Jump to content

playsoft

+AtariAge Subscriber
  • Content Count

    581
  • Joined

  • Last visited

  • Days Won

    2

Posts posted by playsoft


  1. It looks like a bug to me as you now have it crashing at the same point in 2 different scenarios. In Altirra with collision detection disabled the ROM image is unchanged, the collision detection registers simply don't report collisions. The hack tested on real hardware simply loads a 0 instead of reading the collision register.


  2. I hacked the 5200 ROM image from Atarimania to stop it reading the PM collision registers and it still crashes at the same point (the phoenix, not the spaceship) on real h/w:

     

    crash.thumb.JPG.3362f6b419f15ecb1b1bdbb279a91d1f.JPG

     

    As there's no warp speed available on real h/w it does take over 11 minutes to get there, so maybe no one's ever got that far. Or perhaps there's a version of the prototype out there without this issue. I see you can buy a 5200 cart of it from the AA store, does anyone know if the ROM image used here was made publicly available?

     

    From the debug log in Altirra it's going out of bounds in a jump table. I imagine it will be tricky to fix given it's OK the first time round.

     

     

    Xevious_hack.bin

    • Like 1

  3. 9 hours ago, pixelmischief said:

    Thanks for weighing in Mr. Fish!

     

    Tell me.  Why do people still adhere to such a, now, unnecessary limitation.  I get that it was necessary "back in the day"; to maximize adoption.  But there's no need to live with ANY improvable aspect given that there are no commercial concerns and people are doing this for the love of it.  I wonder noone has done a disk version that fixes everything.  Or a 32K cart.  Or something that requires 64K.  I can understand loyalty to the original 400/800, but not to the extent that something both excellent and beloved could be made moreso.

     

    8 hours ago, MrFish said:

    Keep talking, and maybe Playsoft will get interested enough to revisit the hack...

     

    The hacked version is no longer bound by the 16K ROM/16K RAM limit but hacking the game from single buffer xor sprites to double buffer fully masked sprites is a bit much for me. It could probably be done by the original author though.


  4. On 2/9/2020 at 11:43 PM, Jaden (JRH) said:

    That Galaga 5200 ROM is extremely impressive for being just a demo. If Playsoft is going to finish it, then I would love to do the music for it. I put together a small sampling of jingles I covered with Raster Music Tracker if anyone is interested. I hope to see this project finised!

    Galaga.zip 2.42 kB · 7 downloads

    Thank you. It's really nice to see someone being proactive like this. They sound great and normally I'd jump at any RMT music on offer, but in this case I am up against it with both CPU cycles and memory.

     

    The RMT player usually needs 1.5K for its code and 1K for data tables plus the space for the RMT files. Your files are about 2K so that's 4.5K in total. Although you don't get to hear much of it, the demo includes most of the audio converted from 7800 Galaga (typically the sound effects that I couldn't convert, the enemy death sounds, are those which feature most prominently). Attached is a xex, you can step through the music and sound effects with the trigger. Your music is better but the whole package here comes in at 1.8K and will require less processor time to service.

     

    If it does progress further then maybe I could run things by you to see what can be improved. I probably won't get around to looking at this again until later in the year. In the meantime, if you are interested in doing some RMT music, I am currently working on something that needs some good music. Let me know and I'll send you details so you can see if it's something you'd like to do.  

     

    music.xex


  5. 21 hours ago, Gunstar said:

    Does anyone have a 90K ATR made of the latest revision? I found an ATR on Homesoft's site, but it's 180K, and the other sites and here are .xex's. I want to make a 90K floppy disk of the game for my 800 and 810 drive. I use 1050toPC and APE's Prosystem. At the moment I don't have another way to get new content on my 800, and that's easier that re-connecting the 1050 to my 800. Temporarily, until I get my Incognito installed. I don't care if it's a single auto-boot image or a menu disk with other files, just 90K ATR. I can get there other ways, this is just the only way without having to swap hardware around between my 1200XL and 800. The 1200XL is down for repairs, so I can't make my own ATR with the .xex file on it.

    Hopefully this is what you are after. I made a 90K disk image in Altirra, put DOS2 on it and copied the XEX.

    updated.atr

    • Like 5

  6. 27 minutes ago, biobern said:

    What might still be done: The bombs that leave holes in the ground could be visually distinguished from those that do not. In the arcade version, the bombs that leave holes are small flickering rhombuses.

    If it knows that at the point of launch then it's a possibility. I have some other things I want to do, but will take a look later in the year.

    • Like 1

  7. 9 hours ago, pseudografx said:

    Some bug reports here:
    There's an issue with the console keys on the starting screen - when you press one, you have to wait about a second before pressing it again, otherwise it won't be registered. Same applies to the J/C/D keys, only the delay is a bit shorter. Also, attract mode suddently starts even when you were pressing the keys. I think correct behaviour would be to reset the attract mode timer whenever a console key (or J) is pressed.

    They had delays of $11 for the console keys and $0B for the keyboard keys (presumably in frames) during which another key press was ignored. I don't know why they chose these values. I changed them both to $04 and gave it a quick test on both NTSC and PAL machines. It seems OK, no double presses, so see how you get on with it.

     

    In the previous update (post #171) a change was made for it to automatically cycle between the title screen and the demo. This has a side-effect of stopping it going into attract mode because I did that by pretending keys had been pressed.

    updated.xex

    • Like 10

  8. 14 hours ago, eightbit said:

    This is bar none my favorite version of this game for *anything* (that includes the arcade). So fluid, so responsive, great graphics, sound and attention to detail. Thank you!

    Thank you! Although it's been mentioned before I'm not sure I can take credit for anything aiding responsiveness(*). In making the aliens smaller I was able to remove some code that set the size registers, but I doubt that had much impact overall. TIX's great graphics make everything seem better but most of the positives were already there from the original team, just that they were restricted to 16K RAM so it didn't shine as much as it should have.

     

    (*) For PAL users, having the music at the correct speed, instead of playing too slowly, may make it seem so.

    • Like 4

  9. On 1/17/2020 at 4:52 PM, patjomki said:

    I'd vote for a buggy that looks like in the title screen (single wide instead of double wide).

    The buggy on the title screen uses all 4 player graphics. You can't use them all for the buggy during the game because they are needed for other things (the bombs being dropped at you, the explosions, the rolling boulders).

    • Like 1
    • Thanks 1

  10. On 1/16/2020 at 11:37 AM, biobern said:

     

    Good question. Small things. 🙂

     

    - The "time to reach point x - sorry no bonus"-interruption is about 7 seconds in the arcade, about 11 seconds at the Atari. Much longer if you reach a bonus. Too slow if you are a moon patrol speed runner IMHO. ;-) 

    - Musik resets and starts from the beginning if a buggy-goes-uphill-sequence starts. Is this an easy fix?

    - Maybe demo mode could start automatically between title screens? I absolutely love arcade-like attract modes.

    - Flickering 8x8-pixel square on the right side of the screen when the buggy goes uphill.

     

    Very small things, I know. It's already great.

    Bern

     

    I didn't do the uphill music because that looks too tricky but everything else seems OK.

    updated.xex

    • Like 8

  11. 21 hours ago, ascrnet said:

    Hi playsoft,

     

    Thank you very much for adding Joy 2B+ support in version 4.🥳

    Regards

     

     

    You're welcome. I hacked my Mega Drive pad for 3 buttons so I will now be supporting it in the future. One question, if I only support 2 buttons should I still label it as Joy 2B+?

    • Like 2

  12. So you made everyone else play the harder version! 😀

     

    No level data was changed period, so it will play however the original did. You can look at it with this, a hacked version which is generated to test everything out. It still performs collision detection but it won't destroy the buggy so you can make your way through as many rounds as you like (it does cause some glitching with the buggy at times).

    cheat.xex

    • Like 1

  13. Hi Jason,

     

    Not sure I can help as it was OK on my 130XE, booted from cold with Option held and then binary loaded from DOS 2.5 via SIO2PC. It is also OK when I load it from the UNO cart.

     

    I don't know anything about mydos or mypicodos. I have a 810 which I will probably set up after Christmas for something else so I can try a real disk then, but I can't see it making any difference.

     

    Paul 

     

    mp1.thumb.JPG.2a85c8b54555475514f33478c3bebead.JPG

     

    mp2.thumb.JPG.735bf1680ab6f4cb46e0f36ff1349a23.JPG

    • Like 1

  14. Nothing in that area was changed but unless I am misunderstanding what you are saying it seems OK to me.

     

    Beat the default record:

     

    mp1.png.3c9a87675fc40d823920d040be74998e.png

     

    Points get added (2350 + 1900 = 4250) but not until after it prints "You have set a new record!":

     

    mp2.png.163ea178894a57fc8fd8233ed6f9ae47.png

     

    Play again and beat the new record:

     

    mp3.png.c7d96c376379868d03fefc2c54a12b36.png

     

    Points get added (2150 + 2900 = 5050):

     

    mp4.png.c06011f4f99bada4be946a99e96d5e0c.png


  15.  

    On 12/15/2019 at 1:54 AM, Rybags said:

    As it stands this tech demo if made into a full playable game would become at the least second only to the NES home version of Galaga.

    Obviously missing are the collision detection, swarm pulsation and ship capture/dual ship capability.

    I have to say what we have here at this stage is almost beyond what I'd ever have expected the 5200 to be capable of (noting that there's only 16K RAM to work with which puts a limitation on how many rendering speedup tricks you can put to use)

    There's no ship capture/release but if you haven't already found it, press 2 for dual ships.


  16. I was able to turn my Sega Mega Drive control pad into a 3 button joystick by (1) swapping around the wires from pins 5 & 7 so that +5V is taken from pin 7 on the A8 (2) removing the 74HC157 multiplexer (3) making the appropriate jumper connections between (what were) the multiplexer inputs and outputs.

     

    Note this is an original Sega Mega Drive 3 button controller model number 1650 made in China. I also have a model number 1650-50 made in Malaysia but that is different inside, the wires would be harder (for me!) to swap and the chip is surface mount.

     

    This is what it looked like before:

     

    before.thumb.jpg.3a794cfef09c36a74b80d8ccb8f3fbad.jpg 

     

    And after:

     

    after.thumb.jpg.ebbb69b3318c022d63ecf26c5ef99350.jpg

     

    It may not work with all the JOY 2B+ hacks because I get small values when buttons B and C are not pressed (5 and 3) and those hacks may be looking for 0. However it does work with the 2 games that I have done, Scramble and Moon Patrol Redux which support Sega and/or JOY 2B+ (since I am more forgiving with my checks).

     

    Edit: Just tried the hacked version of Defender and it works fine! As does the hacked version of Dropzone.

     

    • Like 1
    • Thanks 1

  17. 8 hours ago, biobern said:

    Back from Sillyventure in Poland. Moon Patrol Redux was an "inofficial thing" there. Some people enjoyed playing it during the festival. Thank you Playsoft and Tix for the incredible work!

    Yes, many thanks for the report 👍

    • Like 2

  18.  

    16 hours ago, JAC! said:

     

    Hi Darryl,

    thanks for the RMT version. I found a small incompatibility in the hack. It assumes BOOT?($09)=$01 which is not guaranteed if the program is loaded from devices other than a DOS boot disk.

    If possible, you should not use the memory locations $8/9/a/b/c/d at all as they have predefined meanings and usage in the OS and esp. the RESET process.

    Bst regards, Peter.

     

    It's in the original RMT version by Fandal and Kjmann and seemed to cause a problem with some xex loaders (e.g. Ultimate Cart) but not others (e.g. Atarimax). I was not sure what it was trying to achieve by using BOOT. From what I remember there were at least a couple of instances of it so the simplest fix seemed to be to set BOOT to 1 by adding it to the XEX just before the run:

     

    On 11/16/2018 at 8:55 PM, playsoft said:

     

    Same hack applied (so that it runs from the Ultimate Cart correctly).

    DK_A8_RMT_$09.xex 26.87 kB · 243 downloads

     


  19. On 11/26/2019 at 12:46 PM, Tezz said:

    Hi Paul, I've continued to put some thought into Galaga this week. The disassembly of the arcade ROMS online is only partial so I intend to put some work into identifying everything in the original and decide on a plan of action for pulling it off in the best way possible on the A8. Perhaps some collaboration would be good going forward with it?

    Hi Tezz, I am full steam on my New Year Disk submission at the moment and will be until the end of the year. After that I'll PM you with details of everything that I've done and we can see if there is any overlap in what we are doing.

    • Like 4

  20. You can read about the sort of things which are possible with the expansion port here:

     

     

    As for breaking games down in that manner, that's what I did with AtariBlast! for the Ultimate SD cart. However, it is one thing doing that for games that you wrote, because you have the source code and know how they work. Doing that for other games for which only a binary exists would be 100s of hours of work for a programmer. Just to disassemble and fully understand a relatively simple game can take many months. Even then, it may not be possible. If say a game uses a bitmap graphics mode it may need 8K RAM for the screen. If it uses double buffering, so it shows one screen while it constructs the other, then it needs 2 x 8K RAM for the display - and that is all the RAM on the 5200 used up.

     

    Even if there was some sort of internal modification available for the 5200, which gave it 48K or more RAM, you would still need programmers to port the games. This would still be a significant undertaking for old games where only the binaries exist. What it might achieve is more new A8 games being made available on the 5200.


  21. Mark, that explains why it's not going right (98 < A4). From the Caverns of Mars 3 topic, Ken used the same thresholds in Beef Drop so you shouldn't be able to go right in that either.

     

    Ideally you should get 00 for left/up, 72 for centre and E4 for right/down. Not many old controllers will be perfect, but your values are a fair bit out. I think you might need to adjust the POT inside your 5200 console to calibrate the controllers - have you ever done that before?


  22. Is anyone else having this issue? I have 4 x original 5200 controllers and the game works correctly with them all.

     

    Mark, I remember writing a controller test program, searched for it and found it here in post #19. You were also having the same problem with that game!

     

     

    I don't know if you ever ran the test program, but could you please try it and see what values you get (I've attached it again).

     

    Run the program and hold the joystick right, the first hexadecimal value is the horizontal POT and should give a value like E4. The game needs to see a value of A4 or greater in order to move right.

     

    cal.bin

    • Thanks 1

  23. On 11/21/2019 at 12:05 PM, biobern said:

    Sorry, I just see: You already did something! The ugly overscan area at the left screen side in fact is _smaller_ than it was at the original Atari module! But still visible on all of my CRTs. PAL and NTSC.

     

    Cheers

    Bern

     

    Yes the screen was put into wide playfield mode in order to reduce the overscan areas. My LCD is pretty good but both my CRT and LCD TV still show some on the left. You can see mine here:

     

     

    There is discussion about overscan here and a possible hardware fix: 

     

     

    In order to "fix" it in Moon Patrol you would have to drop one of the colours. At the bottom of the screen where the buggy is, the four normal playfield colours are green, yellow, light brown and dark brown. The inverse colour is white for the missiles. In order to keep the background colour black you would have to sacrifice one of the normal playfield colours; you need to keep green to blend in with the mountains/city so I guess dark brown would become black. 

    • Like 1
    • Thanks 1
×
×
  • Create New...