SFML community forums
General => SFML projects => Topic started by: Rimdeker on July 11, 2012, 05:50:21 pm
-
[REDACTED]
-
I thought of a method of collaboration that should work fine for a small group of people (this won't work for bigger groups, then you would need an actual CVS and what-not).
DropBox (or something else if you can prove the others and me that what you're suggesting is better/easier/faster/friendlier/sexier)
You know DropBox is also already some kind of version control system, just that's not usefull for coding. I myself am getting more into git and it's actually really simple and unbelievable powerfull! Also you could then open a GitHub or similar repository where you can share your work even more.
(I was thinking of some IM messenger, IRC or even a Facebook group).
I can suggest to use IRC since it's very easy and you can quickly connect to other networks to get help (e.g. #sfml or ##c++ etc.), but since development needs also good communication it's often good to have some voice chat, for the mumble is really the best way to go.
Good luck!
-
I'm interested, PM sent.
-
Hey,
This is interesting. I'll keep an eye on this thread to see where it goes. Feel free to PM me as well.
-
Hmm... maybe... I'll think about it and then I'll send you PM or write post here
-
I think a version control system like git or whatever is a must for exchanging files. They really make life easier. Whatever you want to use for collaboration other than major meetings shouldn't depend on all team members being online at the same time. (Email, forums, etc.) Mumble could work if you want to have a meeting, but I would expect such things to be rare especially if the team is geographically diverse.
I think a small project is perfect to begin with. You mentioned something along the lines of space invaders. I can see a team of 4-5 breaking down such a project into manageable chunks. Ultimately, the most important aspect of working on a team is having effective leadership. While it may seem fun to try work democratically it is rarely productive.
-
I'm interested :) Got free time on my hands and was looking to make a game, something like a side scroller shooter (metal slug for instance)
-
the most important aspect of working on a team is having effective leadership. While it may seem fun to try work democratically it is rarely productive.
I can vouch for that. Last year I wanted to participate at ImagineCup contest and I went the democratic way "let's discus, let's see what's best for everyone etc.". This costs insane amount of time just making the meeting possible. It is best to decide who is the team head (usually the one with the idea or the one making the team) and let him/her decide who does what.
Like it was said, use E-Mail, Forums or something where messages stick and you don't have to be actually online to read and write messages. Facebook groups is a good alternative, easy to setup and has basic file share. For sharing larger files or to have better file management, Microsoft SkyDrive can do that. If you make a Windows Live group you get like 5 gigs for the groups SkyDrive. Or just use a git repo.
It is better to avoid installing multiple applications (last year I was using Microsoft Groove and other apps to share files and it sucked hard). If possible, make it an all online thing, use a link to get to your files or talk with team members.
As a bit of off-topic to anyone who cares, the ImagineCup entry got as far as round 2 in GameDesign because all we needed was some sort of documentation, however when it came to actual coding the team management idea didn't work and we got disqualified.
-
Well, communication is indeed important in a team, however when it comes to coding you can't put two guys to do the same thing. You can make some design sketches with the team, every member sees the design process and adds to it. Also before the design there are the features, the team decided what game they do and what the game can offer to the user.
You can use an iteration plan model, members come and take iterations, make what they can of it and posts the iteration back with notes (something like "I only managed to make the add method" or "I tried something, however it doesn't work" etc.) then someone else comes and takes an iteration (maybe one that's half way done, and adds to it). This way you got constant communication between members (even a dedicated section for discussing how an iteration was done would be at the benefit of everyone).
I can't promise the idea above will work, however it uses part of a software development (bottom-top implementation) model I currently experiment with my project, so far it works as expected. I would like to participate however soon I'm gonna start a team projects with some of my class mates and I fear time will run short from there. Will you be accepting guests that just watch how the project is being developed?
-
The whole point of this is not to introduce people to version controlling or some sort of standardized methods of working in a team or even being extra-ordinarily efficient, it's a rather experimental method and I got the idea from the Super Meat Boy team. The only experience I got in working like this is with my friend and it went pretty well, this time it might go down the drain lol I do not know, but I think it is worth a try.
Yeah some developers might like not having to learn a version control system to work with their team. Me I'm sort of the opposite. I have got to have at minimum a certain set of tools. A good version control system that doesn't annoy me is one of them.
What I like about version control system is that they generally:
1. Provide a process for documenting high level changes made to the program
2. Allow one to quickly view changes made by other developers.
3. Automatically merge changes together with little risk of clobbering someone else's modifications.
4. Selectively control what changes get committed on a per file basis and the ability to quickly review all changes prior to commit.
There are a lot of advanced features too but surprisingly the ones I use the most are the most basic. Commit changes, revert changes, view log, and update (pull).
The learning curve of a VCS is typically much lower than that of your average programming language especially when using intuitive GUI tools like TortoiseSVN for Windows. So in my opinion the difficulty of using a VCS is being exaggerated. It sounds to me like the reason you don't think its worth it is because you don't want to use a VCS.. That's an okay by me, but really you are missing out :)
-
Hi there,
I've got to say that it's kinda a great idea, if the opportunity is still available i'd like to join you guys.
You talked about several tools im currently using on my spare time/work :
- Github / git
- Irc
I started learning C++ about 2 years ago so i'm not a rookie but there are so many things to learn and joining a team will help me in my future projects' management.
Little Question : Is there a motivation/project idea behind this team ? I mean we could think about a bigger project if things keep going well.
I went on the irc channel but there were nobody on it, anyway, Please drop me a pm if it's possible
Regards, Rexou.
-
To help keep others up to date (since I never did post here in the thread) I'd be interested in helping. I already added Rimdeker on skype, and have talked to him a little bit. For anyone with skype that wants to add me, my skype is: nekofuu
I have, I'd say, basic-intermediate knowledge of C++ as a language, and basic knowledge of the SFML library. On top of any help I could throw in as a programmer, I use to fiddle around with pixels a few years back (meaning I use to do pixel art). So until we started a more serious project that wasn't meant for purely learning purposes, I could suffice as a pixel artist until a better one would be able to take over or help out.
-
What my plan was is basically like you have a big group of people in a room but only one computer, one person works on the code for at least an hour and then the next person goes at it with a given task from the previous coder. Everybody can watch as people code and communication is a must, naturally.
(I liked this analogy so much, I had to write it in bold letters lol)
It is a good idea. I myself have found some game development tasks really hard to figure out. For example, I've been trying to figure out how to create a simple 2d lighting engine that integrates with a physics engine. I would definitely love to sit down at watch someone who knows how to do it go through the process step-by-step.
The space invaders idea sounds really good. It is a simple enough game to where there isn't any doubt that it could be completed, but I can see where there is plenty of room for creativity and learning.
-Steven
-
Why more people? I advice that you have a project adequate to the team size. If its too short, you can't really make it a learning experience for everyone in my opinion...
Unless perhaps that you record all the process with commentaries for the others to see..
Pick something a little bigger perhaps, which can be modular, so each task is well defined and can be assigned to a person, who will have leadership over the task.
It is not easy to get right, but if you had a well defined protocol between you, on how things fit together, each could work in a part of the game, under defined constraints.
-
i can suggest to look at terraria http://www.terraria.org/
-
i can suggest to look at terraria http://www.terraria.org/
Why? Is this some kind of publicity? :)
-
i can suggest to look at terraria http://www.terraria.org/
Yes you can, but don't expect that this will be a project to work up to. It may seem easy on the surface and even easier with a team, but it's not trivial and needs a lot of knowlegde in C++ and on game design.
Also I'd suggest you better start at one point instead of discussion you dreams. ;)
-
i'm not taking a part in this project ;D
just someone mentioned they have 4-5 people, so i linked an interesting idea to start from (seriously space invaders for 4-5 people ? :o )
-
Or they could pickup a small scale adventure/rpg where one could plan stats for character , weapons, buffs, enemies and implement them in some text file format.
Other could do the actually coding of making these systems work together, even two people. Then make people complement that rich content with graphics and sound.
It's at least something to show in the end ;D
-
PM sended, waiting for reply :)
-
i'm not taking a part in this project ;D
just someone mentioned they have 4-5 people, so i linked an interesting idea to start from (seriously space invaders for 4-5 people ? :o )
It makes more sense to do a small project that has little chance of not getting completed. Besides, even with space invaders one could do a lot. For example:
1. Varying enemies with different AIs that increase difficulty as the level progresses.
2. Weapon upgrades, power ups, and all the special effects to go with it.
3. Boss battles
4. RPG concepts such as levelling up, attributes, etc.
5. Tower Defense concepts such as building turrets
Anyway, the idea is not to make space invaders more complex and drag it out, but if we have several people we might use the extra resources to make the game more interesting and there are a lot of ways to make a simple game more interesting.
-Steven
-
I 'd say not to stick too much to space invaders, for a group like yours to suceed, you are going to need a lot of MOTIVATION and DYNAMISM between the group. I think those are the key, as the project leader usually keeps working alone after something, because others didnt feel motivated enough.
I think its kind of the same feeling a parent has towards his son :)
Discuss between yourselves what is the medium of everyone's decisions, do something feasible. An RPG is nothing special if you keep it simple.
When people say a RPG is tricky to make, its really about complex inventories, too many items to fill them, too many weapons and customizations, multiplayer perhaps.
Stick with me, if you had in your mind to start a RPG.
1 - Tile maps or sprite placement - Just use a premade editor and parse that file. Absolutely doable.
2 - Physics, add a basic quad collision system.
3 - Make easy to parse xml or anything else formats to specify your content
4 - Make a base entity, and derive it to player and enemy, and spawn some of those.
5 - Deduct hp, earn gold on kill
This may not seem a deep planning, but if you stay focused and work with your mind in place, each of the tasks is very doable in a few hours, depending also if you make wise choices on how to approach issues like the tile map format(thinking TMX as a good example for example, theres tutorials).
The numbers game is funny but once you do a few of those you certainly gain a lot of experience and something like becomes too easy for you, logically speaking. I bet you can do that game in 30 mins instead of 4 hours when you grasp the logic better :)
-
Hi All,
I would be interested in joining this project. I have an average knowledge of C/C++ and SFML and minor media(graphics, sound) skills. Probably will not have a lot of time - a few hours a week - but it sounds like fun.
If needed I have a personal/private server with redmine (issue tracker, forum, wiki, news, files, documents, etc) and svn.
-Chris.
-
Hey, I would like to join group. For next 2months I have about 8h of free time a day, later its about 2-4 (starting second year of college - Computer Sience with Game Dev spec). I mostly code (C++) but I'm trying to do some 2D grahpics. Right now I'm working on Space Invaders myself.