> I would say that if one were to rank these features in terms of value to > Haiku, file system compatibility > > with Linux is fairly low ranking when it comes to features that new Haiku > users would want. But above either of these in terms of value to Haiku is the possibility of another developer, imo. GSoC is really more about getting people interested in and acclimated to the open source environment. The code that we get out of it is a nice side-effect, but if the student stays with Haiku, then their future contributions are not limited to any one area. The best way to get a student to stick around is to make sure they enjoy the summer, and if they're slogging away at some boring task because 'the users' deemed it important, chances are they won't come back. > My criteria of high value is: will it attract new users, will it help keep > them using Haiku. Right now a Linux/Haiku user has the options of > ext2/ext3/ext3. So right now there is a good substitute for btrfs. While > ACPI is not a critical feature, it is not here, there is no real substitute. There's no substitute for a usb-powered desktop thermonuclear reactor either, but that doesn't mean we should build one. Furthermore, when it comes to writing to a btrfs disk, there's no substitute for a write-enabled btrfs driver. At any rate, both projects are useful, but the best project is one that the student will enjoy and see through to completion. Anyway, maybe our student has some input or questions? --Alex