Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uflorida!gatech!emory!dtscp1!joel
From: joel@dtscp1.UUCP (Joel Rives)
Newsgroups: comp.sources.games.bugs
Subject: Re: Galactic Bloodshed patches needed.
Message-ID: <660@dtscp1.UUCP>
Date: 3 May 89 20:22:41 GMT
References: <1501@scicom.AlphaCDC.COM> <12619@reed.UUCP>
Reply-To: joel@dtscp1.UUCP (Joel Rives)
Distribution: usa
Organization: Digital Transmission Systems (a subsidiary of DCA), Duluth, GA
Lines: 37

In article <12619@reed.UUCP> odlin@reed.UUCP (Iain Odlin) writes:
>
>  The game dumps its core when it tries to move a planet around a star.
>
>  It dumps a core when it tries to move a ship.
>
>  It goes into an infinite loop* under certain circumstances (which I haven't
>  figured out yet) in launch, build and a couple other commands.  I suspect
>  there is the potential of it happening in all of them. [*= By this, I mean
>  that you get the prompt infinitely printing itself after itself.  At this
>  point, the prompts will take over the screen and not accept any command,
>  forcing you to control-Z out and skill -9 it.]
>
>  Over half the commands don't check to see if you have enough "Action
>  Points {tm}" so when the game finds out you have a negative amount, it
>  throws you out of the game and calls you a cheater.
>

I have seen NONE of these alledged bugs in Galactic Bloodshed. I have seen 
one bug which is related to the "load" command. It dumps core whenever
it is passed a #shipnum directly. Invoking "load" and allowing it to prompt
for the ship number works just fine though. I haven't bothered to track 
down the problem. I am running the program on a Sun 3/60 running Sun OS 3.5.

I have made a few modifications to the daemon and looked at some of the
rest of the code -- particularly the doturn() stuff and the data structures.
With all due respect for someone who has donated a working (note the release
number 0.8) beta version, i was somewhat paled by the sloppy and nearly
indecipherable coding "style" throughout.

In regards to the daemon mods, they are simple and straight forward. I merely
changed the daemon to accept an integer parameter which would be used as
the update time cycle (in minutes). Note this change requires a redefinition
of the meaning of the UPDATE_TIME global definition in "tweakables.h", which
is used as the default time cycle if one is not passed to the program.

joel