[MUD-Dev] Re: Analysis and specification - the dirty words of mud development?

[T. Alexander Popiel]

In message:  <199806200103.SAA08807 at under.engr.sgi.com>
             J C Lawrence <claw at under.engr.sgi.com> writes:
>
>  Does anyone here have a base of regression tests for even any
>portion of their server or game code?

Yes, but only a very small portion; maybe .1% of the codebase.

>  Does anyone here have a test plan for their server or game base?

Nope.

>  Does anyone here have actually do any structured or automated
>testing of their MUD or game code?

Very rarely; normally we just test the individual changes, and
send the patches through code review.  This keeps us at least
as reliable and bug-free as the other MUSH servers, and more
than this just gets onerous, since we really don't have the
infrastructure to set up automated tests.  (The only testing
channel we have for most stuff is by opening a standard player
connection, running player commands, and analyzing the results.
Unfortunately, this channel passes through at least 6 different
subsystems, just for the most trivial of requests... analyzing
where any fault might lie without higher-order knowledge is
too bloody difficult to be worth it.)

- Alex