Jump to content

AnalogKid

Members
  • Posts

    70
  • Joined

  • Last visited

2 Followers

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

AnalogKid's Achievements

Star Raider

Star Raider (3/9)

233

Reputation

  1. I bought one of the Phoenix roller controller power supplies and I can confirm that it does clear up the input noise issue. I mentioned it to my Collectorvision contact. Who knows, maybe they'll collaborate with the creator.
  2. In case any other developers ran into the same problem, it turns out that if you're using the roller controller with the Collectorvision Phoenix there's apparently a problem due to the fact that your using a separate power supply for the roller controller and there's no common ground between the two devices. As I did, you can code for the issue, i.e. look for a huge jump between the last spinner value and the current, but there is a hardware solution. Someone created an adapter for the port on the back of the Phoenix that can power the roller controller. Phoenix Roller Controller power supply
  3. One of these days I'm going to get around to finishing up Missile Command. Also, I'm currently working on a game for the steering wheel, another vastly underutilized input device.
  4. Excellent use of "animated" tiles. It's so worth the extra code block for character/color transitions because an otherwise static background comes alive using relatively few clock cycles.
  5. That's good news about supporting the spinner inputs and generating the interrupts. Most emulators do not support them and those that do don't seem to be terribly accurate.
  6. I just uploaded updated ROMs with one small fix. The ricochet logic was wrong for the case where the ball is bouncing off the back wall. I remember as a kid playing Breakout on the 2600 trying to open up a hole to the top so the ball would take out the top part of the wall quickly. To download just use the links I already posted.
  7. here's the joystick version: Breakout for joystick ROM
  8. Here's a little Christmas gift for all you Colecovision fans with roller controllers. There aren't enough games that support it and I couldn't find a port of the 1978 arcade classic Breakout so I threw this together pretty quickly to get it out for Christmas. Like the original you can influence the angle by using the edges of the paddle, unlike the original the ball doesn't pass through bricks on the way back down which always seemed a little weird and besides the ricochets are pretty cool. I'm sure it's not without bugs but it's a freebie. I did also code for joysticks but it's just not the same and lacks something. If folks want me to upload that version as well, I will. Breakout ROM
  9. Hi Philipp. I've been developing games with SDCC for a few years now and there is really only one thing on my wish-list and I'd like to put it out there as it could be a great benefit for coders mostly for performance but also for resulting code size. I don't know how feasible such an optimization would be, but, it would be nice if the compiler could take advantage of a programmer ordering their structure members and pertinent code in such a way that that the compiler could avoid adding assembler for calculating offsets for every member reference when the code is using pointers. Basically what I'm saying is it would be nice if we wrote code the right way then the assembly could just INC HL and save 31 clock cycles per member access which doesn't sound like a lot but it adds up quickly if you have a large number of sufficiently complex screen elements. I know it would only be applicable for simple statements as you can't hang on to HL otherwise, but, it's worth considering. For my own games I take SDCC's output and do this optimization by hand via inline assembly, but other coders would certainly benefit. BTW, just have to say, you guys already have a lot of really clever optimizations for the Z80 in there, nicely done.
  10. It could be that I have a wonky roller controller but I was wondering if any other game developers have seen the same thing that I have. I added an interrupt handler for the IRQ that calls the BIOS UPDATE_SPINNER subroutine and what I'm seeing is occasionally there is "noise" in the values stored for the accumulated spinner velocities. By that I mean I see a range of absolute values between 0 and 20, with 20 occurring when I spin the ball like a madman, and while maintaining a relatively constant value there will occasionally be a single very high value in the mix, like between 40 and 60. Also this is occurring with testing code that has a bare minimum of code to avoid impacting the accumulated values. Obviously I can code expecting noise but I was curious if anyone else had observed this with their own hardware.
  11. BTW if you haven't discovered Gearcoleco yet, it has the best assembly debugger that I've found so far. For instance it supports setting memory breakpoints which can save so many hours of debugging
  12. From my own personal experience I found that BlueMSX is very good at revealing unwanted behaviors due to my code not beating the refresh, not beating the VDP send window, writing to VRAM to quickly, etc. Things that appeared fine with ColEm, CoolCV, VColeco, Gearcoleco, and others would appear as problematic in BlueMSX and I could confirm that behavior in both my Phoenix and classic CV. Just my $0.02
  13. When folks started reporting issues I went out and bought a vintage '83 Colecovision with the original voltage regulators, capacitors, power supply, VRAM chips, etc. in order to debug the issues. It was so vintage it barely worked when I got it home until I hit it with some contact cleaner. I was able to sporadically reproduce the reported issues and a couple things were going on, one I was being too aggressive with the VRAM timings and after 40 years not every CV has aged as well, two there was inexplicable flakiness with the value in memory for the enemies remaining count after the enemies were initialized. It was a case of being able to run the same code 50 times and out of those 50 runs, one time the result of the tally was: 1+1=3. The fix for that was to replace the incremental tallying with running a tally just before entering the game loop and also changing the check for 0 enemies to AND off the upper bits before the OR to be absolutely certain a DEC didn't roll the value back to 255. Thirdly there was a code bug that starting with a high enough wave, say 20 or so, if your player died without having shot anything at all, the game got stuck. I've had two great guys helping me test on their own vintage CVs and between the three of us I think we're just about ready to sign off on the update. It has been frustrating and disappointing as I wanted everyone to enjoy the game but I'm doing everything possible to make things right.
  14. scramble_demo3.avi Here's a very early preview of my current project, a ColecoVision port of the first side-scroller arcade game, Stern's Scramble from 1981. It was always one of my favorite games from the 80s. scramble_demo3.avi
  15. The retailer, CollectorVision, only sells physical media.
×
×
  • Create New...