Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-22 Thread Tom P
Hi Stefan, that's a fantastic offer! The jury is still out on the choice of DVCS, but I hope we finalize it soon and we can put your server (and plenty of bandwidth!) to good use. Thanks Tom On Sun, Sep 20, 2009 at 2:23 AM, Stefan Seifert n...@detonation.org wrote: On Saturday 19 September

Re: [Flightgear-devel] gitorious repositories for FlightGear and

2009-09-22 Thread Martin Spott
Tom P wrote: The jury is still out on the choice of DVCS, but I hope we finalize it soon and we can put your server (and plenty of bandwidth!) to good use. I'm sure we'll manage to arrange a couple of read-only GIT servers for use by Joe Average. A round-robin DNS will do the rest. In

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-19 Thread Kishore
On Friday 18 Sep 2009 7:26:12 pm Curtis Olson wrote: Github wrote us back saying: Git doesn't work very well with large amounts of binary assets. They didn't offer further explanation to where the problems might be? Maybe they were just putting the brakes on and didn't want to offer to

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-19 Thread Tom P
Hi I've talked about the FG data repository to Johan, the main guy behind gitorious, and in his own words: Diskspace in itself isn't so bad, it's really more about the bandwidth. How active do you think the repository ends up being? It's difficult to come up with such an estimate. Any idea?

[Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Tim Moore
By way of experimentation, and to move the discussion about source control forward, I've put git repositories for FlightGear and SimGear up at http://gitorious.org/fg. These are somewhat special in that they include all the history of the project back to 1997, as reconstructed from the

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Curtis Olson
Hi Tim, The big hang up in the investigation has been the size of our data package. If we can find one of these established places that is willing to host a couple Gb size project, then the source will fall into place quite easily. That was one potential advantage to code.google.com ... we have

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread AJ MacLeod
On Friday 18 September 2009 14:56:12 Curtis Olson wrote: Github wrote us back saying: Git doesn't work very well with large amounts of binary assets. They didn't offer further explanation to where the problems might be? Maybe they were just putting the brakes on and didn't want to offer to

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Erik Hofman
AJ MacLeod wrote: On Friday 18 September 2009 14:56:12 Curtis Olson wrote: Github wrote us back saying: Git doesn't work very well with large amounts of binary assets. They didn't offer further explanation to where the problems might be? Maybe they were just putting the brakes on and

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Alex Perry
On Fri, Sep 18, 2009 at 9:51 AM, Erik Hofman e...@ehofman.com wrote: One thing I have been wondering since this discussion started; Google seems to have found a nice way to add small diffs for binary data[1]. Maybe they have incorporated that into their repository? If they have, it won't help

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Nicolas Quijano
binary data != binary code. Binary data is ALL the images files included in aircraft, possibly some models format supported by OSG (like .ive), all sound files, that sort of thing. CVS is notoriously bad at handling these files, and very inefficient at doing so. I believe that's what Erik meant.

Re: [Flightgear-devel] gitorious repositories for FlightGear and SimGear

2009-09-18 Thread Erik Hofman
Ah ok, I didn't realize it is aimed at machine code only. Erik Nicolas Quijano wrote: binary data != binary code. On Fri, Sep 18, 2009 at 1:07 PM, Alex Perry alex.pe...@pamurray.com If they have, it won't help us. We're not distributing blobs of x86 machine code.