Hi Sam, thanks for looking through the docs. :-)
[email protected] schrieb:
- I think your inclination to allow interface.start to get the configuration as a parameter is the right thing to do. I can't see any good reason to have to set the config separately.
Yeah, in the mean time I think so too. So I'll probably file a ticket for it.
In the same vein, I don't think the programmer should have to call interface.stop(). Can't TurboMail register itself with atexit to do this? I realize that there will still be cases where interface.stop won't be called (signals, os._exit, etc.) But I'd really like to see TurboMail be exactly one line of code to start, stop, initialize, etc.
Actually this is why we have adapters. The really make it easier to plug TurboMail into frameworks. For example I just committed a Pylons adapter which uses atexit and is only one line of code (+import). Feel free to have a look at the docs [1] to see if this is easy enough for you.
Heck, I'm almost thinking that the import statement itself could by default turn turbomail on and register stop with atexit. But that's probably taking things too far. :)
Yes, I do think so. I really don't like 'magic imports', they tend to mess things up in certain situations. So you'll have to live with two lines of code in Pylons (none in TG1).
fsPS: If you find any spelling errors/grammatical mistakes (I'm sure there are a lot of them), please just notify me :-)
[1] http://www.python-turbomail.org/changeset/137
smime.p7s
Description: S/MIME Cryptographic Signature
