Originally posted by: aidanjm
Originally posted by: drag
I don't think that Beryl will actually end up anywere important. I bet that it's most likely that it will fail and that people working in Beryl will take what they've learned and go off and do something that is actually cool.
why don't you like beryl?
I donno. I just seems that it's a project that has set itself up for failure.
There is no good technical reason for forking for them. Everything they wanted to do could of been done through Compiz's plugin system.
see their announcement here:
http://forum.beryl-project.org/topic-4591-beryl-informations-announcement
For instance they say they forked because they didn't like gnome dependancies. Compiz has no gnome dependancies. Compiz's default set of plugins uses gconf for configuration engine and has other things to integrate it into a gnome desktop. Those all are just plugins, there is no hard dependancies on Gnome.
And other things.. It's all can be done through plugins. So there is no technical reason for the fork. Nada.
By forking for no good technical reason they are not only taking on the burden of their special enhancements.. but they are taking on the burden of maintaining and developing their version of compiz all on their own. That's a big deal and means that they have to put in much much more work then otherwise they'd have to do.
They say they like communicating through IRC and Forums... but that's not what Compiz uses and never used. Like 90% of other open source projects they used a combination of bugtracking and mailing lists to discuss issues and submit patches.
The reasons are pretty obvious if you ever seen how a open source project works.
With IRC discussions none of it gets recorded typically. If they are they are nearly impossible to follow unless you know those people.
Forums get recorded onto the web.. but it's nearly impossible to follow them. You have to praticipate. Ever tried to read a forum topic that is 10 pages long? It's nearly impossible to figure out what is going on after a few days of discussions.
With mailing lists however you get records on the web so that people using google can find answers without making people repeat themselves over and over again like on a forum or irc channel. Also most online things do a good job of tracking threads and forks in the discussions. People change topics as the talks diverge and it's usually pretty easy to figure out everything that was going on.
With bugtrack-style sites you get a formal accounting system for bugs. Otherwise what happens is people get distracted and miss bugs and forget about them. There is no way to figure out what sort of problems or bugs are popping up statisticly wise. They have classifications for bugs, a triage, were more important things are tracked and stuff that isn't so important get put on wishlists.
the advantage to IRC and forums over mailing lists and bug tracking is purely social clik thing. Like high school stuff were people get to know each other and everything is tight knit. You have newbies and outsiders and things like that. That's more then likely why they didn't like working with Novell. They weren't significant enough, didn't feel important just mucking around with their own custom plugins.
However I don't hate beryl or think it's trash or anything like that. I don't dislike the people behind it or anything like that.
This sort of thing happens all the time in Linux and open source. Beryl and Compiz are mearly fancy Window Managers. There have been literally hundreds of different Window managers for Linux. The vast vast majority of them are forks from others or people who tried to be part of one group, but couldn't get along so they forked or tried to start over from scratch. And most of those failed. There is probably only a couple dozen or so that are significant in any way and the vast majority of people only use maybe 6 or so.
What will probably happen is that Beryl will continue for a while. Then they will get tired of maintaining their monolythic-style codebase and people will start talking about getting rid of the old compiz cruft or doing a big redesign based on all the experiance they have gained. Then there will be a falling out about what the focus should be and such.. eventually you'll get a few people going on with their own fork and it'll all just kind a peter out.
Out of that you'll have some developers that learned what went wrong and they'll use their own experiance to start new projects and eventually one or two of those will turn out to be something that will be pretty kick-ass and may end up being something very important. Other guys will end up making important contributions elsewere.
This is just how these sort of things tend to go.
