> I'm not an expert but I don't understand why it will use more space
> than token.
>
Tokens could be saved in numbers (if needed and some users have that) and not 
in text. If saved as a number, then each token uses less space then any 
representation of the token as text.


> Token are now char(20), email address would be max
> char(100) but usually it's not longer than char(60).
>
It would be probably longer. The various RFC's describe a much longer character 
count for email addresses. But I think this should not be a big issue for most 
users anyway.


> For a long address : 
> # echo "[EMAIL PROTECTED]" | wc -c
> 44
> 
> char(100) is bigger than char(20) but I don't think that will be a
> problem for the load of DSPAM. This will take a little more space but is
> it very bad ?
> 
Probably not. I just wanted to bring the space usage issue to the desk and talk 
about. That's all.


> Without computing tokens, you recover time to match (truncated) entries
> in DB ;)
> 
Yes. Clean text is easier to handle.


Assuming I would take the needed time to develop that functionality. Would you 
be willing to code the management interface in the DSPAM web UI?

What about the way the email address is saved? Would just simple email address 
be enough or would you like the whole stuff to support regexp? Some storage 
engines support regexp in SQL. That's the reason I ask.

How would you like the function to work? Should it offer a per user storage for 
white lists or should it offer a layered level of whitelisting (aka: search 
user based whitelist, then search global based whitelist)?

Should those whitelist entries expire? I mean should a given entry be deleted 
after a certain time if not used?

etc, etc, etc.... A lot of questions where I think the DSPAM community should 
bring all their needs on the table and together we choose what to implement and 
how to implement. I have a whitelisting functionality in my setup but not 
inside DSPAM. I have implemented it outside of DSPAM. So I have a certain 
viewpoint but probably my viewpoint is not valid for everyone. I for example 
add automatically every recipient from an outbound message (but only if the 
sender is authenticated with SMTP AUTH) to a whitelist for +/- 90 days. If the 
email address does not send in the specified time range a mail to the original 
sender, then the whitelist entry get's deleted. And so on...

To all the people on the ML wanting that manual whitelisting functionality: 
let's hear about your needs.


// SteveB
-- 
Psssst! Schon vom neuen GMX MultiMessenger gehört?
Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger

!DSPAM:1011,487456dd150929635915936!


Reply via email to