The SolemnSky Project

[MagneticDuck`]

(This is the old, dusty solemnsky thread. Check the new, sparkly one for fun and profit: http://www.altitudegame.com/forums/s...760#post196760)

[Amara]

Well I had fun playing the demo! And it'll get even better once you incorporate mouse control, right?
Right?

Alti2de is the vision of a brighter future that I think everyone in this little community has had at least once and just as everyone else I'd be very excited to see it become reality. Right now our future is in your magnetic hands, duck.

And I'd gladly offer you some assistance with anything art/design related, just note that I'm yet an artist/designer in the making, am inexperienced, have a high procrastination/unreliability rate that rises the more projects I get and I'm operating a system that is on its last dying breaths but the upside is I can draw cool cartoon planes!

That being said, here's my application with pencils of some kind and alti works I've created so far:

http://altitudegame.com/forums/showthread.php?t=9659
http://altitudegame.com/forums/showthread.php?t=9750



Not all are pencils tho

[MagneticDuck`]

You're hired! Your salary will consist in silent respect by a potential future community, and will be supplemented with invisible imaginary internet points. Your job description will be do "do .. you know, art stuff" and your working hours will be "approximately sometimes".

The game demo at this point is just essentially me saying "hey, you can actually make a responsive game with javascript and the game engine I'm building", don't worry about that.

And yes, I'll get around to mouse support.

[Moon]

Any particular reason why you chose js?

[MagneticDuck`]

1) People love editing javascipt. It's a popular and (generally) easy to understand language, so we can expect people to be able to fork and develop our source.
2) People love running javascript. Things that you can click on and immediately start working are able to get more public accessibility.
3) Javascript is a viable option. WebGL makes fast graphics right in the browser possible, and a native client with a bit more power and UDP is feasible with minimal effort using a system like Electron.

I'm not a big fan of javascript as a language, but it seems like an interesting option for our project, at least as a way to get started. Javascript will make the game easy to produce and distribute.

If later, after releasing and marketing solemnsky to the public, we find ourselves in the position of having a lot of people interested in the game and asking for something more, we could of course move on to make a native client setup with the performance benefits that it would bring.

Fun websites can spread like wildfire.

[wyclef]

Have Karl and Lam given you their blessing to move forward with this?

[ABlank]

Wasn't there some sort of browser demo version of Altitude on the old website?

It's still there at http://altitudegame.com/playnow but i couldn't get it to work.

[Mango777]

Magnetic, is there a way to add a debug mode into this demo? Like in Alti? Like velocity indicators, camera coordinates, etc.

[Karl]

my computer couldn't handle placing anymore packing peanuts to dive into

[MagneticDuck`]

Status update!

Here are some facts about solemnsky as it stands:

• A lot of people have expressed interest in the solemnsky project, and we already have around 10 people in our skype chat group. Add me on skype as 'magnetic.duck' if you want to join in.
  
• I've been fairly busy lately, but will have more time in the near future, especially during the school year. I am committed to making this project happen, well aware that it will take a lot of work both in coding and design.
  
• I have made the decision to move from javascript to python. No more web client; although that was a fun idea and could have made getting new players in the game fast easier, performance issues (especially with networking) has discouraged me from expecting a web client to become a competitively playable multiplayer game. Python is new choice because it is fast enough (at last resort using something like cython), has friendly libraries, and is a language I can hope to expect collaboration in; it's easy to pick up and quite a few people are familiar with it (VAN1SH1NG's code is written in python.)

My current goal is now to build a multiplayer game demo on a scalable infrastructure in Python.

Longer term goals, more easily contributable to, include imagining the gameplay of solemnsky (Emi is having fun with this), and deciding the aesthetic atmosphere (is there an artist in the house?!). As important as it is too easily dismissible we should also consider the design of the important social infrastructure that ideally will incentivate a healthy competitive community but in the unfortunate case of altitude are directly responsible for the wide gap between competition and 'the pubs'.

This will be a long project but as long as I stay at a computer and do not come up with a very, very good reason to not proceed, I am dedicated to seeing it through.

-Duck

[MagneticDuck`]

Status update.

Not a lot of visible work on solemnsky has been done since the last status update; Python has been abandoned on the grounds that it's just too slow to natively implement the whole rendering / physics / gameplay / networking structure (things like PyGame work by linking against C libraries, and I'd rather not have to use such a system; writing games in python is more of a compromise than a fully defensible strategy), and also because I have found a bunch better option: haxe.

Haxe is a high-level platform-agnostic language that transpiles to all kinds of useful targets including HTML5 / javscript (with WebGL), native executibles via the cpp toolchain, and Flash. It trumps both of my previous choices, javascript and python, for being simultaneously a well-designed language with support for native releases (which javascript lacked) and a fast one, while also potentially making a web client possible (which python lacked). Porting the small game engine / mechanics code from the javascript codebase should be relatively easy, and haxe + its various toolchains has already been battletested in a large number of indie titles.

The only thing that awaits attention is time on my schedule to make this happen; I will begin to find more time to work on this as my part-time (suprise, I do programming) / other things stabilize. I haven't forgotten about solemnsky, don't worry.

If anybody out there believes they could be useful with the development, I would love to talk to you, send a PM here on the forums or a contact request on skype (I'm 'magnetic.duck').

-Duck
solemnsky

[Mango777]

Lol... this sort of reminded me of something I saw on another forum (KSP forum). A former dev decided to make his own game, and he kept switchingthe engine or something... He originally used unity, then he switched to something called panda, then back to unity... I hope this doesn't happen to this too...

[MagneticDuck`]

I can appreciate the naive concern that the superficial reports I have given may present: that my indecision and apparently unproductive changes of mind will continue into the future. However, I believe that these concerns, upon adequate evaluation, are invalid. It is not from indecision that I draw my changes of mind, but rather from objective decision in the interest of the success of this project; an objective decision that does not take into account the outward appearance of my progress; a decision that holds in utmost importance the choice a capable development infrastructure that can be depended on to not let us down in the future.

My choice of Haxe now is more credible than my choice of Python several weeks ago because now I know what to look for in my language, having failed at it twice.

To reiterate: I do not 'hope' that I will not continue stumbling, I try my best to have certainty of it. It would never cross my mind to not abandon a sub-standard development path with the objective of outward appearance, just as it would never cross my mind to shelter from the reality of the effort that is required to create alti2ude from the ground up in a truly unreasonable wild goose-chase after a perfect development platform.

I'm beginning to port code from the old javascript repository into Haxe today.

-Duck

[Karl]

I wondered about networking w/ javascript but figured webclient was worth the trade off

Python... well yes you have to call C code... OpenGL is a C library o_O. Every language in which you use OpenGL calls C code.

Haxe is the same, you have a Haxe interface that calls the OpenGL C library. You are trading one interface for another. If you want to use box2d or some other neat library you're going to have to create a Haxe interface for it... unless you get lucky and somebody already has. In addition to the massive Java API Altitude uses in the neighborhood of 15 libraries.

I would recommend taking the path more traveled such as UE 4. The source is available on git hub.

Unity seems good too but the source isn't freely available. C# is a nice interpreted language.

If you want a path less traveled then look into libGdx - it's what we use. But it's not without it's many problems.

The world runs on C/C++, using anything else you will have lots of fun problems to deal with... you just hope there are other people on the internet that had those same problems and found solutions. This is less true when there are fewer people using your solution. Everybody hates header files . We use Java for a lot of reasons including but not limited to:
it's API is massive and well documented,
IDEs for Java are light years ahead of C++ ones
Debugging - you can change code and have it hot swapped while the game is still running
So many high quality open source libraries it's a little crazy - we use like 15 or so open source (mostly apache) libraries.

[MagneticDuck`]

Thank you for the input Karl.

> Python... well yes you have to call C code... OpenGL is a C library o_O. Every language in which you use OpenGL calls C code.

Of course this is always technically true; I didn't describe my complaint about python well. Systems like pygame don't just link against the general OpenGL libraries, they link against C logic written especially for the occasion, to do lots of performance-sensitive things that would be written in native code if you weren't using a language that's slower than javascript. It ends up feeling like you're coding your game in a scripting language inside a game engine written in C. Less than optimal.

> Haxe is the same, you have a Haxe interface that calls the OpenGL C library. You are trading one interface for another. If you want to use box2d or some other neat library you're going to have to create a Haxe interface for it... unless you get lucky and somebody already has. In addition to the massive Java API Altitude uses in the neighborhood of 15 libraries.

Haxe is, first and foremost, actually quite different; it's not really comparable to most other solutions like python or java because instead of interfacing with your platform's libraries, you transpile your code natively to them. Can't beat that in terms of performance or platform API. And in terms of concerns of available libraries, the OpenFL development toolchain provides platform-agnostic graphics and audio, the Nape physics engine is available natively in haxe, and a haxe package repository, well, exists fairly prosperously. From investigation, I'm confident there is enough.

Here are some games written in OpenFL.

[Karl]

If you have an IDE that automates the build process and lets you launch from the IDE then it's not terrible, waiting on C/C++ code to compile before you launch is still pretty horrible. Maybe targeting android would speed up the build and debug process.

Debugging also sounds like it will be problematic as the code you're debugging is generated and is a layer of obfuscation between the actual Haxe code.

[MagneticDuck`]

The issue with compilation isn't so much of a problem -- the transpiler works in about half a second, and the rest of the compilation isn't necessary for error checking. Now, the concern about debugging is real... my best bet is probably to use an in-browser debugger on the html5 target (the produced javascript isn't that obfuscated).

[Solikos]

Technical questions aside, as a PM my first question would be 'why?' Is there a spec you're working toward that I've missed somewhere? Sounds interesting!

Edit: To be clear, I did read the github intro to the project; just wondering if this is going to be a near-clone of altitude with respect to the gameplay ('classes', weapons, perk system, etc.) aside from allowing others to mod it (which may actually break the community up into small groups than it already is).

[MagneticDuck`]

Right now I'm simply writing a generalized infrastructure with the immediate result of having a multiplayer plane game that cuts few corners in its technical design. I had reached roughly this zone with the javascript implementation. I will then augment it in subsequent steps. I do not intend to clone altitude, but I will keep many of the basic elements of play, including the plane class / perk system.

I'm not worried about an overflow of productive work from the community, let alone to the extent that it will divide people. Let there be no delusion; to date, I have only been able to contact one person who is willing to put in some LOC in the name of the project eventually. As long as things remain in this state (small profile, in the scope of the altitude community), the development of solemnsky will too remain a small and necessarily dedicated nucleus, primarily myself.

[MagneticDuck`]

Just to remind everybody, solemnsky is indeed still very much alive. We even have a roadmap and all kinds of good stuff. Don't hesitate to contact me if you want to help somehow.

[Oyster]

If you need any help with the text side of things, such as writing/editing anything, feel free to let me know.

[Kafka]

Any idea when the next demo will be released? I'd love to see what you've been working on

[MagneticDuck`]

Well we don't have much in terms of demos right now, but I have a rolling release of the precursor to an eventual online gameplay tutorial here. It's more of a practical application of my game engine / customization framework than anything else, but you can see LEGRAPE's neat plane design zipping around. Consult the roadmap / our skype group if you're interested in keeping up to date.

[MagneticDuck`]

Originally Posted by Oyster

If you need any help with the text side of things, such as writing/editing anything, feel free to let me know.

We might; people were talking about making a compilation of the best gameplay ideas, because right now everything's just floating around... come to the skype group and see if you can find something. (Add me as magnetic.duck).