...and we don't need nine babies either. I am talking about web software development now. A recent Techcrunch article quotes former Facebook CTO saying how it is possible to build a Facebook clone with about 250 people in 2 years. It's an interesting read with some valuable insights and yet, after many years in community software development, this assertion feels flawed to me. That's fine, surely, for we all have a little bit of a guru blood in us and can teach those smart-ass CTOs a lesson or two... but we should probably keep it to ourselves. The only reason why I decided to confer this one is its relevance to the current state of affairs in BoonEx development department.
Interestingly, we have had about 250 people in total working on BoonEx over the range of about 2 years some time ago. Hectic times to say the least, but parties we good. As for the product, well... let me just say it wasn't the brightest time in BoonEx history. We ended up with 3 unfinished platforms half-dozen sites and huge wad of crappy code in Dolphin.
We've had a small team first, a big team later, and a small team now, again [long deep sigh of relief]. What we've figured out is that after you have 2-3 really good developers new ones only add trouble. They start to need managers. Worse, they start fixing each other's bugs. Even worse, they do things differently. One may think that managers can fix this. Not really. For good developers managers are a distraction, a potential trigger for inspiration leaks. I know, I distract our developers all the time. A very small group, given a right direction, creates beautiful code, almost subconsciously churning out elegant algorithms and approaches. It's inspiration. Time, money, managers and, sadly, helpers kill inspiration.
Even with things as mundane as building a Facebook clone, it's not the amount of developers, time or funding that makes or brakes it. It's all about the ideas and direction. World's greatest innovations appeared from garages and napkin doodles, forged in a matter of days, if not hours. Moreover, I strongly believe that once fabrication stage starts taking more than a few months - it's a bold sign of a blackhole, wrong direction or a bad idea altogether. Good product will appear quickly and it's ok that it may take decades to polish it.
Dolphin was born after about two weeks of development. Seriously. We had the first sale the day it came out. In two months is was already #1 community/dating platform on the market. Ten years down, it powers hundreds of thousands sites and is still being polished. Trident swallowed 3 years of work of our best people and we scrapped it.
With Dolphin 8 development, we put forward very high requirements and only let a couple of people touch it. It's going great. I'm sure it'll appear sooner than it would if we had a developers army or millions of VC dollars to blow away.
All in all, I guess I just wanted to say that it may be possible to create a Facebook clone with 250 people in 2 years, but it is absolutely possible to create a better social networking site in one year if you go down to just 5 people.
I am an inch away from throwing in the towel.
and the member of my site doesnot comes to boonex and gives the feedback. so, to face the feedback confidently and maturely it would be nice if Boonex will include "priority/premium/knowledgeable site owners/admins" in the [by invite only] locked discussion of any upcoming logics and features. once after the discussion. Boonex can do the code. see more
reason is we dont have time to do that because we pay to buy a cms, pay to buy mods, pay for the server, pay incentives to members, burn our time/head to do something better day by day. thats it.
say, Boonex has 1million registered and premium accounts.
when the site encounters an error, i get an email. if the same email is cc'd to some see more
Overall I am very pleased with Dolphin, and plan to use it for a long while longer. It get's two thumbs up from me.
I think that is probably right. There is such thing as "a runaway project" - and they are especially a danger for Boonex I think. I'm not sure how obsessed you should be with keeping he development team small - if you do I hope you don't lose at least the "spirit of open source" here..
I think see more
Merging the design layer with the application layer: the result of 250 people who took 1 good idea and made it bad.
the column thing? see more
Much Appreciated