From: thhal at mailblocks.com (Thomas Hallgren)
Date: Tue, 01 Feb 2005 19:33:26 +0100
Subject: [Pljava-dev] Re: PL Java Security
In-Reply-To: <003801c50865$786521e0$1e03a8c0@fhrbek>
References: <003801c50865$786521e0$1e03a8c0@fhrbek>
Message-ID: <41FFCB76.7080307@mailblocks.com>

Filip,
> I have tested this and I think it works fine. I don't like the solution 
> with the temporary function, but I haven't find anything better. Though 
> I think that it could be a good solution.
>  
> Please, if you find some time, take a look at the source or try to 
> compile and test it. If you like this solution, you can involve it into 
> your developement. If you like me to join the developement team, just 
> tell me what I should do for it.
>  
> There are not so many changes, you can use diff to find what has changed.
>  
I've looked at the code that you supplied and it looks fine. I'll take 
some time soon and incorporate it.

I have some plans that might replace the need for the temporary function 
eventually. It is possible (and very cheap) to switch the identity in 
C-code. I can expose that functionality to a limited set of classes. 
Typically, only the ClassLoader and the SQLDeploymentDescriptor needs 
this. The ClassLoader needs it because I don't want to allow arbitrary 
access to the jar tables.

Thank you for your contribution.

Regards,
Thomas Hallgren