Charles Cazabon <[EMAIL PROTECTED]> wrote:
>Dave Sill <[EMAIL PROTECTED]> wrote:
>>
>> He's got a qmail problem that he's trying to resolve--as a service to
>> the users of his utility--and we should be *helping* him, not telling
>> him there's no problem.
>
>I don't think it was quite this bad. He described his problem, and it
>was agreed that qmail would require a patch to give him the cleanest
>possible solution. He then said that patching qmail was not an option
>to solve his problem.
Patching qmail is not just "not clean", it's a nightmare. If you were
managing a system and a user requested that you apply a patch to a
major subsystem so they could use some particular utility, how would
you take that?
Another suggested fix is parsing virtualdomains. How would you feel if
a user wanted you to change the mode on virtualdomains? What if you
don't want all users on the system to know everything about your
virtual domain setup?
>The discussion then refocused on various degrees of ad hackery to work
>around the problem -- parsing virtualdomains, having a .cdb or other
>preparsed form of virtualdomains to do prepend lookups in, etc.
>
>How else should we have handled this, I wonder?
I don't have problems with the general back-and-forth approach. It's
the overall attitude of the discussion that bothered me: the
implication that Jason was just too lazy or incompetant to kludge up a
patch or a parser.
In short, we should have handled this as if we had respect for the
person to whom we were replying, not like we're gods tossing the
mortals occasional gems of wisdom.
-Dave
