[Mailman-Developers] LoopError vs. Approve.py

[Barry Warsaw]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Jul 5, 2006, at 10:26 AM, David Lee wrote:

> (Mailman 2.1.8)
>
> Probably trivial; probably my misunderstanding.  But...
>
> ... at the end of "Approve.py" is some code to detect multiple passes
> through a list ("x-beenthere" etc.).
>
> But "Approve.py" is about handling passwords, whereas loop- 
> detection has
> no relation to such authentication issues.  Shouldn't this loop  
> detection
> functionality be in some other place, not "Approve.py"?  For instance,
> perhaps in its own, almost trivial, handler?

Ideally, yes, that's probably how it should (eventually?) be  
implemented.  It's the way it is mostly due to historical accident.

> (I noticed it because I'm writing an experimental sender- 
> authentication
> scheme as a custom handler modelled on "Approve.py".)
>
> BTW: Overall, as I dig into the Mailman code, I'm increasingly  
> impressed
> with its usually clean modularisation and its flexibility.  So this
> loop-detection (mis-)placement seems out of character.
>
> Many thanks for a good product.

Thanks!
- -Barry

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)

iQCVAwUBRKwPOXEjvBPtnXfVAQIdfQQAqgA88zshuPd1QGTdmqpL68TLx3aEoaWD
jKZTEqhK5AdX0NhWrBysrsXfCITlijnhUF7JFjNL3ewUm8jB6WyimQ/+dhW+/FId
tetu3f7C6jzA6ofb9DAaxnHfYDbf9xG9+jUXwBQTr/B8bhGG4bPAYMF3xq2neGRR
NdL9Cvc+ekE=
=EPSm
-----END PGP SIGNATURE-----