There are two things I'd like to know, and please all coders great and small.
1. What kind of programming challenge do you envisage being feasible for a possible Sumea programming challenge?
2. How many hours of work should each programming challenge consist of approximately, and what kind of investment of time can u put in?
Personally, I'm not sure I could spend more than a few hours a week on any programming challenge, and to code anything substantial takes more than a few hours, so I think some form of challenge every month or maybe bimonthly would be good from *my* point of view.
However, if other people have more time to invest in this, and want more frequent challenges opinions would be nice.
As far as what the actual challenge should be, I think it would be good if it wasn't in anyway "interactive" in terms of the final product. I'd also like to see each challenge build upon contents of the previous challenge, and also maybe steal a bit of help from the artistic challenges?
A programmer contest is different to that of artists, as they can usually put up a static picture of something and make it look great, we'd have to have a rolling demo or something, which would mean more space to hold the exe's - then there's viruses etc.
We'd have to agree on programming language, api's etc, it's a lot of work I'd say.
Although just off the top of my head there's:
A simple demo showing something {glowing, shadowy, colourful, pretty, technical} using just plain C/C++ (inline assembly allowed) code, using OpenGL (becuase its a lot faster to code for than DirectX), with a timespan of about a month or so from start to completion, and the sourcecode available when the comp is finished.
More ideas need to be thrown onto that pile.
I agree with Daemin a demo like the 64k demo's back in the day, would be a great idea. (maybe not 64k)
But I don't really see the reason to set the api to be just OpenGL or just DirectX, it should be open.
As long as its C/C++ and under the set size limit it should be allowed.
Maybe a list of allowed api + libs should be made.
I don't see how "agreeing on api's, language etc etc" is alot of work :) It's just a decision.
I think the reason(s) to choose a particular API are reasonably self-apparent. Obviously there's the consistency of the back-end of the code, which takes out an element that reduces focus on the actual programmer, and also it probably allows for better code recycling from one challenge to the next.
As for space, maybe making space at a premium as part of the challenge could be a good idea. Which is why I'm thinking the challenges should all run off of one preset "back-end" that's organised before the start of the challenge and that way there is a minimal amount of people downloading the same pieces of code over and over again (this is why we have dynamic linking etc ppl!).
As for virii (not viruses :P) I think that's a little bit paranoid!
So Daemin/lava_monkey what were your responses to the 2 q's that I asked above? I think I got some vagueish answers for q1 but not q2? I'm thinking about how difficult the challenge should be given the timeframe (approx one or two months?) so that it's reasonably accessible to all code monkeys. Coding a demo in one month isn't that easy -> you'd have to invest a fair few hours to get something reasonably complete and functional.
Two months would be good, depending on whether or not there's a common back-end (anyone up for it?). On the other hand, it's pretty easy to write a couple of funcs to get your display up & running in DirectX or OpenGL.
As for a subject, it's probably best not to be too ambitious straight away. Something specific & easy, like alpha-blending, would probably be a good start.
Also, I noticed there's been like 30 reads of this post and about 4 replies other than me :P
There's one more question that I think I would like a ***broad*** opinion on.
Do we want to have a bunch of *strictly seperate* coding demonstrations that could maybe be spliced into one big demo. Ie first challenge is like a waterflow problem, second is a ballistics challenge, third is a whatever -> then after x number of challenges we splice up a bit from each to form one long demo.
*OR*
or to have a set idea of a final big demo *from the start* and each challenge forms a specific part of the code base for that demo - with each challenge likely being smaller and more focussed and the code base is built up based on some kind of winning judging criteria.
There is problems with both of these ideas, and any more ideas about the aim of the challenge would be nice.
1. The problem with the first idea is a possible extreme lack of consistency in the demo, some sections may really outshine others or some sections might perform better on one type of machine than another, and hence performance disparity might reduce the objectiveness of the demonstration of some of the coders' work...
2. The problem with the second one is a lack of acknowledgement of the work of the "non-winning" entrants, and difficulty in including/promoting their work and also it might make each challenge on it's own seem far less appealing, however the end result could be more consistent and cohesive.
I personally believe that the second option is better - but that's just my opinion.
Either way, I've just had an idea for the structure of a possible programming challenge, and I'll hopefully get a draft up for people's perusal within a week, and code-monkeys can tell me whether or not they'd like to participate. If no one likes it, then bugger it, we'll have to try another approach to the programming challenge :)
Take a look at Ludum Dare's 48 hour game competition entry conditions. Instead of saying "one language, one api, etc" they let people choose from a range of API's, languages, and even platforms. They won't let you use an ?berengine that requires little or no effort on your part, but a basic framework to start from is accepted.
I agree with GooberMan with the api's a list would keep everyone happy.
As for making a set template the all the demos can run off, I'm not sure if its worth it, I can see it will make source code easier to read and help get some people started. But it might be a waste of time setting it up I'm sure everyone?s got there own way they like doing things and they wouldn't take long to setup they're own demos.
As for question 2 and time frame, a month or less would be good, over a month is abit too long i think, I don?t have too much time to spend on it myself as I?m already working on my own projects as most would be, but I still think I could get a decent demo done in my spare time in a month.
And as for what i guess is question 3, I think separate demos for each challenge is a better idea. That way the winner of previous contests isn't going to have an advantage over the rest, and the specific task for the contest can be easily judged.
So from the feedback I'm getting so far, most people would prefer to have stand alone competitions with open slather on code base/tech? That's okay, as long as everyone acknowledges the fact that if that's what they would like to do, then there isn't any really flashy/easy integrable way of showing off the work beyond this website that I can think of, which is fine as well, as long as you realise that...or can come up with an idea yourself to show off all coders great/small on this site.
The only other problems I see with the "open slather" on OS/API/front end is that the *actual programming* won't be easily judged, not many people are comfortable with handing out their source code. Also having each competition be based around a single demo effect - which could be demonstrated in so many different ways or interpreted - then there'll be the ones that look fantastic, but might just be cleverly-disguised bad-programming with some good art work? It becomes more of an artistic exercise than a specific *programming* exercise.
Plus having one "demo" per competition basically tells anyone who doesn't have alot of time to spend coding projects that don't earn a pay-cheque or any large amounts of recognition to not bother.
quote: That way the winner of previous contests isn't going to have an advantage over the rest, and the specific task for the contest can be easily judged.
An easy response to that problem is that although each challenge would eventually be constructed into one big demo, each challenge, coding-wise, would be entirely non-related to the previous challenge, even if the final product itself may depend on the previous challenge(s).
The one last problem I wanted to address is that not every coder on this site is going to be the greatest coder on the planet - that is - capable of coming up with a stand alone demo on their own in a month. Sure it's easy for the more experienced programmers to scoff at some of the things the less experienced may struggle with, but developing some coding skills, some experience and critique might be very valuable for the less-experienced coders in this forum and I'm not sure how perfectly suited that size a challenge is for them.
And even if you say "oh just give them a shit-easy challenge, it's hardly going to have any big impact or presence or glits/glamour, and hence no one will pay any attention to it.
These are just some thoughts, as far as I can tell so far, these are only my opinions, and the opinions from most people here is that we should just have stand-alone demos - if anyone else wants to chip in, please do!
Although an idea just struck me now, that we could have the first competition be one where everyone produces a plugin for winamp. That should be simple enough, would allow people some flexibility, and with the sound input it would allow for some really cool things to occur. A month timeframe should be cool, and even if people don't finish it, it would allow them to continue it as a small side project for all eternity.
OK I thought I should perhaps put my suggestion forward and think the easiest way to submit a finished game in a short period of time with something better than programmer art would be to use flash. As I found that when I entered a competition using flash there was only a rather short learning curve and the art isn?t that hard to do, thus concentrating on the code can be more of a focal point. I understand that there wouldn?t be as many special effects and cool things we can all do with openGL or DirectX but at least we as programmers can submit a finished game (which I personally find more satisfying) for others to actually play and hopefully get addicted to than watching a fly through demo that our engine can make.... don?t get me wrong I love seeing peoples engine cause it gives me inspiration to keep going.
I think I have ranted enough now but I think flash could produce a finished game in a shorter time frame that can be actually played.
Sam
Personally, I feel making a game in Flash is more of a design challenge than a programming one - and unless Flash games take off real big in the near future, it's probably not the best way to show off our l33t game programming skillz either.
The real issue that needs addressing is how are we going to judge the *coding*? It's practically impossible to judge on things like style, or API, or language, as that's entirely personal preference. The easiest solution is to lay out strict rules & guidelines and see who can work best within the given framework, but that's just one option...
There's like 15 active coders on sumea yet I've only got feedback from about 4 or 5 in total, as much as it's good to have the different opinions, there's not exactly a tonne of people xercising their democratic oppurtunity....before I waste too much time coming up with a proposition for a competition, it'd be nice to get a really clear view of what every(coder) roaming this thread thinks. Hell - even the artists can chime in.
Just my 2 cents worth..
A friend of mine who's in the biz has told me on many occasions that, as a programmer, your certificates and degrees are nice, but if you have a working game or something to show to them, then you're definately a good step towards your way into the industry..
A Sumea programmer challenge would be a GREAT opportunity for you programmers to get your stuff seen. And hey, a whole heap of industry types browse this site (even though they don't participate in the forum as much, I know they're here. Trust me [:)].
So as Maitrek has mentioned, post a msg here letting him know if you're intertested in a Sumea programmer challenge, and if there's enough people to warrant an effort, discuss something that would be beneficial for you guys. Something that's worth your time / portfolio worthy..
People that are serious about coding games, generally have written their own engine, or at least a framework. If you haven't, I suggest you do because you will learn *alot* in doing so.
Assuming only the people that enter are serious about gamedev, they will have something to work with. I think that a simple game is not, or should not, be a huge unacomplishable task. Those that don't have an engine could always use someone elses, and there are plenty of good free engines out there.
I'd say a no for teams, since that would involve more people, would require a larger timeframe, and the project would be more complex (not to mention all the coordination efforts involved).
Also a big 'NO' goes out to the flash idea, simpl becuase flash is a way that artists can make games, not programmers.
OK redwyre I new someone would flame me - but please.... I was just making a suggestion and I had at the start of my original post "Please dont flame me" but I thought people may be more mature than the help wanted people over at gamedev. I would be more than gald to show off my OpenGL skills - I was just thinking of an easy way to show off all skills involved in making a game in the fastest possible time.
Speaking as a coder for the coders, putting anything up saying use (insert web dev product here) as a game development tool is insulting.
However if you would like a challenge, I say pick up either HLSL or CG and create a Shader. To do something with a bit more funk in it rather than Straight O-GL or DX.
Oh and C / C++ I think is the desired language of choice, whilst CG enables you to skip around C++ a bit with it's own compiler does make it easier for the artists to enter without having to grasp the fundamentals of C / C++ Code first.
I myself would use DX9 as my platform of choice for a task like this.
[8)]
0xBaaDf00d
quote:As far as judging the coding etc... let the final product be the basis of judgement.
Okay, that's an obvious statement, but there are plenty of components of a final "product" that could dilute the "coding challenge" part of it. Like I said, neatly disguised crap programming with some great artwork could quite easily win over judges (whoever they may be) yet the coding part of it may have - on it's own - been very substandard and lost out to some of the less well "presented" good code.
Also - making a *game* is a very involved process, and alot of those processes have *nothing* to do with coding in general, which I feel may detract from the value of the coding challenge. I'm going to make a bit of a ridiculous statement, but the only really good way to produce a demonstration game from this site - would be if we had an artistic challenge on the site, a game design challenge, music challenge, sound effects challenge and a coding challenge all agreeing on the one big product.
That's not a very good idea, I'm personally leaning towards keeping the challenge short and to the point and keeping artwork/music/non-porogramming related areas to a minimum, rather having tons of creative content that isn't programming related and a long and involved project.
Having said all that opinionated stuff, my solution to the fact that (near) everyone wants to use their own framework, I think that for the purposes of my propsotion, I may code up some framework for people to feel free to use, or they may use their own if they wish.
"Like I said, neatly disguised crap programming with some great artwork could quite easily win over judges (whoever they may be) yet the coding part of it may have - on it's own - been very substandard and lost out to some of the less well "presented" good code."
I don't think John Carmack has ever been accused of writing good code, but his work is among the most respected. Game programming is all smoke and mirrors anyway. Presentation is not limited to art assets. The sum of the parts != the whole.
If you want a "coding challenge" lets skip the pesky game elements and write a console app.
Scott.
Steal our stuff, dear god noo... lol
I'm sure the artists would help out with content to help fluff up your codey bits.