Amiga Spring '24: 10 weeks of full-time Amiga Coding

[VladR]

My work contract ends at the end of this month (this Friday - 03/29) and I have saved up enough money to support myself comfortably for next 3 months.

I've been considering this for some time now, and it looks like stars are aligned for this to happen right now. I have finally only recently set up Dev Env for WinUAE on my PC that I bought last year.

Since I've done enough work on my Jaguar, XNA/C# and DX11/C++ game projects last 2 years, it's time to do some work again on Amiga and enjoy the creative process

This the list of projects I absolutely must accomplish in the timeframe of 10 weeks (10 = 12 weeks - 2 weeks for the family visit):

1. VRMark - public RTG 3D benchmark
2. Higgs - my C-like High-Level ASM compiler
3. Refund the six souls that pre-ordered my Heimdall game
4. Set Up my Vampire V4 for coding/deployment (flash the core, figure out networking with PC)
5. Transition from CGX to P96
6. Troubleshoot my recent OCS/ECS 4 KB EXE GFX demos running fine on WinUAE (but not A500)

Here's the Pool of WIP projects that I'll be picking from to do some additional features and/or consider releasing:

1. Heimdall
2. Star Raiders
3. OutRun
4. Demon Attack 3D (this is 95% finished)
5. New project Demo - (StarFox,WipeOut)

Of course, as you can imagine, each of those projects has a To-Do List that could easily fill 12+ months, not mere 10 weeks (let alone a small portion of those 10 weeks), as it's very easy to indefinitely keep adding features to the game/engine (been there, done that and absolutely will keep doing that till the day I die ).

Some Details:

VRMark

• Create Star Wars Trench Run Scene inspired by Sega 32X version in 3 levels of detail to account for Pistorm, Vampire and lesser sub-Vampire HW targets (e.g. 060 and 040)
• Gather benchmark data from 68040-060,PiStorm,Vampire RTG accelerators
• Evaluate how much work it'd be to have a version that would run on 030 (and eventually 020 CPUs)

Higgs

• Utilize Expression Parser for function arguments
• Function Return Values Expression Parser
• switch (var): case val: break;
• Classes

Vampire V4 Dev Environment

• couple years ago I got as far as I could with WinUAE and got stuck on inability to deploy builds (cores didn't support SD cards back then, nor networking in an accessible manner)
• BackUp deployment method: Figure out if SD cards are finally supported
• Primary deployment method: Network V4 with PC so I can upload builds straight from Notepad++ deployment script
• MultiOS Boot (ApolloOS + Coffin) and OS support for both in the codebase

Heimdall

• Implement Audio Support
• As I don't have the money to finance the features in its original vision anymore, figure out a reasonable effort estimate for lower price points (perhaps consider the $10-$15 price point)
• Prototype some additional procedurally-generated content features

Demon Attack 3D

• This is closest to release (can play from start to end) though I would like to test the new point-cloud textured planet renderer on V4 to consider real-time rotation for that HW target

I am very excited and can't wait to start

[Rob68K]

Whatever you do, pick one thing and finish it.
Too many projects seem to get part finished when interest is lost through distraction.
It's easily done. I have a lot of personal experience in this.

[VladR]

Originally Posted by Rob68K

Whatever you do, pick one thing and finish it.
Too many projects seem to get part finished when interest is lost through distraction.
It's easily done. I have a lot of personal experience in this.

I understand this may work like that for many people.

For me, the primary reason is not the lost interest, but realization that it'd be criminal to not spend a certain additional effort to make the game that much better.

In other words, if I already spent 9-12 months working on the game, I find it criminal to release it NOW when I know that 2-3 additional months could make a tremendous difference.

To me, releasing the game butchered and early is a complete disregard and disrespect for all the effort (and money) that went into it.


Now, if I was being paid my commercial rate as a Senior C++/3D SW Engineer, that'd be different. I wouldn't have a problem with that.

But if I am footing the bill, then: Sorry - not Sorry


If I have to choose between the two ways how I go into the grave when I die of old age:
1. Plenty BUTCHERED but released projects
2. Select few curated and hand-crafted projects

It's not really a choice...

Now, if anybody is willing to pay me a commercial C++ rate, I will happily do as you tell me

[Rob68K]

You sound like you have it all figured out.
Good luck.

[copse]

This is a great idea please keep us updated. Too many are likely too scared to leave the rat race.

[kriz]

Epic! Please go ahead and do share updates, may you stay strong

[Minuous]

I'm a bit puzzled why you would want to develop on a Vampire when you already have WinUAE which is faster and more compatible, it should save you some time.

[VladR]

Originally Posted by Rob68K

You sound like you have it all figured out.
Good luck.

I understand why you're being sarcastic, believe me

But, I actually, did "figure it all out" - as this is not my first rodeo. From experience, the 10-12 week coding sprees are the most efficient ones. For me. I would certainly not dare generalize this onto others.



Also, I don't abandon projects, I switch between them and long-term they all grow - which is what matters.

However, this event also fulfills another purpose - life enjoyment. I meet so many people who do not experience the excitement, joy and amazement that comes with coding video games.



I consider myself to be a - what's the psychological term here - "A Lucky F*ck" who has been granted this privilege in life and I will not squander it.

I'm embracing it and enjoying it and living my life around it

[VladR]

Originally Posted by copse

This is a great idea please keep us updated. Too many are likely too scared to leave the rat race.

Every now and then I do enter the rat race, save some money and then do this.

Originally Posted by kriz

Epic! Please go ahead and do share updates, may you stay strong

Yes. This is my thread, so I'll keep updating it. Anybody offended with what I'm doing can simply ignore the thread

Originally Posted by Minuous

I'm a bit puzzled why you would want to develop on a Vampire when you already have WinUAE which is faster and more compatible, it should save you some time.

I totally develop on Windows in Notepad++, that's not changing. I understand there's people who prefer coding on the original HW, but I am not one of them. I need my 50" OLED TV for coding and I absolutely enjoy the instant compilation times on PC.

Originally Posted by Minuous

when you already have WinUAE which is faster and more compatible.

Uhm, no it isn't. One of the projects for this quarter is figuring out why my OCS 4 KB demo runs just fine under WinUAE but crashes on real HW.
Likely, that's not issue with WinUAE, I may have certainly misconfigured it.
But more often than not, stuff works fine in WInUAE but not on real HW.


Another example with the benchmark I am working on. It crashed on 5 different HW configs last week despite working fine on WInUAE.

So, I absolutely need a way to deploy and test on real HW (even if that HW is Vampire and not some 060). Otherwise, I am developing only for WinUAE.

And I certainly want my stuff to run actual HW.

[VladR]

I spent last few evenings messing with the V4 (after it was in the box for 2 yrs), but I got it up and running again, installed Alterra tools and flashed the latest core and installed Coffin.

My keyboard doesn't work, so need to troubleshoot that.


But, the latest core does have SD card functionality working so I was able to actually deploy my builds to V4 for the first time !

I was incredibly pleasantly surprised how fast my flatshader runs on V4. I tested all resolutions from 320x200x24 to 1920x1080x24 and I find it still very playable even at 1280x720x24 ! Did not expect that, that's for sure!

I should probably dig up my capture card, install it to new PC and start recording some vids from the real HW, now that I suddenly have capability to run my own builds!

[Rob68K]

Originally Posted by VladR

I understand why you're being sarcastic, believe me

I just thought you were looking for feedback on your plans rather than it just a statement of intent to the world.

My most valuable lesson in nearly 40 years of coding is being able to appreciate the uselessness of an unfinished project. So I thought I'd share that. If we don't share the lessons we've learned from our own experiences when the occasion presents itself, then the lesson is only valuable to ourselves and to not share, to me at least, seems like selfishness.

I think if you're in a community like this, most people would likely prefer you completed and delivered finished code because it's something they can fully engage with but if keeping multiple projects running in parallel is what brings you happiness and that's what's of primary importance to you, then just do it.

No sarcasm intended.

[VladR]

Originally Posted by Rob68K

I just thought you were looking for feedback on your plans rather than it just a statement of intent to the world.

My most valuable lesson in nearly 40 years of coding is being able to appreciate the uselessness of an unfinished project. So I thought I'd share that. If we don't share the lessons we've learned from our own experiences when the occasion presents itself, then the lesson is only valuable to ourselves and to not share, to me at least, seems like selfishness.

I think if you're in a community like this, most people would likely prefer you completed and delivered finished code because it's something they can fully engage with but if keeping multiple projects running in parallel is what brings you happiness and that's what's of primary importance to you, then just do it.

No sarcasm intended.

Sharing experiences is great if it can be done without implied judgement but that's rarely possible on forums. That requires different format - face-to-face discussion and yes - those can be pretty illuminating if you can find somebody in real world close to you to talk to as then the realizations will keep happening long after that discussion.

There will be things delivered - like the RTG Benchmark. I'm not making promises on full games and no - it's not because of the 80:20 myth. That myth doesn't manifest in my case. I don't even argue with such people anymore - I could care less, honestly.

What manifests in my case is that even if I finish the features that would consider the game finished, I am absolutely unwilling to do that, because at that point it feels like if I already spent this much effort, I might as well spend additional 8-10 weeks.
Otherwise it's wasted potential and I'd rather not release it as I would regret it. And I do not want to feel regret.
I've done that multiple times. On multiple platforms.

Of course, in theory there's no end to adding features. There's always Bethesda-level feature-set that can never be matched by a single coder.




That being said, once I refund the six people that pre-ordered Heimdall, I can reduce the price point to $15-$20 and I believe I have enough there to warrant that price point with few weeks of work. But, no promises within the 10-week window.

[Rob68K]

I don't know the background to the Heimdall thing and this is the first time I heard about it.
And that's your business.
From my own perspective alone, I understand the fear of finishing and I know that comes from any number of elements which negatively affected us on a psychological level.
I think that can manifest in 'perfectionism' and that afflicts many a creative who is terrified of releasing something 'incomplete' that will receive negative feedback.
IMO, it's far better to release something 'good' or even 'average' and then move on than not to finish and release anything at all.
My own regrets are far higher for the things I didn't finish than those I did that were shit.

[VladR]

Originally Posted by Rob68K

IMO, it's far better to release something 'good' or even 'average' and then move on than not to finish and release anything at all.
My own regrets are far higher for the things I didn't finish than those I did that were shit.

That is very interesting.. You actually regret not releasing.

For me that feels like I dodged a bullet by not releasing something I deem "perfect"

Small example - the PC RPG game that is based on the game that was actually released in retail - I worked on improvements for over a year.
Then I took a break and put in another 6 months worth of engine/gameplay improvements.

But at that point, I figured I'd butcher it if I didn't spend another 10 weeks on perfecting it - because additional 10 weeks on a 2+ year project is nothing. Problem is I can always say that, forever...

But I do not regret not releasing it at all. I did, recently, spend 3 months porting it from XNA to MonoGame so that I gain access to all major platforms.
But, no regret.

[Rob68K]

Yes, we operate with a different set of priorities.

The details are, ultimately, unimportant.

To my mind 'perfect' is impossible and the idea that something has to be perfect before it's declared finished is just an excuse for self-sabotage.

To my mind. No reflection or accusation towards you. Just how I see it.

[Photon]

Sounds fun, don't forget to also have some free time in your free time

[VladR]

Originally Posted by Photon

Sounds fun, don't forget to also have some free time in your free time

My Siberian Husky gets me out of the coding recliner 5 times a day and walks me up to 2 hrs. Guaranteed.

No matter if it's -55 'F wind-chill, blizzard (we're just finishing with a 3-day winter storm) or 2.00 am

[VladR]

Originally Posted by VladR

Heimdall

• Implement Audio Support
• As I don't have the money to finance the features in its original vision anymore, figure out a reasonable effort estimate for lower price points (perhaps consider the $10-$15 price point)
• Prototype some additional procedurally-generated content features

While researching Sega Model 1 (SM1) Arcade Star Wars on YT, I got reminded why I started my flatshader on Jaguar all these years ago:

StunRunner


I absolutely already have enough features to release at $15-$20 price point and not be embarrassed about the game. I don't need high-poly gfx environments.
I just need to backport the code from the Jaguar source code.

Vampire crunches Sega Model 1 CPUs for a snack. It looks like it would still be higher framerate at 1280x720 on V4 than StunRunner on SM1.



And it's not like there's savage competition where people release their weekend-long click-throughs from Unity on Steam.

A Stunrunner-inspired game would still look great and be quite original on Amiga.

I don't yet have benchmarks from 040 and 060 but I reckon 060 should still run the code fine at 320x240x32, but we'll see once I figure out why timer.device doesn't get initialized on real HW (but works just fine under WinUAE)

[Minuous]

Originally Posted by VladR

we'll see once I figure out why timer.device doesn't get initialized on real HW

Using a Vampire is not going to reveal much about what happens on a real Amiga, as it is quite different.

[VladR]

Originally Posted by Minuous

Using a Vampire is not going to reveal much about what happens on a real Amiga, as it is quite different.

Isn't Aros/Coffin all branching off the same AOS?

In this particular case, this issue has manifested on 3 other non-vampire configs.

Most likely issue is that I am opening the device incorrectly.

I printed out all addresses of all libraries that I am using and they are correct, it's just the timer.device

One of these mornings when I get full focus after coffee I need to go through the docs and see what I am doing different.

What's curious is that there's no error returned from opening the device.