How many people are developing the core Vanilla version?

Polo

Hello. I wonder if it is just Mark that does the core Vanilla version.

I know he wants the credit, but why don't let other people help you with that? That is what Open Source is about.

Minisweeper

Developing on his own is nothing to do with wanting the credit. He's the only one who's developed it so far. Open source is all about releasing the source code openly, not developing in teams. As far as I know he has nothing against working with people but it has problems (particularly when you're not in the same geographical location) - mainly that it makes it more difficult to keep control of the application and retain a complete knowledge of what everything does and where it is. Not to mention that, particularly up till the release of v1, Mark had a pretty specific idea of what he wanted from vanilla and where he was going with it and in many cases it's easier to work towards goals on your own (besides the obvious fact that the development of vanilla took Mark a massive amount of time, effort, and money, which would have been divided if he worked in a team). Then there are issues about team reliability and such ofcourse...

Polo

I can guarantee you Minisweeper that Open Source is about collaborations and of course releasing the source code openly. It is just that, the Vanilla development is slower just because Mark is the only person developing it.

Can't Mark say his goals so possible new developers can develop with his kind of goals? Isn't there a Road Map for Vanilla? If not, then it should be!

leason

Polo, as an author of an open source project I can say that your opinion is a bit skewed. First of all, Vanilla is Mark's baby - I don't blame him a bit for not opening up the repository for global commits. Secondly, open source means you can view the source code of the application. It means you can download the project and make changes in your individual installation. MySQL is open source but I don't think you can make a change and then check in your fix. There are projects that allow public commits, but they also have teams setup to review all the changes and weed out the crap. I can also tell you from experience that most of the changes that come in are not going to be up to the standards of the original developer regardless of whether or not there is a good public road map.

Polo

PunBB was Richard's baby, and just recently he admitted that working alone was badder for the project than opening it up for other developers. And the development of PunBB is going better now.

WallPhone

I believe Vanilla's strength is a combination of its simplicity, consistency, and extensibility. Its a tripod, so to speak. Adding more programmers to a project introduces complexity and reduces consistency. Differing coding styles become distracting noise to those who are learning Vanilla, and much more organization work is needed to manage the project (project roadmap, coding conventions, submission guidelines etc.) The project becomes a product of mob rule. The community is free to create add-ons, which easily enable modifying in most any way imaginable. This would be harder to accomplish with a larger learning curve if consistency and simplicity are reduced. Patches are also submitted all the time, I can think of at least a dozen occasions where somebody submits a fix that becomes part of the core. I also recall Mark mentioning that a select few do have write access to the code repository. There is less incentive to create modifications if features are bundled with the core, and consequently fewer existing add-ons to use to learn or modify. Sure, a lot of people running forums don't know how to program, but that is what makes the community so great--because a lot of those who do know help out. Maybe PunBB is better now. I wouldn't know. I use Vanilla because it does not try to be like the other BBs.

davidh

For projects where simplicity is the goal, having a large subset of developers distributed across half the globe is NOT what is best for open source. The beauty of open source isn't in collaboration, it's in derivation. Meaning that if someone wants to create Vanilla-Ultra, they can simply branch off of Vanilla's code and start a whole separate project to fulfill a different need. I would daresay that the primary reason open source still hasn't caught on for the non-geek crowd is that too many people are making decisions, resulting in convoluted products that confuse the average user. Look at Mozilla/Firefox and how many years it's taken them to get to a point where normal people are actually using their software, despite it has always been a superior technology. For small projects, it is sometimes better to have one person calling the shots, and writing the code. It's more maintainable, easier to get quicker updates, more coherent to read and use (since you only have to decipher one person's style). And Mark isn't afraid to integrate some addons and patches into the core. If any developer could do so imagine the slippery slope Vanilla would become. Nobody around here is saying "where is Vanilla 2.0?" because Vanilla 1 is such a stable, extensible product that it doesn't warrant a major overhaul. But to say the main power of open source is collaboration, that simply was never and will never be true. It only allows for a greater range of software derivations and modules easily reusable in other software packages. I don't disagree with the need for a roadmap necessarily, but it's Mark's call on how he runs his project. He's done a good job thus far, I trust him to make the correct calls on how its development should run.

Wanderer

I can't speak for Mark but when I have a special project going I like to do it all myself, my way, in my own time.

Lots of people offer to help and that's admirable and I don't want to sound ungrateful but sometimes, organising a group of people who, understandably, have their own ideas, methods, principles and practises takes a lot more time than what it would if I just did it all myself.

Now I sound like I'm not a team player, that's not true, some projects can only happen when a team pulls together, in which case the organisational and communication time is justified and necessary.

Mark has done and is doing a bonza job with Vanilla. If his methods and processes ain't broke, no need to fix them.

Posted: Friday, 22 December 2006 at 9:34AM (AEDT)

Stash

Jeez, if Vanilla dev happened faster than no one would have time to understand how to create all these fantastic extensions! Also, for a version 1 product, Vanilla really is a fantastic product. I'm not going to pretend that it's perfect because it isn't, but it is incredibly refreshing to use a version 1 product that gets more right than wrong!

Mark

I would absolutely love it if there were others out there who shared my vision for the source of Vanilla. But the truth of the matter is that most of the time I don't even understand what my brain is telling me until my fingers put it into the computer.

For example, right now I've got some really really really abstract ideas for how to program swell. The thoughts are swirling around in my head, and one of these days they're going to come out into the computer and Swell will be released. If I were to try to explain what I'm thinking, you'd either think that I'm a complete imbecile or totally insane.

The search for colleagues has been on for a long time for me. I read a while back that the most successful stories online are ones that originate from at least a team of two (eg. Youtube, Google, etc). Solo projects take a much longer time to get going - but I just don't know anyone who "gets" me or wants to try.

I've been trying to convince my wife to start programming for years. So far no luck.

And mini is totally right that geographical locations makes it very hard. I've met a few people online who I might be able to work with, but we're in different parts of the world. And if you're going to have a partner in crime - it's best you've seen them in the flesh (I think, anyway).

Stash

"I've been trying to convince my wife to start programming for years. So far no luck."

I tried to convince mine to play Quake (1) back in the day and eventually got her to try it. After 20 minutes her head was aching and she wanted to vomit as we discovered she suffered from motion sickness in first person shooters. Bugger. So just beware that should you actually succeed, she may end up getting warts or going bald because of it

russlipton

@Mark - I thought your description of 'what you do' was refreshing, candid and warm. Thanks.

@Polo - I hope you won't become discouraged. I'll bet that one thing Mark does do for Vanilla - it's been discussed; not claiming originality here - is shape a first-class method for organization, integration and enhancement of add-on's. Pulling that off would be nearly as great a feat as Vanilla 1.0 itself. While I side with the keep-Vanilla-even-simpler gang, most such designs collapse under the weight of the pain involved in keeping track of extensions. Even the simplicity of the 'core' doesn't come for free if one also has the goal of encouraging a vibrant community of brilliant developers.

Like yourself.