| Eric Coleman |
Posted - Mar 05 2003 : 10:27:47 AM
I've seen a lot of community projects fail. I would like to know what makes those projects fail, and what could be done to prevent that failure.
For those that have been successfull at keeping a community project together, what is your secret? |
| cbx |
Posted - Apr 02 2003 : 12:47:20 AM
I would like to add some extry additions to my prev post regarding multi member colabaration and my personal experience.
1: Don't seek out members over the internet. Put an add in your local paper first or talk to friends and spread the word locally that you want to start your perticular project. This will free you from having to deal with much of the problems that are raised when working in a disconnected environment like the internet.
2: Setting up a web site for the project is critical. Do not hold off creating the site until the project is under way. Creating a web site is the first thing you should do. Also equally important but not really nessary. Is to have a fourm on that site so that members and interested people can comunicate effectively. You can setup a wen site without fourms but it can make community feed back a little more cumbersome.
Communication is key. Not having the ability to verbally talk to another real live person 1 on 1 can be a challenge using technology. It seems like I am always having to describe something to the other person more than once. And vice versa. And if you don't believe me just ask Eric, or refer to these fourm posts right here on vbgammer.
http://vbgamer.strategon.com/msgboard/topic.asp?TOPIC_ID=21
3: The next step I would say is to stop working on the project? I bet you are scratching your head right now. Again speaking from my experience I find it much more benificial to create a library(s) of objects, interfaces, and commands that you think will be helpfull when it comes time to work on your project.
For example i am working on a couple of libraries, one of which is a DirectX 9 helper library called DX9Tools. This library will allow me to create DirectX 9 based apps quick and easy. Another library I am working on is called GameTools. And as the name implies will contain a set of generic game related functionality common to most games. For example Phisics, and ai objects etc.
Also you should create any character or level editors first. That way when it comes time to actually create the game or application you have the tools to create the content for that project and do not have to split your focus between two or more projects
So basicly you want to create a game that has no game engine to run it. This will allow you to focus solely on the game engine itself and not have to worry about anything else. Know what I mean?
At the very least create the content for the first part of the game, so when you test the engine you have something to test it with. And yopu will be able to see what the game will both look and run like.
4: When working together on a project try to break the project up into a number of components. That is to say componentize the hell out of it. Or if you are using a language that supports interfaces, interface the hell out of it. But if you are using a language like VB 5 or 6 split the project up into a number of dll's. Then assign each member a component to work on.
5: Unfortunatly if you are the brain child of the project you are doomed to do most of the leg work. Constantly describing to members what you want done as well as how. But one thing to strive for is to keep in contact. Again just because one of the comunity members does not respond or looks like he/she has quit the project don't assume that they don't still want to help. Keeping in regular contact even if the member does not respond lets them know that you are not going anywhere and that you are serious about seeing your project compleated.
6: Also, you can dream big, but before you start coding the project make sure to scale it back by a factor of 10. This way there will not be any dissapointments about things not getting done. And remember writing software can be very flexible, you can always add features at a latter date.
Some other quick tips are ...
1: Do not worry about coding for speed you can do it latter
2: Again keep a lax schedule or no scedual at all
3: If a member does not feel like helping anymore let him go do his/her own thing but keep giving them updates regardless. He/She may see that the project is picking up steam and may want to start helping out again.
4: I can't stress this enough start searching for members in your local area first! Even if they are a 1 hour drive away, they may be willing to commute very now and then for some 1 on 1 disscusion and planning. |
| cbx |
Posted - Mar 29 2003 : 9:35:25 PM
First off don't set a time or date for the project to be compleated. And just because one of the comunity members does not respond or looks like he/she has quit the project don't assume that they don't still want to help. Keeping in regular contact if not on a daily basis is almost mandatory. Also time zones are a problem, but if you can get all of your members in one chat room (preferably with voice or webcam) all at one time that is a very big plus.
But I would say to keep a lax and relaxing schedule between members is number one. If members being stressed for time becomes a factor, people will quit the project.
My two bits.
quote:
Originally posted by Eric Coleman
I've seen a lot of community projects fail. I would like to know what makes those projects fail, and what could be done to prevent that failure.
For those that have been successfull at keeping a community project together, what is your secret?
|
| Eric Coleman |
Posted - Mar 22 2003 : 2:12:26 PM
It seems no one has the time to help with this, so it looks like nothing will happen.  |
| Almar |
Posted - Mar 12 2003 : 01:38:50 AM
Well Eric, I'd love to help, and I know enogh of ASP to do so... I'm just not sure about my time lately, I have 2 sites to program already, one ASP, and one PHP.. and the ASP thing has gotten on my nerves a bit lately :) |
| kaveh |
Posted - Mar 11 2003 : 5:54:40 PM
I think they fail because they're not well organized, and they dont make a complete technical design document before starting to code. People dont know what other functions/methods will be available to them and what data formats to use because other people haven't written them and theres no document describing the interfaces/data structures. So it gets to a point where everyone is waiting for everyone else to finish their part before continuing. Then people get bored, lose interest and go back to their personal projects. |
| Eric Coleman |
Posted - Mar 11 2003 : 4:51:39 PM
I probably should have originally posted this at Lucky's, simply because of the low traffic here. So who is going to be helping me with this project? |
| Excaliber |
Posted - Mar 10 2003 : 2:18:40 PM
I would use it in an instant. I have several projects I'd like to put up actually, and know several freiends that would too. |
| Eric Coleman |
Posted - Mar 10 2003 : 08:13:36 AM
Yeah, this idea is similar to CVS and sourceforge. The difference is that this is more of a tool to help people work together (so that multiple person projects don't fail), than a versioning system. The version tracking and related stuff is just part of the package, and not the main reason for wanting to do this.
If I can't find anyone that seriously wants to help, then I can't afford to work on this by myself. I also don't want to do this if no one wants to use it, so I guess I should also ask if anyone would be interested in using an online collaboration program? |
| Almar |
Posted - Mar 10 2003 : 02:41:53 AM
An homepage which can tell what type of project it is or something? :)
Or is it more going to be an private 'CVS'? |
| Excaliber |
Posted - Mar 09 2003 : 11:26:38 AM
maybe a list of current dependencies that the project is using...just a thought |
| Eric Coleman |
Posted - Mar 09 2003 : 11:10:47 AM
At this point I'm looking for design ideas. The coding is the easy part 
This is a list of things that need further elaboration, i.e. to be designed. If anyone else can add think of something to add to this list, then please do!:
- Communications
- Bug tracking
- Task delegation
- Task completion
- and general communictation
- message board
- email list
- etc
- Time Line
- Task and bug tracking
- Milestones
- various tools to keep the project on a schedule
- Files
- Versions and comparision
- Storage and archiving
- seperate stand alone program for storage
- Tracking
- What else am I forgetting?
Remember, this is to be desgined for VB game programmers.  |
| Krisc |
Posted - Mar 08 2003 : 11:57:06 AM
I know some ASP, and I would love to know more...If you need me, I can do a bit of coding but currently I am working on a project with my friends for a contest in late March, but after that it is pretty much free time for me so I can work on this project if you need me...
btw, over at http://NeHe.GameDev.net they have a Zelda project going and it is going very nicely. {I am not in it.} |
| Rag on a Stick |
Posted - Mar 07 2003 : 6:04:06 PM
Chicken tastes better than egg. If the chicken had have come first, then man would have eaten it. A dozen eggs came first, man tried a few of them and threw the rest out. Those eggs hatched, and then there were lots of chickens and lots of eggs. It is really quite obvious |
| Peter |
Posted - Mar 07 2003 : 5:59:29 PM
Everyone knows the rooster came first :) |
| Eric Coleman |
Posted - Mar 07 2003 : 5:54:46 PM
This is like the chicken and the egg paradox |
