Rocks'n'Diamonds

Holger,

I was just reading a post where Filbo pointed out to another user that this is a one-man project. It made me wonder, if something happened to you, would another person be in a position to carry on from the project from you? i.e. not only have enough knowledge to be able to carry on, but also have the passwords for github and your own website and the forum etc, and also, legal authority? In other words, do you have an heir, just in terms of RnD and your other free projects? Or is there a process baked into github whereby an orphaned project can be 'adopted' by someone else? - even that wouldn't be be enough, would it, when one considers how much is on this website.

Hope I don't sound avaricious, I am not thinking of myself in this capacity. I'm actually about to reach out to the vendor of an Excel add-in that I use and ask similar questions, only that is a paid product, so I have (I think) more right to ask him such questions. I am very appreciative of the fact that this project is free, and gives its users a lot.

John

All of the source is visible, so anyone could create a fork -- all legalities aside.

./COPYING is a copy of GPLv2. So the legalities are something along the lines of 'you can copy and modify it, and distribute modified copies, as long as you always make your version of the source available'. I am not a lawyer.

There are several RnD 'forks' extant, mainly done in direct cooperation with Holger, like some of Alan Bond's & Jürgen Bönhagen's works.

Basically -- if Holger were to disappear, it would be possible for others to continue his work. Whether anyone would is an entirely separate question, of course. It seems to me that it is rare (though not unheard of) for a project to be taken over and subsequently continued in the same spirit and manner as the originator. It is more typical for a project to either end with its creator, or to develop a substantially different tone and trajectory under a new director.

None of which should, we hope, be relevant any time soon...

(The project would be inconvenienced if, for instance, these forums suddenly disappeared because the artsoft.org domain name went unpaid, or the hosting hardware or its power or ISP connection were to be shut down -- so there are some sensible legacy questions about those things. But these seem like minor inconveniences compared to the containing context of losing the vision and coding of Holger himself!)

I don't think that this is "an un-nice thought", although I know what you mean, of course.

Instead, it is a perfectly valid concern regarding each "one-man project". I am aware of this, and some (but not all, unfortunately) of my decisions regarding R'n'D hopefully reflect this.

filbo has already pointed out some of these things, and hopefully has already answered some of your questions and concerns.

Here are some more answers and thoughts by myself regarding this topic:

It made me wonder, if something happened to you, would another person be in a position to carry on from the project from you?

No and yes.

With "no", I mean that I know of nobody who is even remotely as interested in R'n'D and its legacy as I am, or would be willing or able to put the same amount of time and dedication into it.

That said, there are quite a number of people out there which are generally interested in keeping those games alive that also R'n'D cares about: There is Supaplex Online (based on the C code of the Supaplex engine in R'n'D, where you can play myriads of Supapley levels), then there's Krissz Boulder Dash (although "members only", unfortunately, but you can simply apply for an account and play ten thousands of Boulder Dash style caves -- or play all those caves directly in R'n'D), or the Emerald Web (where you can play ten thousands of Emerald Mine style caves with the game engine that was initially used for the EM engine in R'n'D). And there are still more fans of these games out there than I would have expected (like in the Boulder Dash Forum or the Supaplex channels on Discord related to "Supaplex Online".

So all those games will live on, even if R'n'D might fade away one day.

And then there is another, direct "yes" to your question, of course:

Everybody with internet access can immediately get the source code and all assets of R'n'D, fork the project and carry on with my project just as he or she likes (as long as he or she complies with the GPL license, which should be no problem at all for any open source developer).

Here it is (on my own web site, and on GitHub):

- https://git.artsoft.org/rocksndiamonds.git
- https://github.com/ArtsoftEntertainment ... amonds.git

i.e. not only have enough knowledge to be able to carry on, but also have the passwords for github and your own website and the forum etc, and also, legal authority?

To clone or fork R'n'D on GitHub, you do not need a password, and you are legally free to fork or use R'n'D, because it is Free Software in the sense of the GNU Public License (GPL), which gives you a lot of freedom to do nearly what you want with my project.

I agree that it is a different issue with my website and the forum (and the high score server). And indeed I should care a bit more about what happens to the web site and the forum if I die tomorrow!

Until then, you can easily, recursively download the complete web site and / or web forum using "wget" or "curl". That's indeed what AI scrapers did the last few month without interruption (and in a stupid way, fetching the same pages over and over again), but I think it is very legitimate to do this with any web site you like and fear that it might disappear without a trace tomorrow -- in fact, I did this many times, and many of these web sites disappeared forever since then.

In other words, do you have an heir, just in terms of RnD and your other free projects?

No, at least none that I know of! That's why I make as much as possible of my stuff publicly available on the web.

Or is there a process baked into github whereby an orphaned project can be 'adopted' by someone else?

As mentioned, you can simply fork it!

Hope I don't sound avaricious, I am not thinking of myself in this capacity. I'm actually about to reach out to the vendor of an Excel add-in that I use and ask similar questions, only that is a paid product, so I have (I think) more right to ask him such questions.

You may ask as many questions about R'n'D, life, and the universe as you like, and I will do my best (as time permits) to answer them.

I am very appreciative of the fact that this project is free, and gives its users a lot.

Thank you!

And here is another random thought about what might happen to R'n'D once I am six feet under (although you really don't have to wait for that ):

Most of the time so far I was convinced that nobody would bother to ruin one's day trying to read or understand or change or extend that over 30 years old, nasty, rotten source code of R'n'D. But these times are about to end, just as I write this sentence!

If not already today, then surely tomorrow, you will be able to just open a browser and type:

"Hey dude, grab the R'n'D source from GitHub, and create a package ready to install for <my favorite operating system>!"

And a few minutes later, you will be able to press a download button and install R'n'D right away that refused to run on your shiny new computer, where the last available "official" R'n'D package before its author died refused to work.

Or, when you're at building and packaging R'n'D anyway:

"Wait, before you do that, please add the "plant" element from the EM game engine also to the R'n'D game engine, and make sure that it interacts with other game elements in the R'n'D engine just as it does with other game elements in the EM game engine! Oh, and while you're at it: Could you please add a nice title image and title tune for the game that matches the style of the in-game graphics and in-game sounds and music? You know, pixel art, chip tune, and all that. Thanks!"

That should work just fine in the very near future!

P.S.: R'n'D already contains its first AI generated patch. Unfortunately, it does not fix the bug I hoped it would fix.

P.S.: R'n'D already contains its first AI generated patch. Unfortunately, it does not fix the bug I hoped it would fix. :(

Huh, pretty much exactly the experience I expect with AI coding...

Holger wrote: ↑Fri Jun 19, 2026 4:05 pm Or, when you're at building and packaging R'n'D anyway:

"Wait, before you do that, please add the "plant" element from the EM game engine also to the R'n'D game engine, and make sure that it interacts with other game elements in the R'n'D engine just as it does with other game elements in the EM game engine! Oh, and while you're at it: Could you please add a nice title image and title tune for the game that matches the style of the in-game graphics and in-game sounds and music? You know, pixel art, chip tune, and all that. Thanks!"

That should work just fine in the very near future!

P.S.: R'n'D already contains its first AI generated patch. Unfortunately, it does not fix the bug I hoped it would fix.

Wow, that sounds amazing. Of course it leads to a lot of thoughts:


1. It seems that with AI, the technical side of things becomes less daunting, but conversely, the requirements of marshalling and management and good decision-making (in terms of what the user/developer asks for) will become more daunting.

An analogy might be: in the 1990s and 2000s, as China took over more and more of the world's manufacturing, the cost of manufactured goods fell, causing deflation in the West, but in response (I'm not sure of why) the cost of services provided in the West rose, causing inflation in the West. Or so I recall reading.

In other words, it will be a bit like in I Had Trouble Getting To Solla Sollew; the troubles never disappear (in total); rather, old troubles go away, and new troubles arrive in their place.


2. AI could make every Tom[-fool] an editor/contributor; for example, I might decide to contribute a bunch of new elements. Won't responsibility for accepting or rejecting the changes fall on your shoulders, Holger? At first you might be minded to accept changes that seem benign, because after all
it is nice to no longer be the lonely developer. But after some time, you realize:

- the source code has doubled in size, with many variants of algorithms for movement, transformation into other elements, etc
- likewise the graphics files
- and the executable
- someone has increased the maximum level size to 512 by 512, and wants the release to include a sample levelset of 100 such large levels - I think they would be 16x the size of 128 sq levels, which are about 33 kb, so ~528 kb?
- the available namespace [numberspace] for new elements is dwindling fast, as the highest byte-pair number heads rapidly towards the maximum (which I had assumed was 9999, but now I am not so sure)


3. Version control and knowing who did what and why will become even more critical. Domain knowledge of github will become/remain critical, to keep multiple AIs from running amok.


4. I might continue to spend a lot of time speculating on enhancements etc, but from now on in 'AI speak'. Can you help me get started by, from time to time - or just once - posting to the forum, details of how you went about making an AI-mediated change? Eg the name of the tool, the text of the back-and-forth of the conversation, fish-hooks and so forth?


5. I can't believe that the limits of this approach won't be hit pretty quickly. For instance I am interested in the idea of saving level state from a played level back to the source level, or better yet, to a derived series of files that are 'spun off' but work as level files in their own right. I find it hard to believe that I will just be able to tell Claude, or a free Google AI agent, what it is that I want, and that it will be able to get there, without my having domain knowledge of the source code, file formats, etc.


Interesting times lie ahead!

John