Fedora Core 6 and beyond – the way of the future
So FC5 test3 is right around the corner, this means that FC6 planning is hopefully about to begin. I figured I would throw in a few hopes for FC6 and the releases beyond.
Firstly, anaconda should support multiple repos during install, so to allow installation of Extras packages.
This then allows us to unleash KDE, Red Hat and Fedora have long been acosted with alligations of maltreating KDE. Letting the KDE community take a hold of their baby will allow them the same freedom and control that the AbiWord community now have over that package – despite early concern about the move, it has now turned out really well, better packages of AbiWord for the users (here I care especially about one AbiWord user – me) and greater freedom to upgrade for the developers. Hopefully the KDE community will enjoy the same experience rather than consider this move be a move to make KDE a secondrate citizen.
Replacing the Init system as was planned for FC5 but didn’t make it.
Upgrade to GNOME 2.16, if we keep the 9 month cycle, this would give us roughly 3 months of time to polish the desktop. This would, hopefully, give us a really nice work environment while not being way out of date.
Improve memory use, while always a worthy aim, if we hope to make fedora run well in the OLPC machines, we need to trim the fat. SystemTap might be in a state to help us find problem areas.
gcjwebplugin, this is the single most needed gcj enabling technology I see, not azureus, rssowl, etc. This is what users crave, working java-applets. This might be beyond the FC6 scope though.
XULrunner, I absolutely hate the current system where I have to have mozilla installed to enjoy the superior Epiphany interface.
Hardware database, we really need a program to add to the development cycle that gathers information about user hardware in runtime – HAL based of course. This would greatly aid development I think as well as aiding users in picking the best possible hardware combinations for Fedora. This has some issues with user consent, but I think most users would consent if they knew it was safe and it helped the development team.
I would really like to see Fedora ship TOR, disabled by default, but with a nice little interface in system-config-security enabling “anonymous connectivity [x]” (an UI even Havoc would like I hope).
Open up the translation access, as a translator I’m rather in doubt as to how I commit changes to the various modules and how to get the latest po files. Setting up a l10n-status.gnome.org like page on fedoraproject.org would be extremely helpful.
Switch to Tango, I don’t see how creating an entire brand of icons is a good idea, Tango sets us up with a ton of good icons and even more importantly a good naming standard. Diana, who does lovely work btw., could maybe contribute to Tango instead?
Better torrent support, it’s good enough to distribute Fedora Core, why can’t we have support in the distro for it – it’s great technology. The official client btw. is the finest compromise of options and simplicity of design I’ve found, I would love some better GNOME integration though – gnome-btdownload doesn’t really qualify for my use, I can’t easily set the port to use and I can’t set a global max upload rate.
A move to Elektra, this would really be controversial, but I think Fedora would be a good place to vet such technologies when they get good standardisation. This definately isn’t FC6 material though.
A nice rhn-applet replacement that uses yum/pup to notify the user of available updates visually using libnotify and the notification area. Ubuntu has a very nice little program to do this, maybe we could look at that work for ideas – for a user it works really well at least.
It would also be great to see MonoDevelop and other Mono based applications entering Extras, but the pace seems slow right now for reasons I can only figure as lack of interest or lack of reviewers of the spec files. I don’t really know how to fix this aside addressing the issues sighted – I guess this is a low level entry project for someone in Fedora.
Get Palm syncing working correctly (as well as other device syncing, but I have a palm so that’s what I care about personally), it has been broken rather badly for a while now without a real fix – opensync seems to be the replacement of choice.
I would also like to see something like the security as debugging I blogged about earlier.
And finally, collaborative bug detection for the development cycle, this would capture a lot of bugs I hope – bugs that would otherwise be hard to spot or pindown.