Klax, Atari ST to Jaguar conversion

[+CyranoJ]

One command to assemble, one command to link... that's a 'clusterfuck' ?

 

What do you want, batch files with pretty icons?

 

 

Of course they were using AtariST / Amiga. Those machines had already proven and fully working development environment.

Those machines were the ones that coders were already familiar with very well and could produce something simple (yet still nicer than the older consoles could) very quick..

 

And you keep forgetting the dire financial situation of Atari. It made most economic sense to get as many titles as possible developed in shortest time possible. Ideally using high-level languages - be it Basic or C.

 

Just look at the development environment of Jaguar. It's a brutal clusterf*ck in 2014. Think it was better in 1994 ?

[VladR]

I did not talk about ASM dev env, since we talked about higher-level languages.

 

How many languages did Jaguar have a dev environment for upon release around 1994 ? From what I've read, not even a C compiler. It was ASM or bust.

 

For a CPU-only simple game, I don't see any reason why bother to mess to 'develop' it on jag, if you could use a slower, but infinitely more productive set-up.

 

Let's say Atari (or whoever) tells you they give you $1000 to develop some game.

 

Would you rather spend double / triple time trying to do it on jag or would you just use the dev env you already have on ST/Amiga and just test the builds (occasionally) ?

 

Developing for fun is different - time is irrelevant. If you do it for the money, then it's a whole different ballgame...

[+Gemintronic]

Just look at the development environment of Jaguar. It's a brutal clusterf*ck in 2014. Think it was better in 1994 ?

 

 

No. no! Brutal Clusterfsck was the unreleased sequel to Ultra Vortek.

[+CyranoJ]

How many languages did any other console of that time period have? Pretty much everything was written in assembler, on pretty much every major console..... (The PSX, Saturn, C++ Dev systems came in the following generation)

 

Oddly enough, back in 1994 there wasn't Visual Basic 2014

 

I did not talk about ASM dev env, since we talked about higher-level languages.

 

How many languages did Jaguar have a dev environment for upon release around 1994 ? From what I've read, not even a C compiler. It was ASM or bust.

 

 

Edited February 22, 2014 by CyranoJ

[VladR]

That's very unrealistic for several reasons:

1. You are not erasing. So you will either have to buffer the pixels under the object, or redraw them (or use another object)
2. You are not waiting for vsync. If you are even 1 cycle over the frame limit you will waste an entire frame. If you don't use vsync, you will get unstable frame rates and it'll be unplayable.
3. 5 cubes isn't 40. By your calculations if you can do 5 cubes at 100fps (which is really 10 cubes at 50fps in PAL land, not really anything to hoopla about when you put it like that....) thats 12.5fps for 40..... add in that buffering so you can remove them... and yeah I think you'll be down to around 6fps.
4. VJ is about 2x the speed of the Jaguar for tests like these... so make that 3fps.
5. The rest of the game is still required.

Now, that's not to say I'm not glad you are trying this, and I'm very pleased that VRbasic is coming along nicely and will encourage more people to play with the Jag, but please, don't go making wild 100fps claims and giving everyone false hope

 

So, I found some time yesterday evening and this morning and implemented the bottom pile of 25 cubes with some additional optimizations.

 

Here's the screenshot:

 

For benchmarking (under VJ, since my jag set-up is not complete yet), I rendered 1000 frames of those 30 cubes. It finished in 18.5 seconds, which means 1000 / 18.5 = 54 fps

I am pretty sure, I could shave off another 10% easily and cross the 60 fps threshold, but for now it's good enough.

[LinkoVitch]

Here is a picture of a goat:

 

and for benchmarking 10000*45+2-(my shoe size*the number 4)/12 = 37485.16666666

 

YAY! This clearly proves that it's the most awesome thing evar!

 

 

(LinkoVitch may just be getting a little fed up of seeing completely pointless benchmarks..)

 

progress is good, but can we stop with the completely silly benchmarking with VJ please?

[GroovyBee]

progress is good, but can we stop with the completely silly benchmarking with VJ please?

How about slightly silly benchmarking in VJ? Is that next door?

[sh3-rg]

the number 4

 

This made me happy.

 

+ the goat, of course.

 

But I like reading Vlad's VJ benches, so there!

 

Looks like you might as well just finish the game now, Vlad! And I'm happy to do 50/60 pal/ntsc benchmarking and video if you like when you're ready with a public build.

[VladR]

Here is a picture of a goat:

gbgoat.gif

 

and for benchmarking 10000*45+2-(my shoe size*the number 4)/12 = 37485.16666666

 

YAY! This clearly proves that it's the most awesome thing evar!

 

 

(LinkoVitch may just be getting a little fed up of seeing completely pointless benchmarks..)

 

progress is good, but can we stop with the completely silly benchmarking with VJ please?

Guys, you keep changing your attitudes every 10 minutes.

 

First, it's impossible to do a C code that would render the scene in more than 3 fps.

 

Now, that I get it running in ~60 fps, it is suddenly completely silly and pointless ?

 

 

Go on

[VladR]

And, if, as you say, it's so easy to get incredibly high framerates under VJ, rendering a 3D scene solely on 68k via C code (every pixel rendered is via C code), why don't you go, and spend an afternoon as I did and create the same 3D scene to prove me wrong ?

 

Certainly, it cannot pose a problem for ya, now can it ?

 

 

@sh3-rg: I didn't plan on finishing whole Klax, at all. I'm merely satisfying my curiosity of exploring the possibility of Klax 3D written in high-level language and running purely on jag's 68k without any cheating (e.g. DSP,Blitter, GPU, ...). Besides, I never even played Klax in my life, I have other jag games / projects I'd like to finish, when I get to them.

[sh3-rg]

VladR you seem to have gone all Gorf on us

 

3D for the sake of it, not finishing anything, laying down pointless challenges...

 

You know the "DSP, Blitter, GPU"? They're not "cheating" - they're what makes the Jaguar the Jaguar and not some hypothetical non-existent emulated hyper clocked 68k-powered Virtual Jaguar.

 

It's good that other people seem to motivate you to push round pegs into square holes, I'm all for that. Screw what anyone else says, do your thing. But your square peg is being driven in by Thor's hammer and the round hole doesn't really exist, it's only a figment of your computer's imagination...

 

Here's a picture of a virtual goat:

 

 

 

ps, please don't take anything too seriously, just keep plugging away as usual ;0)

 

[LinkoVitch]

Guys, you keep changing your attitudes every 10 minutes.

 

First, it's impossible to do a C code that would render the scene in more than 3 fps.

 

Now, that I get it running in ~60 fps, it is suddenly completely silly and pointless ?

 

 

Go on

 

You haven't yet posted a single genuine frame rate! All your frame rates are approximations from VJ, If it's not on the real hardware it's irrelevant! This is the "Atari Jaguar" forum, not the "Masochistic way to code a PC" forum! Why don't you post a binary of what you have written then you can get some more realistic frame rates from people who can run it on real hardware.

 

The 68K has a top clock rate of about 12-14MHz IIRC, you are writing code for one running at probably 10-20 times or more that speed, how can you expect that to be anywhere near an accurate benchmark!?

 

And, if, as you say, it's so easy to get incredibly high framerates under VJ, rendering a 3D scene solely on 68k via C code (every pixel rendered is via C code), why don't you go, and spend an afternoon as I did and create the same 3D scene to prove me wrong ?

 

Certainly, it cannot pose a problem for ya, now can it ?

 

Because this is a forum for the Atari Jaguar, not the Atari Mega STe, or Falcon. Using the actual hardware of the real machine is NOT cheating, writing your code for an emulator that runs faster than the machine is more akin to cheating, as you simply don't have that amount of CPU resource on the real deal.

 

If I were to write Klax, for one I'd use 2D, if I were to write it in 3D, I'd be using the GPU and Blitter, why? BECAUSE THAT IS WHAT THEY ARE THERE FOR!

 

I'd suggest you actually learn the system you are writing code for!

 

It's great you are writing code for it, the more the merrier.. I'd even be happy to see you finish your 3D port of Klax, but I think you are selling yourself down the garden path with your refusal to look at the real system, it's capabilities and actual resources. Why would I waste my time trying to force code to run in a way that doesn't make use to the hardware! What next, lets start writing 2D games using only the OP to display frame buffers and the 68K to do all the screen drawing and bitmasking!? yeah that doesn't sound like a complete waste of time at all.

 

IMHO you should spend your time optimising your 3D routines so that the grunt is done by the GPU and blitter, leave the 68K to do simple control. Tie it in with your basic so a user can pass points to a 3D renderer and the GPU just goes and does it even. Don't just shoe horn everything onto the slowest, oldest chip on the system.

[+Sauron]

The moral of the story is...run your benchmarks on real hardware. You should do that before doing anything else, as at this point you're not even sure your binary won't cause a RSOD when the Jag boots. VJ is all fun and games (and a godsend to the Jag community), but it's still got a ways to go before it's anywhere close to being a perfect emu of the real deal.

[VladR]

Uhm, guys - how many times have I already said that this is merely exercise in pushing C to do real-time 3D on jag's 68k (aka techdemo - proof of concept) ?

 

If I was coding entire game, then using other jag's chips would not be cheating.

 

 

Having to repeat the above countless times, now that is pointless...

 

Now, you can bitch as much as you want, but this exercise was a huge fun for me and I thoroughly enjoyed coding it

And yes, with just a little bit of encouragement, I could have been talked into spending one week to actually finish the whole game, although some of you voted heavily against klax (it being the product of '80 or sth).

 

But I get it - jag has a brutally rich game database, and countless active developers, so you're picky as to whether some stupid 3D game is even desired.

 

But, there is good news ! My second allocated-to-jag weekend expired yesterday (well, it's only Sunday morning, but I gotta go to work today), so I can go back to my next-gen PC coding, so for next 4 months there should be minimal disruptions of this forum from my side.

 

And you can focus on all other 37 currently-developed jaguar games on this forum that are tested on actual HW.

 

Win-Win

[LinkoVitch]

Glad you had fun writing code, that's the only reason to write it (unless you are getting paid to do it).

 

If you are truly "pushing C" then I think you would be better working on enhancing the compiler to give better performance from it's compilation of the C code. Writing C code isn't exactly 'pushing C' it's not like you are breaking new ground for C, it's a very robust and capable language. If you want to push what can be done with C on the jag you need to start working on improving the compiler side of things.

 

I tend to post on here only when I am not actively working on my own code/games/utils. So your thread hasn't been a distraction from any of that, no need to worry.

 

It could do with more people producing games, you are quite right there. Shame so many just seem to want to throw together 'tech demos' of things they have no intention of completing or attempting seriously, dangling a carrot and then delivering nothing at the end of countless posts about how much progress has been made. You are right, would be awesome if more people bothered to sit down and write actual games and release them for people to enjoy rather than spouting lots of noise and doing nothing.

 

One day hopefully, after all seems like there are still people willing to code for the cat, and even release tools, libraries, help, and advice to others (even if not everyone wants to make use or listen). Hopefully in time more devs will come to our cat and produce worthwhile stuffs.

[sh3-rg]

wow, it's like Seb's Removers library never existed and

 

this is merely exercise in pushing C to do real-time 3D on jag's 68k (aka techdemo - proof of concept) ?

 

...

 

Having to repeat the above countless times, now that is pointless...

 

...

 

with just a little bit of encouragement, I could have been talked into spending one week to actually finish the whole game

 

Hang on, it's not a game but it is a game and with a little rimming you would have finished it in a week?

 

And you're the same Vlad who hasn't got joypad input working yet?

 

 

Seems more like you put a little effort into making something in a completely arbitrary way and used it as an opportunity to troll Jaguar fans and other devs because that's what you get off on.

[VladR]

wow, it's like Seb's Removers library never existed and

 

 

Hang on, it's not a game but it is a game and with a little rimming you would have finished it in a week?

 

And you're the same Vlad who hasn't got joypad input working yet?

 

 

Seems more like you put a little effort into making something in a completely arbitrary way and used it as an opportunity to troll Jaguar fans and other devs because that's what you get off on.

There is (not that there ever was) no point in trying to argue with you reboot guys

 

How many times have I repeated & explained my initial goals for klax, yet you always adjusted them for yourselves ?

 

So, no - I am going to resist your trolling attempts for just another flamewar Each of the points of your last post has been discussed ad infinitum in this or other threads.

.

 

Of course, I expect from each of you at least 3 more zoofilic/quasi-funny pictures (or whatever it is that you are getting off on) so you can add at least something substantial to this discussion

[108 Stars]

Reboot is constantly releasing games. You are not afaik.

Instead of daydreaming about what would be possible and how, and losing yourself in discussions about that, why don't you just go ahead and do your Klax 3D? Then come back and show it when it's done. That would be constructive, and nobody could doubt you made it.

As long as you're only arguing with the guys who actually release games at a frighteningly fast rate you lose.

[sh3-rg]

You're not arguing with "reboot guys", most of the replies pointing out the futility of benching in VJ were from others, and they were largely not argumentative at all, but rather informative and helpful. But if your doing a .blame reboot makes you feel better, knock yourself out with that, you're not the first and you won't be the last.

 

My replies were supportive until you stubbornly insisted on ignoring every single piece of advice from people who have been there, done that and shipped the software, especially when you did so in a very arrogant and self-important way. That's the point where I formed my own opinion that your actions were no longer genuine, and that your previous and repeated comments regarding AA and its users that you made on another forum were in fact still indicative of your currently held beliefs. I don't think you came here to add to any discussion or produce something for Jaguar fans, I think you took the opportunity to engage in a little ego-boosting for very little effort.

 

If it's a discussion of substance you're really after, here's how to add something substantial to the discussion :

 

• run your code on a Jaguar and tell us how great it really is when it runs in real-world conditions, not when you click the ROM in VJ, press a button on your stopwatch and "do the math", otherwise all you're telling us is how efficient your PC is, which isn't really all that interesting to Jaguar fans or programmers looking at your work.
• rtfm. the Jaguar is great at plenty of things due to its design, the kinds of things you are potentially squandering your precious virtual cycles on by doing them manually or in a completely arbitrary way. You can retroactively say what you are doing is an exercise in pushing the 68k with C all you like, but that seems far from genuine and ends up looking more like a band aid for your lack of Jaguar knowledge.
• make some realistic and useful tests - get joypads reading in, set up a save-restore for the backdrop under your objects, introduce some animation (as the 80s original had), learn how the game you are merely half-heartedly cloning actually works and what made it so special to fans of the game.
• be realistic - do you believe you can honestly write klax in a week on your own? If so, go for it! But over-stating your ability just brings into question your other statements, especially when you're talking to people who have experience producing games on the target platform you are beginning to dip your toe into.
• consider how responsive people might be to your efforts if you quietly produce your works and only them when they better represent what you believe you are all about. you don't see typo mumbling on in little busts every few months about how fast his code is in an emulator, he just drops finished games into people's laps. be more like that and less like someone who likes to start thing, make big promises, and jump off to do something else before even seeing it run on hardware.

[+LS650]

More useless arguing in the Jag forum. Nothing ever changes, I guess.

[JagChris]

you don't see typo mumbling on in little busts every few months about how fast his code is in an emulator, he just drops finished games into people's laps.

 

 

Typo is WAY bad ass that's why.

[+CyranoJ]

Not arguing, merely pointing out the futility of the benchmarks...

 

but those 40 odd blocks (or was it 30) need to rotate down the ramp... and since your not using sprites you'll have to use rotational transformations (but you knew that, right?) - all those MULS won't slow it down at all, will they? (Just asking.......)

Edited February 23, 2014 by CyranoJ

[Goochman]

My $0.02 isnt worth even $0.02 but seriously who cares if he says it runs in 1000fps in VJ - if you disagree so be it, but to post 50 msgs on why he continues to be wrong is pointless.

 

He did something as an experiment, he doesnt have his Jag setup yet, he's poking around and testing some stuff.

 

I think at this point VladR understands his FPS are not going to be machine exact on VJ. Now can the FPS crew move on since this port of Klax isnt going to happen.

[sh3-rg]

"to post 50 msgs on why he continues to be wrong is pointless."

 

Probably futile, but pointless? Why is it pointless? Because he's realised the error of his ways? Or because we're supposed to let him continue to mislead people who don't know any better?

 

Which is more pointless, a meaningless metric or pointing out that the metric is meaningless and stating the reasons why it is so?

 

I don't know about you, but I get to a point where I stop giving someone the benefit of the doubt. There's only one reason to continue with the ridiculous claims at that point and it's to get a rise out of people, and it's obvious that has succeeded.

 

"I think at this point VladR understands his FPS are not going to be machine exact on VJ."

 

It's not a matter of a little variance in accuracy, it's orders of magnitude off-the-chart plain wrong and willfully misleading.

 

But still, that's not the real issue people have here. It's not just that they are not machine-exact, but that by their very nature they are inherently more misleading because they are claims made based on conditions that exist outside the bounds of a working game framework - a tiny loop of code abstracted from anything like working game conditions. Even if they were on real hardware they would be utterly meaningless because doing some arbitrary little non-test of one tiny fraction of what would be required within a working game environment is misleading.

 

It looks more like a case of someone trying to blow his own trumpet to an audience he relies upon to not understand a drop of what he's claiming. The thing is, while he might get away with that elsewhere, he's not going to when his audience more discerning and knowledgeable such as AA's.

 

He entered this topic when GGN made a joke about making Klax in 3D at a very low frame rate:

 

But it has 3D! You could unleash the POWA of the GPU and Blitter and make Klax 3D!!!!... and of course make everything run at 10FPS which would be a great feature for a frantic puzzle!

 

That's obviously a dig at all those games on many systems that have tried and failed hard to force a game into 3D just for the sake of it and end up failing to meet the minimum requirements for the original. Maybe he took that personally, or as a challenge, who knows or cares? But that's when he began his rapid cadence of non-game reveals and non-game-condition claims with an equal return of replies pointing out the lack of merit to the measurements. It seems that he was attempting to prove that this claim of 10fps to be wrong, and by a long way, but he's not offering anything worthwhile to counter it. And when he was called out on this he pulled the old "oh but if you'd just give me a bit of love and a few hugs I'd have taken a week away from programming my important PC project to supply you meagre fools with a whole, complete game, but not anymore! It's all Reboot's fault!!!111one". It's utter bullshit and clearly gives away his real intentions in entering this topic in the first place, anyone who falls for it is as much of a fool as the guy himself.

 

He's only rocked up at AA since js2 closed. Prior to this he never hid his dislike for thiis forum and its users, in fact he openly relished a bout of AA-bashing... maybe because that kind of thing was fashionable there in the past. Who knows. Nobody cares.

 

All we do know is he's shown up here with a lot of bravado and is having the spotlight shined on his "beautiful" offerings for very little effort indeed. Bravo.

[JagChris]

Holy crap. I didn't mean to start all this