Lots of energy expended on Mono the past few days, attempts to violate bodhi in ways unspeakable and abominable to humanity turned up a bug. Not quite the progress imagined but now our users should have a nearly warning free Mono experience starting up Banshee, I say nearly since Boo has yet to be fixed up. I am unwilling to start deeply tinkering with this to bring Boo 0.9 into Fedora as this is not my package and it is rather complex, also the existing Boo package no longer compiles on F10 for some reason.
Most importantly we now have our own mailing list, to which I shall soon send out a proposed action plan. There is a lot of ground to cover to make Mono all it can be for Fedora users. E.g. our packaging guidelines have not been touched in years and definitely need some reviews and updates. Additionally there is work to do to bring a complete stack to our users and developers. Much excitement ahead.
Jeremy Katz kindly informed me that he has a patch to allow splitting out the debuginfo packages correctly for Mono which will bring much needed size reduction to our packages.
Also I am writing up documentation on how to correctly invoke Monodoc and having generating documentation where available be a required step for the packages covered by the SIG. This will require a full review of all our packages, and would make for great work for a beginning packager to get involved with. I am hoping to be able to setup tasks all the way from willing testers to experienced packagers, in the hope that everyone can contribute to the best of their ability and time. I find it important to be welcoming to any new talent however small, just something like having a group of people to turn to to do test installs of packages and confirm that bugs are closed or still valid is extremely helpful.
Finally Debian and Ubuntu are doing some interesting work patching Mono to reduce size and increase performance by forcing a full transition to Mono 2.0 which we definitely need to examine more closely.
Now that Fedora 10 is out of the door, I´d like to look ahead 6 months to things I wish to see on my Fedora 11 desktop (and beyond).
- The beginnings of a consistent all enveloping theming, I feel that Fedora develops software that improves gradually with each release and our artwork starts from scratch pretty much every 6 months. I would love to see a consistent professional look that we can carry for a 2 years or so with gradual improvements. Something like this might also be able to push innovation in the technology we have so we can do the things visually gifted people can invision. OpenSUSE and Ubuntu already do this and I think it does them a world of good in terms of presenting the technology we have to the users.
- Acceleration for my r600, with work progressing the way it is on the radeon driver I don´t know if this is in the scope for F11 but I have my hopes. Especially since unaccelerated my desktop is pretty useless perticularly for playing back video (HD content makes the matter even worse).
- Suspend and resume for both my machines, I mean where does one even begin to file a bug. Really good guides to weeding out the information developers need would be excellent. I feel bad generally just filing bugs with the title ¨dies on resume¨ it just seems like poor style and a technique likely to waste developer time which I am not a great fan of.
- btrfs preview support, I am a glutton for punishment, this seems like cool technology and with the way work progresses towards 1.0 I am hoping we can see a preview similar to what we have with ext4 currently. Not sure F11 is the right time frame but I have hopes ( idealistic vain hopes but still hopes).
- Banshee as the default media player, with it´s aggressive development plan and responsive developers it seems to make a lot more sense to me as our default. It also presents much nicer bling and cooler features. I like that we can count on releases happening often with major improvements happening each time. E.g. in the pipeline now is things like rockbox integration meaning we might be able to offer users a way to convert a player to use only free software and free formats soon (something that seems like an issue for some due to codec patents).
- A netbook spin, as is evident from existing data nebooks are not a good target for our default desktop. The screen size is not perfect for the existing interface (maybe time to look at sugar here?) and they have different requirements for their partitioning to not wear out the SSDs needlessly. Many little tweaks to be done.
- An equvilant to SUSE Studio I am starting to believe heavily in applicances and while Fedora was early to the table making this possible we haven´t really taken the work much further since then.
- I´d like to see our Mono stack improve, there is a SIG now but we have yet to plan the first meeting and there are other failings in this area such as getting more maintainers and getting a complete stack so developers can rely on things being there instead of having to bundle them. One important step would definitely be getting Moonlight into rpmfusion since RH Legal have blocked it in Fedora proper. (this item I will likely invest a lot of time in during the next 6 months along with Banshee as always).
As a personal todo list aside the Mono work, I would love to do the developer/feature interviews as podcasts for the Fedora TV feed. I just have no idea where to start but I have a lot of time on my hands and it seems like a good time to do things like this now that free tools like Jokosher are coming of age.
I tend to stumble over things during randomly reading the daily news, things that really get my juices cooking.. and no I am not talking about nudie pictures. Cool technology, here’s the little list I gathered during my vacation:
Codethink Ltd is doing a revision system based filesystem overlay called WizBit, what this enables is something I have wanted for a long time. All the things a DVCS does for source code but made available for regular files and integrated in the desktop. One simple usecase for this would be saving documents automatically, the user does not have to specify a path or a name since those now are just attributes of the files entry in the underlying DVCS (in WizBit’ case Git), they can be filled in later. The user can thus be safe knowing his work, the thing he really cares about as opposed to the file, is contained for him even if he forgets to click the save button (long term maybe he wouldn’t even need to, the file name and path could be filled in automatically by the application in use). I think in all this recent talk about which system to pick for a project is overlooking WizBit’ excellent example, it is not the DVCS, it’s how you interact with it and integrate it with the job you are trying to get done. E.g. as a Fedora contributor I don’t interact with CVS, I use the scripts provided to help the maintainers that wrap around the underlying system. Regardless, WizBit looks very cool, I look forward to playing with it and seeing how it can play a role in making Fedora a more attractive desktop for me and users everywhere. One thing I wonder about, if we do something like use this technology to do filesystem overlays to seemlessly combine a remote store with the desktop. After my recent harddrive failure I have been wondering about putting my data on something like Amazon’ S3 but this type of service currently does not tie in well with the desktop, I don’t get the experience I desire – namely having a minimal system on my machine but still having it look in every way that all my data is there when I have internet connectivity (maybe even with clever caching to have popular files locally).
Red Hat, everyones favorite Linux vendor, has an interesting research project called razor, which aims to unify parts of rpm and yum to have them share an efficent storage, as a user I hope this is one of those changes I will never notice outside maybe a performance increase. Not every change needs to be wizbang, with PackageKit being the way we interact with the package system, the average user will never notice razor but it will make his life better.. hopefully. I look forward to seeing where this goes, being a research project nothing ever needs to come of it which in a way fills me with joy, these little playful projects are part of what is so much fun about computers.. exploring.
Another Red Hat project, Spacewalk the freed future of the Red Hat Network Satelitte product. I’ll probably never use it, I have very few machines but Fedora isn’t all about me. I can see this becoming a really powerful tool to position Fedora for admins. With all the work the Fedora community puts into making Fedora rock on the desktop, I think it’s nice to see that we also cater to the admin population. Fedora is a platform which you can make into what you need, the defaults for the spin we promote might be set for a desktop but that is easily changed with the tools we have available.
Just 3 of the projects that got me excited, it is one thing I love about Linux, there is always something interesting going on. Just look at the F10 Feature plan as it grows the next month or so, enjoy looking forward to seeing it in action on your Fedora desktop soon and maybe you can catch a glimpse early on Fedora TV vod/podcast. Which is probably if I have to mention something which I am personally the most excited about Fedora TV would be it. I see it as a good advertisement channel for Fedora and a good way for us to show off new features, interesting bits of work, maybe some day conference talks and video interviews. I would love a good Fedora insider podcast to compliment it for the heavier topics, lenghty interviews and that kind of thing.