Fixes to "Armor..Attack" binary

[gliptitude]

Oct 9, 2013 16:49:43 GMT -5 parabellum said:

I'm using a disassembler/tool I've coded from scratch, since all the available alternatives have the problem you're pointing out.

Would this tool be available to others?

Oct 9, 2013 16:49:43 GMT -5 parabellum said:

Well theoretically it should be possible to replace it with anything since I'm not limited by the binary size anymore (the original coder was definitely hindered by the 4Kb limit), but I'm short on ideas...

I haven't seen it yet, but the spinning explosion sounds really cool. ... Other ideas? How about a skull and crossbones which would begin tiny and zoom into focus? Maybe too many vectors? (it looks really cool when vector graphics begin invisibly tiny and then fluidly zoom in, like in the arcade Zektor. I can't really recall any Vectrex game that does this). ... Are you able to replace or modify ANY graphics in the game???

Oct 5, 2013 15:52:10 GMT -5 parabellum said:

I might have a go at Fortress of Narzod afterwards, but I really can't promise anything for sure. Debugging a 'third party' binary is a quite intricate and time-consuming task, and this one is 8Kb (double the size of ArmorAttack), which won't help

But about the hit bug, if I remember well, an enemy that has been hit has its brightness decreased, hasn't it?

Well thanks for considering it. If it's one of the more difficult "debuggings", at least it would be one of the most justified because it is such a popular and original, but also frustrating game. .. Maybe you are right about the flickering indication. Hard to tell though because the display flickers at other times as well. At any rate a decisive sound effect (or simply one hit kill) would be much more satisfying I think.

... I'm not sure I have any great ideas for your current project/favorite game. ... But several other GCE games have glaring problems holding them back, while others have some natural improvements that would make them better.

What if Star Castle was modified so that the fortress rings actually lined up with the overlay?

Dark Tower has a lot of slow down in the minotaur sequence. Perhaps sprites could be simplified?

Dark Tower is impossible! Perhaps a coordinate indicator could replace the directional indicator?

... I have had a lot of fun using Vectorzoa's drawing tool, and displaying my own sprites on Vectrex. Also V-model is easy to use and provides an intuitive drawing environment. I think lots of people could come up with cool vector objects using these tools. YOU COULD HAVE A SPRITE HACKING CONTEST!!!

[cNp]

This is great to hear you might be able to fix some of the niggles in the original code.

My son and I play Armor Attack a lot and often run into a bug where 1 player dies when the already dead, other player gets shot (even though they aren't meant to be on the screen anymore... I guess their jeep is just made invisible but can still be shot?)

cNp

[hcmffm]

Oct 10, 2013 12:47:33 GMT -5 gliptitude said:

Oct 5, 2013 15:52:10 GMT -5 parabellum said:

I might have a go at Fortress of Narzod afterwards, but I really can't promise anything for sure. Debugging a 'third party' binary is a quite intricate and time-consuming task, and this one is 8Kb (double the size of ArmorAttack), which won't help

But about the hit bug, if I remember well, an enemy that has been hit has its brightness decreased, hasn't it?

Well thanks for considering it. If it's one of the more difficult "debuggings", at least it would be one of the most justified because it is such a popular and original, but also frustrating game. .. Maybe you are right about the flickering indication. Hard to tell though because the display flickers at other times as well. At any rate a decisive sound effect (or simply one hit kill) would be much more satisfying I think.

Darrin and I have played Fortress of Narzod quite a bit during the game of the month contest of the Retrocade Magazine. Again, I've played it during the Vector War. I'm convinced that there is no problem with collission detection. You just have to hit an enemy twice and it will be killed, then. The difference (no hit/first hit) can be seen in the brigthness of the enemy, with the right adjustment of the brightness one can see the difference. Rather than changing the game to a "one hit kill" and thus changing the difficulty level, I strongly suggest to just make the difference in brightness larger.

[Rapetou33]

Oct 12, 2013 16:32:30 GMT -5 hcmffm said:

Oct 10, 2013 12:47:33 GMT -5 gliptitude said:

Darrin and I have played Fortress of Narzod quite a bit during the game of the month contest of the Retrocade Magazine. Again, I've played it during the Vector War. I'm convinced that there is no problem with collission detection. You just have to hit an enemy twice and it will be killed, then. The difference (no hit/first hit) can be seen in the brigthness of the enemy, with the right adjustment of the brightness one can see the difference. Rather than changing the game to a "one hit kill" and thus changing the difficulty level, I strongly suggest to just make the difference in brightness larger.

I suscribe to this point of view.

[gliptitude]

Sorry if I implied (or said) "problem with collision detection". I understand the multi-hit concept in Narzod. It even says so in the original instruction manual, and we have discussed this before. (I believe it says something like the farther away the enemies are the more hits it takes) ... I JUST FIND IT INCREDIBLY ANNOYING.

I see I am outvoted thus far though.

[vectrexmad]

Oct 9, 2013 11:35:41 GMT -5 Rapetou33 said:

Franck, The new "spinning" explosion is quite good and gives an impress of move (less static than the original one in my opinion). What kind of explosion could you create ? Debris, desintegration with blinding aura ... ?

For those who aren't able to burn a rom or have VecMulti I made a small video for Armor Attack II.

For the period up to 0:37 secs you see the AI problem I mentioned in an earlier post. This is present from the original Armor Attack. Basically, just stay in the centre and all you have to worry about is getting hit by the helicopter. If you are skillful you can shoot down all the helicopters and attain a very high score.

When I was unlucky to get hit you see the new explosion routine Parabellum has implemented. I guess Parabellum borrows the graphics from the helicopeter rotor as I do see some similarities between the graphics.



Parabellum I notice there is no explosion sound when the vehicle gets hit?

On the whole nice improvements to make the game more solid.

[parabellum]

Oct 10, 2013 12:47:33 GMT -5 gliptitude said:

Oct 9, 2013 16:49:43 GMT -5 parabellum said:

I'm using a disassembler/tool I've coded from scratch, since all the available alternatives have the problem you're pointing out.

Would this tool be available to others?

Not at the moment ; the sources are rather confuse and mixed within ParaJVE's workspace.
I'd need to refactor & package this properly before releasing the tool (and I have no time for that right now).

Oct 10, 2013 12:47:33 GMT -5 gliptitude said:

Other ideas? How about a skull and crossbones which would begin tiny and zoom into focus?
[...]
Are you able to replace or modify ANY graphics in the game???

The skull idea would not fit well with the game's visual IMHO (all other graphics look rather realistic). Plus, in 2 players mode it would look a bit weird don't you think?

Well as I said, theoretically one can change any graphics.
However, the program stores the current rotated version of the vehicles/missiles in RAM (so that the program only computes a rotation once, when a vehicle turns or when a missile is fired) ; therefore more complex vector lists would not fit in the overcrowded RAM.

Oct 12, 2013 9:08:48 GMT -5 cNp said:

[...] often run into a bug where 1 player dies when the already dead, other player gets shot (even though they aren't meant to be on the screen anymore...)

The original program held a counter of the number of active players (1 or 2), and decremented it whenever a player was destroyed. In 2 players mode, when a bug fooled the program into killing the same player twice, the counter reached 0 and the round was over.

I've already found and fixed a bug where the 2nd player's dead jeep would still register collisions with tanks ; and by looking at the code, I think a dead player cannot be hit by a bullet/missile.

Do you know any other cases where a dead player would die a second time?

Oct 12, 2013 16:32:30 GMT -5 hcmffm said:

I have played Fortress of Narzod [...] I'm convinced that there is no problem with collission detection.

That's my feelings too.

Maybe it would be more fair if a distant hit (not killing the enemy) would increase the score.
Otherwise, I don't really see the point of weakening distant enemies (by shooting at them) since they're destroyed by a single close hit anyway.

Oct 13, 2013 18:31:52 GMT -5 vectrexmad said:

For those who aren't able to burn a rom or have VecMulti I made a small video for Armor Attack II.

I guess Parabellum borrows the graphics from the helicopeter rotor as I do see some similarities between the graphics.

Parabellum I notice there is no explosion sound when the vehicle gets hit?

Thanks for the video Ian.

You're right, the explosion graphics is derived from the chopper's rotor (as is the original) ; I've just added a rotation and a bit of randomness to the drawing's angle, scale and position.

The explosion sound would really be a plus I reckon, sadly that's the part I'm less at ease with.

Also, at 2:10 your video definitely points out flaws in the checks for vehicles collision : your jeep just goes through a tank, unharmed.
I haven't noticed any straight bug in the collision routines, I think it's just that the vehicles hitbox are too small.
Because they are rectangle-shaped (rather than square), the routines use their smallest dimension to avoid false positives ; but this clearly ends with a false negative when 2 vehicles are facing (because then, it's their greatest dimension that matters)...

Maybe I could increase the hitbox a tad ; with a bit of luck it would be enough to prevent this kind of bug (but it could also add a new bug where two vehicles passing each other side by side would trigger a collision)...
What's the best alternative I wonder? :)

Franck. 
^{(edited : typos)}