Our application is due for the Summer of Code in less than a month (see FAQ). The applications will include a few parts that I think are key and that are likely to change from last year’s application:
- Project ideas
- Previous involvement: successes, challenges
- Application template specific to Gentoo
- What we will do to ensure students stick around after the summer
We definitely need a nice big set of project ideas. The rest of them will probably be shorter answers. How do you think we should answer these questions?
Striving for greatness 
Ideas for the `Gentoo Clustering Live Media`:
– Convert to USB based media and leave the CD/DVD _and_, if possible, the squashfs approach
** Much time was lost implementing hacks to get NFS roots exportable from CD/DVD (ro)–> SquashFS (not NFS exportable) –> AuFS (which no longer supports NFS in the current versions!)
** Being USB (writable media) based, this now makes it a perfect candidate for a “cluster by night” setup.
** The environment is ready to use _and modify_!
** Many hacks had to be done due to the CD/DVD being ro, especially for the initial setting up of the Master and exporting the node’s NFS root.
– Find an alternative to AuFS for the node’s booting mechanism.
** AuFS (http://aufs.sourceforge.net/) is very dynamic and its support for NFS export has been shifting on and off lately.
:: Unionfs v2 might be a decent alternative
:: using a bind mounted tmpfs and copied-over files might also be an alternative (as used by the gentoo LTSP project)… I personally dislike this approach since we have to maintain what is required to copy over and configuration changes made on the server won’t propagate (ie: changing $NODEROOT/etc/c3.conf as nodes are added)
– `Dialogify` and enhance the configuration scripts to make the interface more “professional” and interactive (ie: scan available NICs and display in the configuration options).
– Implement parallel tools such as TAU and +…
The list obviously goes on and on, I’d be glad to ‘coach’ someone into the dev environment for the live-usb (catalyst, git repo and all) if anyone really seems interested in picking some of this up. Quite frankly, I see this as multiple projects.
Hack the Synaptics graphical package manager to use Gentoo tbz2’s. User can adjust the PORTAGE_BINHOST, /etc/portage/package.*, and global USE flags through app. This would make for an easy-to-use, graphical, binary package manager.
I was going to initially do this with the openmoko graphical installer, but I thought that it would be better to capture the desktop market first.
Let me know if you’re interested and I’ll put an application into the GSOC 2009.
Christopher,
Why not PackageKit?
I’ve never heard of it until you mentioned it actually, and then I googled PackageKit Gentoo and your name incidentally came up from last year’s GSOC!
http://www.gentoo.org/proj/en/userrel/soc/
Was this project completed? Does it work for strictly binary distros?
The only other package managers that I’ve used with Gentoo included Kuroo and Porthole, neither of which supported binary package repositories (AFAIK).
When I originally wanted to tackle this project, I was just going to heavily hack around with gentoolkit and build it as a library / backend for the openmoko installer.
I find that strictly binary distros are particularly suitable for netbooks, MID’s, phones, etc, where cross-compiling is preferred or compiling on-the-device should be done at most once (due to lack of resources).
Christopher,
Unfortunately the applicants for that project didn’t make the cut, so it didn’t happen. I’ve added it to this year’s idea page at http://en.gentoo-wiki.com/wiki/2009_Summer_Of_Code_Ideas to encourage someone to apply for it again.
I would like to work for Gentoo this summer under gsoc. I was going through the proposed ideas. I suggest you tag the ideas on basis of difficulty and skills required.