Your proudest work to date.

[megatron 0]

Today, I finally got SFML installed.

Using only the socket example and the API documentation ( The first time I have used the docs to learn how to use something ) I have created a primitive chat server and client.

I have been working on since 11am this morning, and ready to publish it 12 hours later.


For anyone interested here is the code ( NOTE: Because of terminal IO, you receive any new messages after you send one, it sucks, I know. My next phase will be using the SFML graphics library to make an interface for it )


What is your proudest work so far? Do you still have the code? Doesn't matter if it's a simple function to a whole library. Share it, and tell me why you are proud of it.


I am proud of this one because I have never been able to use network stuff before, and I have never been able to decipher an API to see how it actually works. So, I'm pretty chuffed.


EDIT :

Code removed it was too long, if you want it, PM me.

[Little Bobby Tables]

const std::string IP = "127.0.0.1";
hehe now i know your ip and will h4xx you cause I'm l337

my proudest work is between my python irc bot and the front end i wrote for my lexer.

[megatron 0]

hehe now i know your ip and will h4xx you cause I'm l337

You are talking to the guy who knew your name the first time we spoke over IRC. ;)

How is your compiler coming along?

[Little Bobby Tables]

i didn't really try to hide it lol.

How is your compiler coming along?

terrible. I'm trying to write it in nim lang, but its slow coming

[megatron 0]

Slow coming ain't the word, it's been weeks since we last spoke, made much progress?

Though you have given me an idea, I'm gonna turn this chat program into a remote compiler for when I'm not at home. Fancy a little project?

[LB]

I am most proud of my code (in multiple projects) that uses complex template metaprogramming. Not only is the complexity completely unnecessary, it is still basic compared to what JLBorges can do.

Other than TMP, I'm proud of the fact that out of the hundreds of incomplete projects I have, some of them are actually mostly complete and in use by a few hundred people:
http://www.LB-Stuff.com/MMF2/Extensions/

[megatron 0]

JLBorges has helped me a lot with my programming, if I were religious I'd name him God.

That's a nice collection of things you have built over the years, I'll be sure to take a gander tomorrow.

I hope I can make a decent collection of tools.

Other than TMP, I'm proud of the fact that out of the hundreds of incomplete projects I have, some of them are actually mostly complete and in use by a few hundred people:

Which is the most used, and why?

[Lachlan Easton]

I think my proudest work would have to be a modular mission scripting system I wrote for a game. Second proudest would be http://goo.gl/y35F3U
(don't buy it, it's shit, but at least it doesn't crash anymore)

[megatron 0]

Wow that's pretty cool!

I'm hoping to make a game (Now I have SFML installed finally, which was a bit of a ball ache)

Did you make it in SFML?

[chrisname]

My half-finished and horrifically bug-ridden x86 OS kernel, I suppose.

[megatron 0]

Wow, whether half finished or horrifically bug-ridden a kernel is an amazing feat.

I attempted, a while back, I got the boot loader up and running, after that however, trying to write the rest in C didn't go down so well. The tutorials I used were out-dated, and I couldn't copy files to a floppy.


How far did you get? CLI/GUI? File system? CD drive up a running?

[chrisname]

Text output was the first thing I did. It's important to be able to get some kind of output early on, mostly for debugging. Some people prefer to do it with graphics or over the network but those are very hard to program, which means you have to do a lot of debugging before you can get the output that enables debugging. So I went with text.

After that, you have to do a lot of technical foundation-type stuff with various data structures, basically setting up the CPU and hardware, before you can do anything interesting, and you don't really get any visible results from that. You also can't usually fully implement the features you want to implement because you always have to finish something else first. For example, if you want a double-buffered GUI you need to have a memory allocator, but you can't debug your memory allocator without having some kind of output, so you have to write each thing bit by bit and jump between them in a very haphazard way. And since I was learning while I was going, and writing it on and off, it took me a long time - it's been about two years on-and-off now.

I did get to the point of a mostly functioning memory allocator and keyboard driver, but because of all the bugs (mostly from all that two-years-old code) I've been trying to move as much code out into modules as I can. That way, I can separate all the different pieces of logic and it will be easier to see how they interact - because a lot of the bugs come from faulty interactions between modules that work by themselves. But ofc to do that I need a module loader, which is going to consist of an initial ramdisk parser and ELF executable parser. But for those, I need the memory allocator to be fully up and running. Iin order to reduce the amount of code that's in the kernel, I first need to write a lot more code. So, you see, it's all very circular, it's a kind of dependency hell. And that's why after two years, it's still only half-functioning and full of bugs. That and the total lack of design, because, like I said, I'm learning as I go.

[megatron 0]

Text output was the first thing I did.

Same here:

Print:
	lodsb
	or	al, al			; al=current character
	jz	PrintDone		; null terminator found
	mov	ah,	0eh			; get next character
	int	10h
	jmp Print
PrintDone:
	ret

I always prefer text over graphics anyway, always seemed simpler and easier to implement, I never got as far as graphics when developing a kernel.

I never got to those stages (Data types and structures etc) nor a keyboard driver ( That was next on my list on the tutorial, I think) never even seen an allocator before. Seems intense.

I would love to work on a kernel, my programming is nowhere near your level, I'm sure. Though I'd happily contribute if you need an extra pair of hands.

[twiggystardust]

To be completely honest, I'm most proud of what I've been working on for a couple weeks. It's a Final Fantasy Type RPG, like one of the first 3. I'm not going to share the code though, it's A LOT. I will admit though it's not in C++. I'ts in Java. Although I've never fully finished any of my own games, I'm bound and determined to get this one done. It's what I've been working up to since I started programming a couple years ago. I'm still finding new hurdles and finding ways to get over them. It's been fun so far, and I get more excited every day when a new addition works, and even when they don't work, the hunt to find the problem(s) is always fun.

[megatron 0]

That's excellent! Can you tell me about the game?

What you find fun in programming I also find fun, when something works, or when you fix that bug, it's so.. gratifying. It always allows you to keep your senses sharp, and the feeling of self accomplishment is a real confidence boost.

I've tried making an RPG in the console, but what I wanted was too big a job for myself, and in my area ( or on the internet for that matter ) I could never find programmers to help me. It's good to see you are determined to get it finished. :)

[twiggystardust]

Well there's really not much to say about the game so far. I haven't written a story or anything even like it. I asked a friend to write a story, and I told him I'd mold the game around it, but so far I have a tile map, movement around that map, the menu to display(you can go to each selection(except the save function, that's for later) but you can't actually do anything with the menus). That's really it so far. The next step I've planned is some way to implement a dialog system, I've got some ideas, I just need to read up on text output with Java, then it's on to the battle system.

I know exactly what you mean about fixing a bug, especially when it's one you're having a hard time figuring out. I've had times in the past where I was stuck on a single problem for a week. The entire time it's disheartening, but once you fix it, the elation is unparalleled.

You know I was thinking about doing an RPG in the console first, it'd probably be the best way to get all the integrated systems working first then add graphics. But I started programming graphics before I really did much else, so I just told myself "To hell with the console, I'll make a graphic based game."

[NoXzema]

I do have an IRC parser I implemented. It's my first IRC client library that's asynchronous, has a full fledged parser (using Lex), and is rather optimized. It also has a good design, enough that I can turn it into anything or it can be used by a 3rd party. Only ugly part I don't like is where I mixed C and C++ which forces me to use malloc in C++ code. Not sure how I feel about it yet.... originally, I was going to make it entirely C (aside from the ASIO implementation for networking) but string formatting is *so* much easier in C++ than it is in C...

[3hM2Nwbp]

A bulldozer that is capable of traveling 45mph. Only had it up to 25 at 1/2 throttle...anything above that and things will break.

Dodge 2.5L inline 4 cylinder gas
Dodge Dakota flywheel/pressure plate/clutch/first 3 inches of transmission cut off
Deleted emissions monitoring inputs from ECU
Custom 1/8" thick steel gas tank
Touchscreen LCD instrument cluster powered by arduino
Rewired ECU module (rev limiter piggyback to lower max rpm)
Custom hydraulics to work blade off of existing power steering pump.
Custom body panels in progress.
Over 30 cases of beer were harmed in the making of this bulldozer.

Time invested: 2 months
Cash invested: $1,000 (plus a car that was going to be scrapped that I saved)

https://plus.google.com/app/basic/photos/116741526010375708873/album/6068416258817135697

* Watch the video and hear her roar...then you'll know why I'm proud of it.

[twiggystardust]

Luke: Pretty sure he was talking about code, but that is badass!

[3hM2Nwbp]

I had to code the arduino instrument cluster and calculate the input curves for various sensors. Not 100% linear calculations at times...

:)