2009/2/6 Martijn Faassen <faassen@startifact.com>:
Yep. Also, as I said before I think we also need to use deprecation warnings for imports that are not classes for persistent objects (until Chiristian writes the tool to upgrade them :)).
As far as I understand in a recent discussion people indicated they didn't like deprecation warnings anywhere, as they are forced on third party users of code that itself isn't deprecating anything. Since Christian's tool is as I take it well underway couldn't we just rely on this? Also since it actually *fixes* the state of a ZODB, instead of just giving a lot of warnings and then leaving the helpless developer with writing some scary custom upgrade code.
I don't think there was a consensus about that. For ZODB objects I think its okay for now not to use deprecations and to use the upgrade tool, but for the imports of other things that were moved elsewhere, I still think we need to bug developers that they need to upgrade their code with deprecation warnings, so we can eventually remove old imports. Oh, and BTW, if we use the "zodb fix tool", it's also okay to raise deprecation warnings about ZODB objects, as it's really easy to fix with the magic tool and again, we can eventually remove the deprecated import and make our code more clean. Let's discuss it once again :) -- WBR, Dan Korostelev