It's a work night and I'm in the mood for some opinions then help with a project. If anything I had a certain library I would like to build down the road, and I want to use Java as my platform anyways for a bigger project, and would like to use Java as a means of prototyping/debugging/testing to make mass data management a breeze for a upcoming project and hope it can replace sql lite.

We all know C++/C are great at building portable, slim, good performance applications for your environment, but have many traps for young players wanting to release their applications on a any day machine, which is why I notice a lot of us use functional programming to test out something before spending hours converting that code to C/C++. Just like the big boys, Google. ;) Of course at the end of the day, it's all about what gets the job done. Even if you use Erlang to do heavy number crunching, and you manage to make it out perform it's self with some hacking.

Anyways my question is to everyone is. What is your prototype language? Did you pick it because it's what your most comfortable with? Or do you go out of your comfort zone to learn something new because the language supported your needs for prototyping. Or is it basically, I have no prototype language, you use what language you plan to launch with?

Go. It's as easy as creating a new folder and a new file in my "src" folder and, 3 lines after, I am ready to experiment.

The nice thing is that it has the performances and the infrastructure to bring your "prototype" to full production without forcing a rewrite once you hit a "complexity wall".

I been meaning to give Go a try I'm wondering how well Golang compares to Java with socket programming.

I been meaning to give Go a try I'm wondering how well Golang compares to Java with socket programming.

Considering the pedigree that's exactly what Go is good at. The entire networking stack, from the low level sockets to the very high level HTTP, page templating, json marshaling/unmarshaling is top notch.

I only have real experience with UDP sockets since I used Go for the game server of my game... apart from some weird naming (ie. "Dial" to create a socket) and the fact that you cant (by design) make the sockets non blocking.. the thing was really fantastic to work with.

I tend to prototype in Python for tasks that involve text-processing, and C++ for graphics-related tasks. Occasionally I'll need a stricter type system, and I'll pull up Scala, or even Rust (though it really isn't mature enough at this point).

C++ strikes a lot of people as an odd choice for rapid prototyping, but I've found that even many experienced C++ developers don't seem to realise just how powerful and compact the language is. It's not uncommon that I'll take a 300-400 line Python program, and transcribe it to ~250 lines of C++, with dramatic performance gains...

I only really develop 2D abstract puzzle games, and generally what I do is prototype them by writing them with shitty graphics to Win32 using the GDI in C++. So for example, my last game looks like this on iOS (link) but started out looking like this (link). Actually the GDI version of that one doesn't look that bad because I went through the trouble of using sprites with per-pixel alpha but if you want to make things easier you can just use the regular Win32 BitBlt(...) routine and use sprites with solid white (or whatever) backgrounds for everything.

I find it pretty easy to write a Win32 game if it doesn't have to look good once you have done it once because you can use the shell of the game over and over and if you need any kind of complicated GUI widget (e.g. text boxes), well, you have Win32 at your disposal.

I have tried doing prototyping in PyGame but it didnt really work for me because

1. to be honest Python slows me down: I find that I personally can write code faster in a statically-typed language because I will just like lay out all the classes based on the first principles and then start building a game with them and let the compiler work for me. You can't do this when any variable can be any type.
2. I end up re-using a lot of the code from the prototype as is in the production version. That game above was written to iOS using cocos2d-x which is a C++ framework. My prototype was C++ + Win32. The architecture of the prototype was sufficiently modular that 75% of it, i'd say, could be re-used without changing a line, and another 10 or 15% could be massaged into equivalent cocos2d-x-based code.

Also tried prototyping in Java but this was worse than Python for me. It was basically as complicated as doing a Win32 prototype without me having a lot of background knowledge about it.

That last point is the key thing: I happen to be really comfortable with Win32 because I used to write desktop applications to Win32 professionally in days gone by (i.e. the 1990s) so I think the takeaway from what I am saying is that you should probably prototype in whatever language/platform/environment on which you are most familiar and on which it is possible for you to easily get a sprite on the screen.

I tend to prototype in whatever environment I'm developing.

For a non-game related example, I was developing a webserver and wanted to add a Restful interface. I could have used another language, but that didn't make sense. But I also could have just created a quick example server, but instead I chose to prototype inside the application. There was much trouble integrating the libraries, fighting conflicts, and learing how to add the interface, but it wasn't tested or robust, and no one used it, so I didn't care. I was able to play with it.

Then came the fun part. I wanted to use the interface for something, but had no way to test it. Rather than create an app or something, I wanted something in the system that was still a prototype. So a wrote a quick javascript restful testing console that let you enter a URL, and some data, and could GET, POST, PUT, or DELETE to the URL. I learned more about my system, plus fought problems with cross-sight scripting, javascript restful calls, and learned a lot about the restful way. That restful console ended up becoming the testing tool of choice for the whole system. I never would have went that route if I chose a different language or copy-pasted a one-off example.

As for games, I try to prototype in the game. Many times, I find that I need to do something that I didn't anticipate. Such as tweaking variables at runtime, or allowing for a prototype level to be set while running. These in turn teach me about the game, I find bugs never would have found through normal gameplay, and sometimes surprisingly fun stuff come from prototypes breaking the main game.

For games, I prototype in Unity with c#.

I would prototype using whatever language you intend to do the real application/project.

I tended to work on larger/complex ones which had to initially show a large portion of the functioanlity

You want to prove that primary features work (and possibly within performance specs), and you like to know that whatever libraries you use HAVE the features you need working.

SO one question is of how trivial redoing it all in another language is, versus having ALOT of work that gets thrown away.