Kurt_Woloch

Here are my times for classic games for this week: Windows / Visual PinMAME: Baby Pac Man - 66 min. Atari 2600: Chaotic Kitchen (WIP) - 11 min. I didn't play very much this week because I'm still hard at work on the Atari 7800 version of Baby Pac Man. I played the version of it on Visual PinMAME, which is a bit difficult to describe and to file, so I'm not sure if it's actually eligible here... to explain, there's a PC program for Windows called Visual Pinball which is a similar to a modern version of Pinball Construction Kit. There are various tables for it, and it also has a function to pair with another program, Visual PinMAME, whose job it is to run the original ROM's of pinball machines and connect them to the virtual table. So the table for Baby Pac Man has been recreated in Visual Pinball, but it runs the original ROM, so if you play the maze part, it's pretty much arcade perfect. Don't know how to file this... but I think I played and named this one before somewhere in the last 10 years... possibly when there was no tracker for modern games yet. Other than that, there was a new update to Chaotic Kitchen which I played, which now has got all the six screens of the original arcade version and the pepper, but still some bugs as well.

Thank you for the feedback. Just to address some of the points: @Defender_2600 + Jinks: Yes, it was probably too generous. @Swami: There's no such thing as a simple "setting" for the collision detection... in the first version I was assuming that the flippers were always ending at the same column. I amended that for the second version, but now it was basically assumed that the lower two flipper positions were ending at the same column as the 3rd one counted from the bottom (which extends the farthest and has the smallest distance between the flippers). I've amended it again now so that if the flippers really are at that position, they do deflect the ball, but this only happens for one frame... on the other frames, they don't extend that far. You're correct about the right flipper inconsistency, which actually was a label mismatch in the code on my part. @Imstarryeyed: Thank you very much for your feedback! I'd have one specific question for you... if you hold the flippers in the up position so that the ball slowly rolls towards their upper end, how fast does it actually roll? I.e. how long does it take from the middle of the flipper until it comes to a stop because it also hits the ramp to the side of it? So... here's another update. I amended the collision map for the flippers again, as well as fixing the ball sticking to the flippers and the inconsistent behavior of the right flipper. I've also tried to bring the flipper behavior closer to what Imstarryeyed described in his spoiler. The right saucer should also work better now, though there do seem to be some glitches still, especially that the ball under some circumstances sticks to the "sharp edge" of the right saucer (left and down from the "1"). I will try to fix these in the next days. Can you find any others? Other than that, I think the physics should be pretty much complete, except for the upper ball which is one big challenge left. What do you think about this update in general? Is the flipper behavior better now? BabyPac.A78

That's a nice update... there are now all six screens present as well. Also the starting position of the player is now correct. You probably didn't work on the enemy AI, but this time around I noticed that Mr. Pickle tends to turn around on a platform even if there's no intersection if it finds it's going the opposite direction it needs to go to reach the player. However, this sometimes means that Mr. Pickle never reaches the ladder that would lead to the platform the player is on (I noticed this at the 5th screen with its long platforms). With the scoring, I noticed that it's resetting at the start of each level... I tihink it's not supposed to do that.

OK, thanks for the feedback, that's what I'm looking for. Yup, the scoring for the pinball part isn't implemented yet. As for the spread between flippers, the table was designed by Bob (with some help probably) and it would be difficult for me to try to change it. What I can change, however, is the collision detection on the flippers. Actually, I'm a bit unsure how to properly calculate the boost the flippers give to the ball... I know it's supposed to go the more to the side the closer the ball is to the edge of the flippers, and that doesn't seem to work really well so far. Anyway, here's another update. I've tweaked some things. Based on your feedback I closely looked at the collision map and I made it more generous regarding the flippers... I think there's now only one or two pixel columns in the middle where neither of the flippers is able to reach the ball. I also fixed the right bumper, the right upper curve and the right wall of the playfield, though there's yet some work to be done around the right saucer. I also capped the speed the flippers are able to give to the ball, this should also improve the handling a bit. What do you think about this version? BabyPac.A78

OK, I think it's time to post a new binary. I've worked on this for the last two weeks, and now I've reached an important milestone... the flippers are now working! It's not perfect yet, though, and there are still many bugs, especially in the right half of the playfield which is still pretty much broken. This is most obvious in the right bumper, the upper target area and on the right edge of the playfield. Also, the few sounds I did include are repetitions of the sounds of the maze part, and should clearly sound much different in the final release. Also, the rollovers, spinners and targets are still non-functional, and there's no scoring in the pinball part yet, and also no way to return to the maze part. Oh, and the background color changes currently are there to help debug the game (the color depends on what object handler is being called, which again depends on the ball's position on the playfield) and shouldn't be there in the final release. Still I think this is a substantial improvement over the last binary I posted two weeks ago, so I thought I'd share it with you. BabyPac.A78

Here are my times on modern systems for the past week (October 1st through 7th)... Sorry, I didn't play any modern games this week. I'm still busy with doing the pinball part for Baby Pac Man on the Atari 7800.

Here are my times on classic systems for the past week (October 1st through 7th)... Atari 2600: Amoeba Jump - 6 min. Chaotic Grill - 26 min. in 2 sessions I didn't play very much this week because I'm still busy with doing the pinball part for Baby Pac Man on the Atari 7800. However, Amoeba Jump and Chaotic Grill received some updates by their respective authors, so I tried them for a bit and gave feedback for Chaotic Grill.

Yeah, that's a good point I forgot to mention with... I think in the arcade version it works such that the screen is made up of character cells (8x8 pixels each) and the chef can get onto a ladder or platform as soon as he's in the correct character, which gives 3-4 pixels of room each way. Translated to the Atari 2600's proportions, there should maybe 3 pixels of room at the top and bottom of a platform, and 2 pixels left and right to a ladder. Sorry, I can't give you more details about the Level 4 bug since I don't remember what exactly happened... I only remember that before it happened I had multiple enemies (at least 2) pretty much in the same spot riding down the same burger part, and then they ended up being stuck in the air.

Good job! As a veteran Burger Time player, here are the things that occured to me when playing your latest version: - The background music is missing (OK, maybe the 2600 wouldn't be able to do a good job at it anyway) - The pepper is missing (obviously) - Screens 5 and 6 are missing completely - In screens 3 and 4, the chef's starting position is wrong... it should spawn at the bottom of the screen, not at the top - There was a bug in screen 4 where a green burger part was stuck on its platform and didn't fall when I tried to get it loose.At the same time, multiple enemies were running left and right in the air between two platforms. However, after placing all parts below it, the level was won. - The enemy AI is still wonky at times, quite often they just stop in their tracks and don't know where to go. You can trigger this by being at the same scanline as an enemy, which quite often will stun that enemy (out of whatever reason). - The flickering is pretty bad. Does it have to be THAT bad? Or is that because the players are also getting used to display the burger parts? - The winning animation doesn't look quite right... in the original version, the chef raises and lowers its arms completely. Sorry for being so critical with your game...

Here are my times for the past week (September 24th through 30th) on modern games... Browser Based: Coloruid - 2 min. Tuper Tario Tros - 26 min. Tuper Tario Tros is a variation on Super Mario Bros. where you have to switch to Tetris in order to build bridges for Mario which are necessary to overcome certain obstacles. In Coloruidm, you have to re-color patches of different color to their surrounding color so that eventually the whole screen becomes one color. The problem is that you only have limited moves to achieve this.

Here are my times for the past week (September 24th through 30th)... Apple II: Raster Blaster - 11 min. Arcade: Burger Time - 10 min. Atari 2600: ChaoticGrill - 2 min. Atari 7800: Baby Pac Man - 7 min. Commodore 64: David's Midnight Magic - 2 min. Pinball Dreams (Preview) - 12 min. Intellivision: Deep Zone Gorf - 15 min. MS-DOS: Mario Bros. (VGA) - 21 min. PC (Windows 95): Mario's Cement Factory (Simulator) - 23 min. This week I played quite a few games, but only in short sessions. I'm now stuck with having to finish Baby Pac Man for the Atari 7800 since the original programmer, Bob, has at least temporarily left AtariAge and abandoned this game. Actually I only wanted to be helpful with the physics routines. Anyway, in order to accomplish this, I played some other pinball games for a comparison. Raster Blaster by Bill Budge is one of the earliest, it was made in 1981 and was the precursor for Pinball Construction Set and David's Midnight Magic. It's got pretty believable physics, but it seems a bit wonky at times and runs at a slightly inconsistent speed. I only played David's Midnight Magic for a short time having control problems in VICE. Pinball Dreams has only been ported to the C-64 in the last years, and even then it's not completed, but a preview. The pinball gameplay works pretty well, though, but it has got no scoring and no sound, only background music. ChaoticGrill is a new clone of Burger Time for the Atari 2600 which is still in the works. I helped out with analyzing the enemy movements, and for doing this I also played the arcade original for a bit. Deep Zone Gorf is an Intellivision game similar to Gyruss, so similar that someone else only slightly changed it and attempted to sell it as Gyruss. Then there are some Mario games, though not all of them belong to this tracker. I replayed a MS-DOS version of Mario Bros. which isn't bad, but not as well playable as the arcade original. Then there's also a simulator for the handheld game Mario's Cement Factory which I found when looking for the MS-DOS version of Mario Bros.

Well, I was listening to some songs lately which seemed particularly fitting in this context... This is roughly how I was feeling after Bob left the building... with the "woman" in this context meaning Bob (not sexually, of course!)... and I should now be the "star" continuing with this project. And in our context, in the following song the "distance" means the way from what Baby Pac Man is now to a finished product: https://www.youtube.com/watch?v=Cu6ze1wmSxA

Well, I would propose coding the simplest thing possible, which would be having the enemies always continue their line towards the player if lining up, consistent for all enemies in all cases, and then focusing on getting the other game elements right. Inconsistencies in the enemy AI can be tweaked later, I think, if necessary. When falling down with a bun, I would let the enemies continue in the left/right direction they were going before the fall until they reach the end of their burger, then going up until hitting the first platform, no matter how long the way up is.

Well, it's fun it I get to do it in a way that's fun for me. For me that would be doing a little challenge every day, but not too much of it... and showing off things often ,-) To that end, I've now decided to post a binary for the first time since I started working on the source code. There are some changes to the last binary Bob posted, part of which are still Bob's work. The colors in the maze part have changed a bit, and the design of the pinball table also differs notably from the last binary posted. Unfortunately, the pinball part is still far from being playable, but you can now serve the ball by pressing down and watch it go up and bounce around the table a bit. Some of the bounces are clearly incorrect, most notably it will sometimes go through the saucer on the upper left which normally shouldn't happen. It also doesn't react to the flippers yet. In fact I haven't implemented any collision detection for the right half of the table yet, but I still want you to see what I'm up to currently. Anyway, you can try it mutliple times and the ball will go different paths because there's a bit of variation in the direction it will be served. I hope I didn't step onto anyone's toes doing this... BabyPac.A78

Well... there was a bit more going on behind the scenes between Bob and me. We were trading PM's, and I was sending him multiple iterations of changes to his source code. He was actually always being very nice. However, I just tried to send him another PM, but it doesn't work anymore... it says that he has left our conversation about his source code, and he can't use the messaging system. It almost seems to me like he has left Atariage for good...which is a bit of a pity because he made some more changes (like moving the tunnel and fruit indicators) which would now be lost.

Oh dear... seems like I seriously mis-worded my last posting here. Unfortuntately, as a high-functioning autist, while being good at math, I have a hard time to see the boundaries of what not to say in order not to offend anyone. In this case it seems to me like I tried to make excuses for why my code is as badly readable or understandable as it is, but really ended up blaming someone else for something that went wrong. Bob, I want to apologize for everything I said which might have hurt you, for criticizing any part of your code (I know you told us not to when you posted your source codes in another thread), and for everything I claimed being able to do which I might not really be able to pull off because of time constraints. What would you all think would be the correct way to proceed with this project now?

Thinking about it more... probably I could have. One problem might be my extensive use of temporary variables. In the code I posted here back in June, there were quite a few variables, especially in the multiplication routines, called ResultL, ResultM, ResultH, TempResultL and so on... well, in the source code Bob has found a place for nearly every important piece of data to store, and has almost filled up the zeropage, leaving over only 5 bytes. However, he has also declared the variables Temp0 through Temp15 (all one-byte variables) which are also being used throughout the maze part. Therefore there was no room to declare additional result variables. So I rewrote the code for the multiplication and division routines, and also the code surrounding them, to use the variables Temp0 through Temp13 instead, which does make the code harder to read because it's not that obvious anymore what's the meaning of the temp variable now used. Moreover, even throughout my pinball code, I'm using several temp variables for multiple purposes, sometimes overwriting a temp variable with something different just after it has been read. This also lead to errors on my part where I wrote data to a wrong temp variable and then wondered why that temp variable didn't hold the value it was supposed to have, or I accidentally switched low and high bytes of 16-bit values. If I would have coded this from scratch, I would probably have thought about which variables of the maze portion don't need to be retained while the pinball portion is running. For instance, the position of Baby Pac Man and the ghosts in the maze don't need to be retained because Baby Pac Man always starts out of a defined position (depending on how the pinball part was finished or if it's the start of the game) and the ghosts always start out in the middle. But the code came from Bob and I tried to work within the confines given. Speaking of that, I think my assessment of free ROM was wrong... I thought I had a bit over 1 K to work with, but I overlooked the fact that the part from $9000 to $9FFF is still completely unused, so that would be 4K more, and I'm pretty confident that I could implement the remaining pinball features in this area (given the time needed to work on it).

Hmmm... could I have done anything in my code to make it easier for you to follow it?

I've played the arcade version of Burger TIme for a long time and have learned to anticipate the enemy's moves... yes, I think what Xybot67 says is correct. However, they will first walk left or right until they reach the end of "their" burger and then go up towards the next platform. As for the movement on intersections, there is one big rule... if an enemy reaches an intersection, it will generally check if any angle it takes (not continuing straight) will take them towards the player. If so, they will take the turn. For instance, if they come down a ladder and there's a platform on the right, they will take the turn if the player is to the right of their ladder. If the player is to the left, they will take the platform to the left if there is one, otherwise they will continue straight. As far as I've seen, all enemies follow this pattern. However, there seems to be inconsistencies in what they do if they line up with the player, especially horizontally. Which means that if they are on the same platform as the player, they will sometimes stay on the platform, even through an intersection, but sometimes they will still take the ladder up or down. I haven't found out what triggers this inconsistency. It might depend on which enemy or which round it is, which platform they are on or even if the player pushes down or up or if the chef is on a bun waiting for the enemy to step on it (this just happened to me). However, there seems to be no such inconsistency vertically, i.e. if the player is on the same ladder as the enemy coming towards him, they will always stay on that ladder if they can. That's the behavior in the arcade, but I found some differences to that in various ports. For instance, in the Flash version of Burger Time all enemies take all intersections available if they line up with the player either horizontally or vertically. And in the MSX version, Mr. Hotdog prefers to take platforms leading to the player vs. ladders and only take a ladder if the player is in the opposite direction of where he just came from, while Mr. Pickle prefers ladders over platforms, and Mr. Egg follows the player roughly like in the arcade. On the TI-99 version, only one level instead of two is added to the fall of each ingredient for each enemy riding it, and if the enemies ride all the way down to the completed burger, they will walk to the far right before going up again. The C-64 version by Interceptor's Micros doesn't let the enemies ride the ingredients at all, and in some versions they get killed by a ride, or only get killed if they fall to the bottom.

Here are my times for this past week (September 17th through 23rd)... sorry, I haven't actually been playing anything. I was busy taking part in programming an Atari 7800 version of Baby Pac Man, and while I've been testing it as well, you can't really call this "playing" because the part I tested (the pinball part) isn't playable yet. As of now, you can only launch the ball and watch it going through a curve and disappear. Therefore, no times for me this week... the Atari 7800 wouldn't be eligible here anyway.

Here are my times for this past week (September 17th through 23rd)... sorry, I haven't actually been playing anything. I was busy taking part in programming an Atari 7800 version of Baby Pac Man, and while I've been testing it as well, you can't really call this "playing" because the part I tested (the pinball part) isn't playable yet. As of now, you can only launch the ball and watch it going through a curve and disappear. Therefore, no times for me this week.

Well, at least that Einstein discussion is distracting from the fact that Bob and me are still silently working on implementing more and more features of Baby Pac Man's pinball part. By now I've got the deflection matrix implemented roughly correctly, but that's hardly visible on the surface. And I just added checking for the lane leading to the left flipper, even though only part of it can be reached by the ball currently. Now I'm working towards the upper left curve. To be able to test that, the ball is served at an angle that sometimes reaches this curve.

Well, the problem about the binary is that Bob said we shouldn't post any binary before we have anything to show... and what's been improved so far isn't really playable. The current binary on my side, in fact, is even a step back because I'm in the middle of implementing the reflection matrix. I think I've now coded all of it, but it usually takes about just as long to get it to work correctly. Right now, you could serve the ball, and sometimes it would hit the right bumper and then fly away at a crazy speed because that speed isn't calculated correctly yet. If the ball hits any other object, it goes right through it because the collision detection for all other objects except for the right surface of the left bumper hasn't been implemented yet. We do make improvements, but it's still slow going. Most of the changes I'm doing are "under the hood", so they don't make the pinball part playable yet. Meanwhile Bob is doing different changes (and implementing the changes I did before that), so he might have yet another binary.

Here are my times for this past week (September10th through 16th) on modern systems... Sorry, I didn't play any modern games this week.

Here are my times for this past week (September10th through 16th) on classic systems... Atari 2600: Red Sea Crossing - 4 min. Seawolf - 17 min. I had only been playing two Atari 2600 games for a rather short time. Red Sea Crossing seems to be widely available now (also as a ROM), but it's actually not that great, thus one game was enough to realize that. Seawolf (by Xype) is your typical submarine shooter, which is similar to other Atari 2600 titles like Air-Sea Battle and Missile War. Apart from that, I continued work on the pinball part of Baby Pac-Man on the Atari 7800, but I don't have anything nice to show yet.