So my previous model, the 2nd prototype, took 8165 bytes from the 27k obtainable and stored a efficiency from 8-20 averaging at 14.
It sported 2 completely different gamers with ~4 assaults every. This quantity is right here only for reference as that model was very streamlined for reminiscence utilization and a bit for efficiency.

So, present model sports activities 783 bytes for the system, 106 for a dummy background after which what you see on the desk for the states stand, stroll, duck, leap, pushed, push.

In basic,I might love to have the ability to hold every state measurement to a max 500 bytes and to carry out at 4.
There might be 4 sorts of characters utilizing this technique:
.Player, largest one in measurement at ~6kb.
.Full enemy, at ~5kb, is a clonable enemy with loads of strikes, every clone makes use of a unique set.
.Enemy, at 4kb, are distinctive bosses.
.Short enemy, at 2 or 3kb, are less complicated smaller enemies, just like the hawk and the canine.

Back on the desk, it’s clear now we have the stand, stroll and leap states going over what I intend.

Standing is tough as a result of it’s the state the pc is checking plenty of joystick combos, that could be very time consuming . This state is predicted to get greater in measurement and heavy on efficiency as I nonetheless have so as to add checks for the assaults. A drastic answer to lock it up could be to skip drawing the standing participant if he’s already there.

Walking is presently a monster in measurement and perf and may worsen when including the assaults that requires motion. While I don´t consider I can get its measurement a lot shorter, I´m sure I can get a 1/third velocity enhance by altering the primary IF layers to an ON GOTO factor. So issues are fantastic right here too.

Jumping is already fairly optimized however nonetheless wants the flykick added to it.

Now why all this deep craze about perf and reminiscence?? Shouldn’t they be in examine by now?
Yeah yeah…however see, stand, stroll duck and leap are heavy processing motions that might be occurring more often than not throughout sport, so the sport velocity is outlined right here. The state that takes longer, say 5.5, will set all different states to its.

While I might all the time like sooner and smaller, what there may be up to now is ok. At 5.5 we’d have the sport going at 13, 1 level sooner than final prototype. I consider now it’s time to do a closing cross on these basic states and work with the discovered values on forth.

edit: the right way to know the sport velocity from the chart? Choose any mixture of two states from STD to PUSH, add their perf quantity and add 2 (BASE) to the end result.

Joke apart, yeah, blue on blue does that factor.
The first stage, comprising 3 screens, is to depict night time to early morning, issues ought to use the darkest colours during which case I solely have darkish blue. It additionally suits the primary man higher as blue denims because the different 7 colours are a greater match for particular guys, like policeman, shoreman, no shirt man, and so forth. After first stage the background colours change, so it received´t be an issue. I may additionally repair that when doing closing cross on them.

And here’s a defeated warrior…discover its soul going away…no, it’s simply soiled pixels calculations.

IF G<>255 THEN IF G<>127 THEN IF G<>125 THEN IF G<>253 THEN GOTO 64

The G variable is the results of pooling the enjoyment buttons as in G=PEEK(65280). It swimming pools each pleasure buttons and returns a complete bunch of numbers like 126 or 254 for the precise button, 125 or 253 for the left button, one other couple numbers for each buttons pressed and one other quantity for none pressed. All this mess means I’ve to try this lengthy set of IFs to search out what I would like and that’s what hogs it…If I may do a single IF examine, it might be a lot a lot sooner.

So researching across the interwebs I discovered a submit by Jerry Stratton on Stackexchange the place he proposed the next deed:

…and that simply works! It does with a single IF name, IF (1 AND NOT BU)=1, so that’s actually great!
I’m not expert sufficient to know that bit operation (1 AND NOT BU) however  Jerry additionally did an article on the topic, test it out right here:
https://www.hoboes.com/Mimsy/hacks/coco/what-are-8-bits/

Bit-wise operations like that may certain velocity issues up and whereas he does an excellent job explaining it, I´m not fairly into the extent of absolutely understanding it to the purpose of utilizing the approach on the remainder of the code, however I might be fantastic with the joystick button answer.

Now again on the sport, I bought punching in, you are able to do a sequence of as much as 3 punches earlier than being again on the stand state. Punching reaches a brief distance so you must be very shut. I bought power and stamina methods occurring too. Each punch takes a little bit of stamina away and if it will get to a sure level, the participant will get dizzy. If you might be dizzy and get punched, the participant does an uppercut for two hits. I bought low kick in, may be performed 2 instances in sequence. I bought a robust punch and a excessive kick, that you simply carry out whereas strolling and eventually, I bought the flykick again in. So now the participant carries 7 strikes.

Some of those strikes are nonetheless being tweaked and a few certain wants redesigning and higher frames however I believe I’m now getting nearer to a neat gameplay. Here is a video I recorded taking part in it with associates, what do you guys assume up to now? (skip to 1:20)

You’ve actually performed wonderful work right here, and I recognize you sharing your perception and classes discovered with the remainder of us.

Just a few questions, if I could:
1.  Is any improvement persevering with on this sport?
2.  Have you tried utilizing the speed-up POKE but?
3.  Have you tried working it in an emulator and dashing up the “throttle” to make it run sooner general?
    (3b.  Also I’ve a dumb noob query right here and surprise why code working in an emulator cannot run simply as quick because the HW you are working it on, thus eliminating the necessity for a lot optimization.)
4.  Could you share your code from this sport with the remainder of us?  I might like to take it and work on it for my very own characters, presumably substituting within the bigger character design that you simply explored at first.

Again, effectively performed!

I’m resuming a sport I envisioned and commenced in my teenagers, and it is humorous/scary how the instructions and ASCII codes got here again in my head after so many many years!

I’m largely occupied with making a lo-res sport as a result of my artwork expertise are so poor and I believe going with hi-res will expose how awful I’m in that division.  Not to say that I have to finally put all these hours studying Rainbow and goofing round on the CoCo to make use of for one thing spectacular.

Thanks once more for the curiosity. If you might be into COCOs, examine the discord and FB teams, lot´s of motion occurring, specifically this month.

Thanks. I’m too previous to even know the right way to use Discord.   I did join the FB group although, though I do not see a variety of graphics/sport design dialogue which is my main curiosity.