the pkgen in TranQL and the pkgen-builder in
> Geronimo (since pkgen-builder is going to be dependent on the Geronimo
> -builder infrastructure at least in the form of EarContext and
> DeploymentException and setting up dependencies on GBeans).

That sounds like a good option.

> 	Still, any way you slice this, I'm going to need access to the 
> current PK generator code.
> Aaron
> On Sat, 2 Jul 2005, Jeremy Boynes wrote:
>>This code needs a fair bit of Geronimo infrastructure to run. TranQL 
>>currently does not have any dependency on Geronimo to avoid circular 
>>dependencies. Is there a way this can be made independent of Geronimo?
>>This patch also contains code under the OpenEJB license whereas TranQL 
>>is all AL2.0 - do you have permission from the original contributors to 
>>With these two issues, would this be better in OpenEJB?
>>Aaron Mulder wrote:
>>>	I've got the primary key generator changes ready -- so they're now 
>>>XML elements instead of manual GBean definitions.  It's not 100% done, but 
>>>everything that worked before still works (including itests and TCK), so 
>>>I'd like to put it in and work on the remaining tweaks in the public eye, 
>>>so to speak.
>>>	However, this requires two new modules (a tiny bit of glue code
>>>and a -builder module), which I put in TranQL in my tree.  It turns out, I
>>>don't have permissions to add new modules to the TranQL CVS tree.  I don't
>>>remember if I'm not a TranQL committer or if there's just some special
>>>karma needed to import new modules.  In any case, could someone make me a
>>>TranQL committer if I'm not, or give me permissions to import the code?
>>>	Aaron

View raw message