PaquoCastor wrote: ↑Sat Feb 17, 2018 17:48Intriguing. I can work design. Right now all we have is an engine right? So what sort of platformer does this community want to make? I could just say keen like, but I'm pretty certain that little design choices could easily become hot-button issues. See checkpoints, but also health system. And then there's always mixing genres. Action puzzle platformer anyone?
From what I learned peaking into other game projects, I'd say the most important thing is to design a design document. It doesn't have to be binding, but if all contributors can agree on a set of guidelines and themes, it would probably be easier to have the whole feed into itself. Just my five cents. At the very least it might prevent what happened with Freedoom, of course that started mostly with just art assets, if I recall correctly.
Awesome! Yes, all we have right now is the engine. My vision for it is pretty broad-minded; I could see bringing in all sorts of new ideas into the general framework of the platformer genre. I would love to work up a design document. Perhaps a shared Google Doc?
How about we all contribute to this document to try to come up with something we can all hold on to?
https://docs.google.com/document/d/15Dp ... sp=sharing
Mobiethian wrote: ↑Sun Feb 18, 2018 19:49What do I think? I do think it's nice that you want to come up with something completely original for DOS. Yet at the same time, I don't know why we all can't to do this same type of thing or put in the equivalent amount of teamwork using Keen. A new 2D engine, stereo music, 256 colors, the while nine yards.
KeenRush wrote: ↑Tue Feb 20, 2018 10:27I'm of two minds here. On one hand, like said by others, I find it a good idea to bring something new and well-made to the DOS platform. But on the other, I'm instantly thinking like Mobiethian: why isn't it Keen? And from that, I get to my usual lines of thinking: if I'm going to use my time and effort for anything, it should be a mod (because logically that's the only worthy platform, right?), and whatever the idea is I keep thinking "how would I transform this into the Vorticons engine without losing too much of the original idea".
For me to get involved, this should be something that I couldn't translate into Keen 1 with even moderate success. So, it would need to be something entirely different.
You guys have a good point... this is a Keen forum and we're supposed to be making Keen games, right? I've put a lot of thought into this. Perhaps this new game
could be a Keen game after all! I wouldn't be opposed to that. But I kind of feel like we're in need of some originality, that's all. I won't argue, though, if we all decide to make it Keen.
Nisaba wrote: ↑Mon Feb 19, 2018 22:54I'm not quite sure yet what I could possibly contribute but I'm truly interested in your vision and I promise that I'll see what's in my power to help. it's just that I have a project on my own that takes a good amount of sweat and tears. but nevertheless I'll be happy if I can donate in one way or another. I guess I'm good in tweaking existing stuff and add detail especially to graphic related things. English on the other hand is not my strength though, but others did already sign up for storytelling and punchlines.
that being said: count me in!
Nisaba, I appreciate it! I'm grateful for whatever help you are able to provide, even in the realm of storytelling!
KeenRush wrote: ↑Tue Feb 20, 2018 10:27* No grid. The entire level made of objects which may be placed with pixel precision. Some of the objects are completely neutral, while others have collision detection (no need for slopes for me, just basic rectangle is ok), and some are more complex like enemies or points - but the enemies and collectiables could be of a different class altogether. Objects have an order of drawing, so they can be layered and create background and foreground easily.
This is a very fascinating suggestion. I think with the way I'm going to implement objects, this could work! I could even have this sort of thing side-by-side with the default tile engine, so that either one or both could be used.
KeenRush wrote: ↑Tue Feb 20, 2018 10:27* Parallax scrolling background, maybe with two layers for more atmosphere. Should be easy to implement.
* Foreground effects like snow and rain. They would add so much atmosphere. Likewise not that difficult, and there are a billion examples of the falling snow effect from which to see how it's done. I'd really want these!!
I love these ideas. In fact, I already had it in my head to do parallax backgrounds. I never thought about foreground effects though! This could be super awesome. I don't suppose either would be difficult to implement.
Commander Spleen wrote: ↑Tue Feb 20, 2018 12:42The freedom I would most like to see would be in terms of sprite functionality. Easy implmentation of new characters, behaviours and player abilities, maybe borrowing one of KnP's best aspects and using the event-oriented programming grid. In fact, I think that is exactly the kind of feature I would recommend as a primary focus. I'd be willing to contribute to designing this but probably wouldn't be able to commit to consistent involvement or programming for the foreseeable future.
My end goal is to have a sort of universal object/sprite/enemy/etc system which has a config file for each object and maybe a graphical editor of sorts to edit that config file and set its properties and behaviors.
Great ideas. Very good input... this is the sort of thing that should probably be in the aforementioned "design document".
Commander Spleen wrote: ↑Tue Feb 20, 2018 12:42The first question is what language and library to use if it's to be built for raw DOS. C/C++ would be the most flexible but also the most complex. Allegro 4.2 (the last to support DOS) would be my library of choice here, though it would run in protected mode rather than within conventional memory.
You could write something fairly decent in QuickBASIC (such as this) but run into limitations pretty quickly. I'm pondering TurboPascal for some future projects as a middle ground to increase efficiency and features while maintaining some simplicity of syntax. It was used to program many classic DOS games and could be viable for this engine.
Roobar wrote: ↑Tue Feb 20, 2018 14:46Keenmaster486: I remember there was a quick basic project decades ago. Was it yours? I don't remember at the moment. I remember it was pretty good and felt really keenish, but it took ages to load, which I didn't like. Non the less its interesting to see some demo or something. Is it a galaxy or vorticons style?
Ah, QuickBasic Keen... nope, that wasn't me, but I certainly was inspired by it. The previous few iterations of my engine were indeed written in QB, though. My original goal was to exactly clone the Keen Galaxy engine. I got pretty close eventually:
https://youtu.be/a-3VlnlOCGM
But I eventually ran into the same limitations that the original QB Keen guy ran into (surprise!). QuickBasic is just plain very slow - at everything. Even using the DirectQB library which sped up all graphics routines significantly. Once, I did a test in which I compared the speed of various loop and conditional statements - and FreeBasic was a whopping 20 times faster than QuickBasic on the same machine, after compiling the exact same program with no modifications.
FreeBasic does run in protected mode. Right now the engine demands 8 MB of extended memory, mostly because I'm loading ALL the graphics into memory, and my allocation is all static, which means there's a lot of empty space. Probably there are better ways to do it.
I'm definitely not averse to using C/C++ or Pascal or something. My only problem with that is, I am very inexperienced programming anything but Basic-like languages in DOS. I've used C and C++ in more modern OS's quite a bit, but I find it confusing in a DOS environment. Someone would have to volunteer to help me out if the engine were to be rewritten. I should mention, though, that I do have a full copy of Turbo Pascal (which is probably what I'd want to use if we were to abandon FB, but I don't know for sure).
My engine is Galaxy-like, mostly because I used the Galaxy engine as inspiration, and stole the Keen 4 graphics and levels for testing purposes (you can see them in the demo). The philosophy behind the engine is to (hopefully) be pretty general, though, so you can make it feel however you like. Already in my new engine the physics are much faster and different than Keen, but that can be changed in a configuration file.
I'm feeling pretty good about this overall. If we can keep this up we can make something great.
Here's a link to what I currently have for the engine:
https://drive.google.com/open?id=1TjtYJ ... zzZwIEgNNf
Run engine.exe in DOSBox (make sure you have cycles=auto in your DOSBox config, or turn the cycles up to 30000 or more) and use the up and down arrow keys to select which level to run (just the Keen 4 levels for now, testing purposes). It's a bare bones demo. Just proof of concept mostly. In the IMF folder you'll find a whole bunch of music files. You can play around with editing the config file, the level files, the source code, etc.
I'd sure appreciate comments on my source code. Right now there are a lot of things commented out that need to be deleted, and a few things that are on my list to either fix or optimize. For instance, I have plans for a real main menu and to clean up the main loop code and place it in a separate subroutine.
Let's all contribute to that "design document"! I can see this being very cool if we set our minds to it.