[TransWarp] PROPOSAL: Split "SEF" into "CIS" and "SEF"

[Phillip J. Eby]

At 06:03 PM 6/9/02 +0200, Ulrich Eck wrote:

>I've run into several inconviniences when developing the Database stuff
>recently and would like propose changes there as well:
>
>I don't find it very usefull to have all connection-types in one file, if you
>want to supply more than just a few connection-types in longer terms.
>Also several Models need more than just one File and should be placed into
>seperate directories:
>
>for my Development i use a structure like:

Just a reminder: as I said previously, the Database package will see 
*heavy* refactoring in the 0.3 release cycle, and is effectively 
"unsupported" during the 0.2 release cycle.  Something like your 
structuring may take effect in the 0.3 refactoring, or it may not.


>Perhaps there is another thing of interest:
>NLL.Utils.InstancePool:
>a first draft for transparent InstancePooling that
>has a first implementation: TransactionAwareInstancePool.py. It creates a
>number of Instances  (e.g. Database) and blocks other threads until a
>Transaction has finished its work and an Instance is free for use again.
>We talked about this shortly and an example can be found under:
>examples/Util/InstancePool

Hm.  This strikes me as problematic to use with CIS bindings, unless the 
pool is a proxy of some kind, and then it needs to also know when to 
allocate/release an object.  My architectural assumption about threads and 
pooling is that you'll create separate instances of the App-level object 
and pool *those*, rather than individual components within the 
application.  This works well in Zope, and it localizes the lifecycle 
issue, while avoiding proxies and pretty much all possible thread 
synchronization.