Programming challenges contests ideas

Sounds great to me, this would be a fun way to learn new techniques. I'll try to think of some ideas, but in the mean time I also want to say... bring back the 20 event competition! that was a programming challenge.

That would be a lot of fun I think!

20 event was great, the same guy won it 3 times in a row (he partnered with different people from memory), hopefully some new contenders can challenge the defending champion.

That said, the 4 frames x 20 event rule was rather silly, you could do all sorts of hacks to make use of events on other frames, making it essentially an 80 event challenge.

I'd like to see a 20 event, single frame challenge.

Some people have their heart set on a title screen, which you can do with a single event and an active as big as the screen (on click -> set active to invisible), you can also tie other events into it (on click -> set active to invisible, shoot object from player).

But if the consensus is a title screen frame is essential, you could set the rules as 2 frames, first frame must have 5 or less events, 2nd frame must have 20 or less events. That encourages more creativity than say 4x20 events, whereby it's the people who just put in the sheer amount of work to optimise those 80 events that win, not necessarily the most ingenious 20 events.

I also think it should be limited to 1 or 2 frames.

Back on the original topic - a couple of challenge ideas/themes -

1) Top Down Enemy AI / Path-finding - this could be a bit too complex for the average clicker, so perhaps it could even be split into two separate contests eg. "Simple/easy to understand solutions for top down enemy AI" and a different contest for a "top down enemy AI/path-finding" preferably extension-less.. to be cross platform friendly.

2) various challenges involving Fusion 2.5's built in physics engine - I feel like the physics engine is not being utilized to it's full potential because most people (myself included) don't fully understand how it all works (except for that one guy who sells various examples, he has a real handle on it) so many challenges could be based on creating certain movements/mechanics etc. using the physics engine.

Yes please, bring back the 20 (25) event challenge. Also it would be great to see a compilation of previous 20-event challenges uploaded somewhere, I learnt so much from the few of those I managed to scrape up.

The more I think about it, the majority of problems in programming have been solved in other programming languages, there's plenty of examples of pseudo-code that provide these answers. The difficulty is getting these answers into Fusion's format to execute the solution.

If someone wrote a web-based javascript Pseudocode to Fusion Converter that stepped through the events required to execute a solution, this would solve a heap of issues. A programmer could find a function written in C++, use an existing javascript that converts it into psuedo-code, then convert it a step-by-step demonstration of implementing it into Fusion.

Apart from fusion-specific examples like Physics or Extension tutorials, I can't see many problems that haven't already been solved somewhere else other than this forum. If we're going to address them all one-by-one through challenges and examples, it's a lot of (arguably unnecessary) work.

Obviously to include all the functionality of other languages like object orientation and database integration would be a pain to code, but a converter that can handle if, else, while, for and array would accomplish the majority of tasks people need in fusion.

Quote from Jeff

Or I wonder perhaps we don't care how many frames a person uses but its capped at 20 events?
Perhaps we could even up it to 25 events just to get a little more bang out of it.

I also like the idea about requiring certain elements like physics or maybe an building an application that does "X"

That could be interesting, I have to say. I've been waiting a long time to be able to enter the twenty event competition, so it would be kinda neat to finally be able to participate.

I think 30-40 events (maybe 30?) sounds like it could be better if you are counting it across all frames, since then it would be close to half the events people were able to technically get in the past. Maybe that's too many, though, maybe it should remain at 20 or 25 events, I don't know...

Whatever changes are made, though, I really would like to be able to participate in a twenty event competition, since they sound like so much fun. Technically speaking, I've actually had a game ready for the old ruleset for a few years now. At this point I might just decide to build it into a more complete game instead (I tried giving it a unique concept, after all), but that doesn't change that I actually know how to make a game within those limits. As a matter of fact, at one point, due to the absence of the 20 event competition (and because I thought it would be interesting in its own right), I actually tried making my own take on it, which was 5 events per frame, as many frames as you want (and 5 global events allowed), but it was just a challenge and not a competition, and it never really took off (there was only 1 submission other than my example, as far as I can remember, but I never looked at it since it seemed to have lots of blood. Also, I had actually prepared a second example of my own but never released it for some reason, even though I personally think it actually was pretty impressive (a five event Metroidvania engine, albeit using global events due to object selection issues forcing me to use collision conditions+needing a few extra events because of it)).

Ryan - this is a good idea, but having spent countless hours "transcribing" code and pseudo-code to Fusion there's still plenty of quirks and workarounds to be solved to make it actually work. Very often there are even simpler/more efficient ways to accomplish something in one event that would require a bunch of events if it were lifted from say JavaScript verbatim (that said, the reverse is also true).

An example is object instancing, which Fusion handles in its own very specific way. This is both a blessing and a curse depending on how it's used. You can save a whole lot of unnecessary overhead by not foreaching if you have a good grasp of the internal scoping process.

I think that the challenges would be a great way to see how other users solve various problems within Fusion that might have several solutions or that make creative use of these quirks that might not seem immediately obvious.

In the spirit of competition I submit Joshtek/Jeff's example reduced to 1 event using some tweaking of the bouncing ball movement:

EDIT: Added self-collision (1 extra event)

you crazy SolarB - I've never looked into it too closely, do you have a clear explanation of what "randomizer" and "security" do in the movement?

Wow SolarB! Using the bouncing ball movement would be the last thing i'd think of. I probably would've tried for hours attempting to make custom movements/AI etc and it would still not be as good as that.

oops I just checked on security and it should be the other way around! It counts the number of similar direction bounces and applies a random bounce after a certain number. The randomiser makes the directions less predictable. Caveat: on an open plane with this example bounces still happen in 8-dir and will cause objects to get stuck in certain cases. It would be good if there were a similar direction wheel instead of "# of angles" such as "restrict angles" that would hard-limit available directions to the ones chosen.

// end of derail

My suggestion for a comp would be to have Clickteam or by a community vote decide on a theme/genre of project and have an artist lay out a single-frame level with all graphics included. The winner of the competition would accomplish the goal of the project in as little events as possible (under 20/25) using whatever tools they deem necessary. Judging could take into account the 'cleverness' or 'elegance' (whatever that means) of the coding, event structure and ideas. Further points could be awarded for multi-platform capability etc. So yeah, pretty much what Jeff suggested XD The project style would be complicated enough to stimulate creativity but easy enough that it would actually be possible to do...

I think that art should be part of the comp rather than pre-provided. Otherwise the two things will be too divorced in people's minds, which already tends to happen too much imo.

Except for the fact that art can be very hard to make. You can't honestly expect that every awesome programmer in existence is also (at least) a half-decent artist, can you?

I don't really like the idea of Art being provided... the entries will have less variety... and it gives an advantage to the really talented programmers over a lesser programmer who is a bit more artistic etc.