6 messages in net.sourceforge.lists.courier-usersRe: [courier-users] Re: whitespace pa...
FromSent OnAttachments
Patrick PriceFeb 5, 2001 9:03 am 
Sam VarshavchikFeb 5, 2001 3:30 pm 
Patrick PriceFeb 5, 2001 3:59 pm 
Sam VarshavchikFeb 5, 2001 4:17 pm 
Patrick PriceFeb 5, 2001 5:11 pm 
Sam VarshavchikFeb 5, 2001 5:55 pm 
Actions with this message:
Paste this link in email or IM:
Paste this link in email or IM:
Atom feed for this thread
Paste this URL into your reader:
Subject:Re: [courier-users] Re: whitespace parsing in userdbActions...
From:Patrick Price (sysa@moment.net)
Date:Feb 5, 2001 3:59:47 pm
List:net.sourceforge.lists.courier-users

At 11:30 PM 2/5/01 GMT, you wrote:

Patrick Price writes:

userdb entries are *very* touchy about having *one* tab after the username. One space, or an extra tab, or no tab, cause the user to fail authentication.

Which source file is this handled in? Improving whitespace skipping (tokenization?) would sure save me a lot of headaches.

makedat/makedatprog.c

You may have a case for smtpaccess or aliases, but not for userdb. There shouldn't ever be a reason to touch the files directly -- this is why you have the userdb script.

I'm thinking of actually writing scripts to manage smtpaccess and aliases.

-- Sam

I wasn't aware of userdb until you pointed it out. :^) I've been editing all my text files then running makeuserdb which didn't complain about extra or missing tabs after the username and then they wouldn't authenticate.

How about getting away from text files completely and working directly on the dbm .dat files?

What is the reasoning for having the text files to begin with?

-Patrick