Call me Cassandra, but I bet in 2013, while sitting in front of a Linux box I'll wonder whether there isn't an utility that does all that I need from GnomeFooBarBarella without 2.7 gigs of dependencies, and then the link to it goes to berlios…
Well, it might be cheaper to create a read-only mirror of all of the projects on BerliOS and host it for free for a while than to have to re-create some obscure bit of software you need for an old system (or to recover data from an old disk, etc.)
It's a shame, because they had a great service, but it's very hard for them to generate buzz nowadays as they have been very much overshadowed by github.
It's not about the buzz. It's about addressing the needs of the users, all of the users, and building an ecosystem that provides them with the necessary tools to discover, learn, participate and improve. That and building up brand loyalty through meatspace efforts.
One cannot make the case for a model that has converted 50,000 over an eleven year span in comparison with another who has surpassed 1,000,000 users in a relatively short span. Common sense.
_It's not about the buzz. It's about addressing the needs of the users, all of the users, and building an ecosystem that provides them with the necessary tools to discover, learn, participate and improve. That and building up brand loyalty through meatspace efforts._
Not that this hasn't been discussed to death here at HN, but how do you explain the success of github comparing to bitbucket for example?
_One cannot make the case for a model that has converted 50,000 over an eleven year span in comparison with another who has surpassed 1,000,000 users in a relatively short span. Common sense._
I don't think a github user has the same role as a berlios user, at all. So those 50.000 vs 1 million are not comparable.
Sure github is centered on distributed development, but among all the forks that spawn all over the place even on pastebins (gists) or wiki pages, which percentage ends up pushing code back to the parent?
Don't get me wrong, I like github as much as any other developer do. But I fail to understand why must we refrain from state that things such as hype do play a major role on github success. Is it just because github is a huge success and does provide a good service for a fact? When did both became exclusive?
This is like assuming that a popular band is either good or a hype. Why can't it be both?
My point: berlios, sf, savanah, never really aimed towards being a social network. There's nothing wrong with that approach.
I guess I should've qualified my statements first by alluding to the fact that unless you're a Ruby dev or come from that kind of background, Github may seem foreign and weird in a fuzzy way. Github is in many aspects a reflection of a particular community's philosophy which may not fit with others.
I guess one has to consider the lifespan of hype. I've been a paying customer for two years now and it doesn't get old. I don't follow the Githubbers around, I don't wear their schwag and I only go to a random drinkup if it's convenient but if there's anything I tell emerging developers it's "Get thee to Github". They're always innovating and sharing with not just their customers, paying or otherwise, but also with the developer community as a whole. Their efforts transcend just a couple of commits.
Also, they're making money as they're supposed, with hard work, innovation and a commitment to excellence. To me, it's exciting to be a part of their experience. It's empowering. And if I look at it from a financial perspective, I make money from using their service.
It all goes back to asking if using a particular service or product is going to get me laid. Github will. Unfuddle and Bitbucket not so much. Hype and buzz is something nice to have in a fast-paced world, but being indispensable is awesome.
Forks, pastebins and wikis? That's called opening the door to the reclusive developer communities of the past. Do you know how hard it was to get a dev to fix a bug in their code ten years ago? Nearly impossible. Patches unwelcome.
Sure, a service and technology will emerge in the coming years that will surpass Github, just like any other business, but I'm willing to bet right now that they'll be based on Github's model with a lot of lessons learned from that entire community.
"It hurts us all the more that we are forced to discontinue the hosting for BerliOS by 31.12.2011."
Is hosting for something like that really so much of an issue? I would have expected Frauenhofer to have plenty of serverspace around to keep it up for a little longer.
The question is: is Jörg Schilling (who maintained BerliOS) now unemployed because FOKUS decided to close BerliOS, or do they close it down because Jörg Schilling left FOKUS and nobody's interested in maintaining the mess he left behind?
(j/k)
Isn't a couple of servers enough to host such a service? This is the matter of 80€/month... (at least this is what I'm paying right now for 2 i5 servers with 1Gb/s unlimited traffic and 1TB mirrored drives).
Yes, but it's the Fraunhofer Society... I worked there for a little bit and you really notice that it's partially state owned and usually full of people trying to get their PhD and then leave. That also means that they get funding for writing papers, not for actually maintaining something.
That also might be why they have the weird chart on the frontpage (http://i.imgur.com/TTSFn.png), it's probably straight out of some report :-/
I guess my 'rant' kind of shows how annoying I think this kind of institution is when it comes to dealing with something as quickly moving and interesting as the internet. Although they employ a lot of people and have a large sum of money, they don't seem to be able to create any of the cool things people actually use or really innovate. The main focus is usually on something that they're able to sell to somebody managing somewhere at the State. Something like "we're developing the next b2b exchange format for selling services in the cloud." This usually means writing a lot of papers and reports and projects, looking for somebody big in the industry to 'work with' and then releasing a somewhat crappy demo implementation of it that then dies after the last paper has been written.
Sorry about the rant, I somehow get annoyed when I see that much time and money being wasted. I'm sure they do some cool stuff, but the majority of it just seems like a fire to throw money into.
