"Laser Strike" COMPUTE! Magazine program conversion

[+OLD CS1]

@oddemann Thank you.

 

I let the program run for several hours playing itself.  It did not show any strange behaviors nor had it crashed at the times I checked on it.  That is a good thing.  But I did run into an interesting occurrence.

 

The computer logic is, as I stated, rather simple.  It appears that it can find itself in a dead-lock situation.  In particular, this game was the perfect storm.  In GRID 2, the computer player apparently found the two-part explorer, four-part bomber, and five-part mother ship very early in the game.  That left the three-part fighter to find.  The spacing of the random shots leads me to the early-discovery conclusion.  Note at this point there is only one area large enough to contain the three-part fighter, and where the computer should make its shot.

 

However, the computer did not take a shot for over three hours!  I finally terminated the program and started it running, again.

 

  

 

It seems that this situation would be rare enough not to impact a real game of human v. computer.  I spent time reviewing the logic in the original game which not only appears similar to mine (since I largely based my logic on the original's observable behavior,) but also does not appear to contain logic to break out of such dead-lock.  My concern is, as the shot-taking logic is transparent, I cannot be certain that an appropriate spot was not chosen and rejected.

 

I have begun considering a way to prevent this from happening, which could also increase the speed at which the computer takes its shots.  I will have to set up a simulation of this game to test the selection logic for failure.  While I want to pursue this further, I am still not convinced this issue will have real-world impact on games.

[+OLD CS1]

Dead-lock corrected.  I forgot to clear the detection variable for the solution proposal fit check when switching from horizontal check to vertical check.  The unexpected consequence is to make the logic a bit more effective... now that it checks all four directions.

 

Only other change is that a computer v. computer game ship deployments are not hidden, but are covered quickly after the explorer is placed.

LS-083020.zip

[Keneg]

Downloaded this today.  Played one game against the computer and had the computer play itself twice.  Very nice, if a bit slow.  Probably the best you can do in console BASIC.  I understand extended BASIC is somewhat faster.

[+OLD CS1]

@Keneg Thank you for the feedback.  Yeah, I spent a good bit of time working on some of the routines, which I have used in other programs.  I think where this one gets sluggish is in placing the cursor.  I made the decision to show a white dot over a hit ship as you move the cursor.  This involves a calculation of the character to place in that location, and I think it is complex enough to cause additional slowness on top of what we would expect from console BASIC.

 

I thought about this over the weekend and decided I might have something I can do to help with the sluggishness, but I am not sure if breaking the formula apart will give any speed increase versus the program logic which will be required to replace it.

 

Extended BASIC actually running this game as-it-is is appreciably faster, and I will have an Extended BASIC version released to complete the magazine type-in style.  The TI BASIC program is something like 300 lines, which is larger than I would normally expect for a type-in but, in all fairness, the version on other computers are built from multi-statement lines so I think it still fits the spirit.  The Extended BASIC version I hope would have felt right at home within the same pages.