Post a Comment
Timevault looks promising.
https://wiki.ubuntu.com/TimeVault
A lot of things still need to be worked out but it has potential.
A more elegant solution without a gui is rsnapshot. sudo apt-get install rsnapshot.
Timevault isn't bad, but it seemed buggy when I tried it.
My home solution:
1) Set up rsnapshot to backup /etc/ {weekly,daily,hourly}.
2) Set an rsync cronjob to rsync the backup dir to another server using an scponly user so that a passwordless key isn't a big deal.
3) Profit!
http://www.rsnapshot.org
Reading this it is clear that Ubuntu is still very focused and heading in the right direction with each release.
They have covered a lot of territory, and most certainly have their work cut out for them, and I wish them all the best.
Out of all that tasty work, I find it hard to believe the most exciting thing to me was the new default theme and icons, lol.
Hopefully the constant moaning about the default Ubuntu theme is going the get quieter gradually now? All those people who never liked the Ubuntu default theme and colors now have a chance to share and propose their ideas for new themes...:
https://wiki.ubuntu.com/HardyTheme
https://wiki.ubuntu.com/HardyIconTheme
However, with a bit fear I've been reading some new, sometimes even a bit wild theming ideas... - but I'm sure that the final decisions and themes are probably going to, like they should, emphasize usability instead of changing artistic tastes and fashions.
Personally I've been rather happy with the default Ubuntu theme, especially the Human icons. I'm happy that Ubuntu, unlike many other distros, have preferred warm earthly colors in their default themes, instead of the common (and rather sterile?) blue and gray. Orange and beige colors look warm, friendly and nice, although the default Ubuntu wallpapers do have been a bit too dull brown also in my opinion.
You might have been happy with a mishmash of differing icon sets, but for a lot of people it simply wasn't good enough.
I don't have any faith in a new theme being any better, considering they're doing the initial planning less than six months out from release. Canonical needs to suck it up and get some experienced artists on the payroll and get Ubuntu to start putting its best face forward. To date it's been a complete and utter shambles.
I think it needs to include the PulseAudio project that is being worked on for Fedora's nextGen audio soundsystem:
http://fedoraproject.org/wiki/Interviews/LennartPoettering
Edited 2007-10-31 23:47
> "Oh, great another competing sound system".
It's not another weak sound system :
Demo :
http://dev.gentooexperimental.org/~flameeyes/mezcalero-pulse-demo.o...
Ubuntu's biggest challenge, imho, is how to avoid doing too much with too few people. They're already spread perilously thin; and as the new! exciting! gloss fades as the project matures so the pressure increases to do yet more. They're achieved a remarkable amount already, but while anyone can come up with a list of ideas, executing them well is very difficult. In this sense, I don't think this article is particularly helpful. In a way, it is part of the problem, not the solution. Resources are always limited - money, people, time (only six months to the next release). How best to use them is the question.
"""
Resources are always limited - money, people, time (only six months to the next release). How best to use them is the question.
"""
It's fortunate that their leader is an expert in that area. You don't run a successful business for years without being able to manage resources well.
Edited 2007-11-01 02:27
These kind of roadmaps are mainly written for adverising purposes -- they help to generate the buzz and the momentum that the distro needs. They also create the illusion that lots of new and exciting changes are taking place in the distro, even if most of the promised changes actually fail to materialize. It's all about selling the sizzle, not the steak.
Well sometimes things that are proposed for a certain version in the specs don't get implemented until two or three versions later. So they eventually do show up, but they take their time about it. It can be annoying but it all works itself out in the end.
Hopefully slickboot will actually show this time. They've done some work on it already by reducing the amount of switching to only two (I don't count grub). I'm hoping that they finally realize that usplash is a dead end and will probably never be up to the task of being a competent splash manager. Something like fbsplash or gensplash would probably look better, though I think Splashy is the way to go. I guess they really don't have to get complicated with the splash, they can do something simple and clean like OSX, bootsplash -> splash with throbber -> display manager. It should be simple enough to emulate. Basically start gdm early and end it late (until the desktop finishes loading if possible).
I don't actually know what Ubuntu's naming system is, but from the past five released one can guess that they use one letter of the alphabet at a time in a sequential order. What happens when they reach Z? Do they begin again from A, with possible confusions with older releases? (Granted, that's some eight years into future with current release schedules, so people will have forgotten.)
That wasn't a wrap-around, it was just before they started using the iterative naming scheme. As for X, they could use Xenopus or Xenops; since Xenopus is a rather ugly species of frog, hopefully it will be the latter.
The wraparound at Z will be some nine and a quarter years from now, which isn't really in the forseeable future; and they're already releasing two versions with the same letter (hoary/hardy). So try not to stress about it too much.
I wasn't saying that they did a wraparound but they were asking about H's so I was pointing out that they did it already.as for the X, okay so you have the animal but what about the word before that. I really think they will skip that one or change the scheme way before that.
Edited 2007-11-02 02:36
Letter X? Ah, but that is all too easy...
Xenophile Xerus
http://en.wikipedia.org/wiki/Xerus
Xhosan Xenosaurus
http://en.wikipedia.org/wiki/Xenosaurus
Xanthic Xantusia
http://en.wikipedia.org/wiki/Xantusia
Xiphoid Xiphosura
http://en.wikipedia.org/wiki/Xiphosura
Xylophagous Xyleborus
http://en.wikipedia.org/wiki/Xyleborus
Xianese Xysticus
http://en.wikipedia.org/wiki/Xysticus
Xenian Xenicus
http://en.wikipedia.org/wiki/Xenicus
(there could be much more, but I spare you from it...)
They could also switch to, for example, plants when they run out of animals:
Xeric Xanthoria
Xyloid Xylem...
One things I can't stand about Debian and its derivatives is this insane obsession with turning perfectly usable and capable configuration files into conf.d/ sub-directories full of mini-config files. First it was inetd, then logrotate, then apache, then pureftpd. Now they want to split up network configuration.
Their main reasoning is that a typo in the config file causes the entire file to be scrapped. Sounds to me like it's more a parser problem that needs to be fixed. Something along the lines of reading things in as sections and only invalidating the info for the section with the typo. IOW, they need to fix the interfaces(5) parser instead of splitting up the config file to work around this bug in the parser.
There are lots of applications out there that can handle these kinds of situations (Apache, Samba, the rc(8) system on NetBSD and FreeBSD, to name a few).
How it is possible that these people can't see how annoying it is to open a dozen files, try to remember what is in each file and how they relate together, when trying to change a config option?
There should be 1 (main) config file per app, not 1 config file per option (which is what pureftpd on Debian uses).
For a dual boot setup, as usually Windows was originally used to configure the PC, Windows should be used to repartition the disk before the installation and to install Linux.
1) This prevent errors as partitions are named differently in the Linux installer and in Windows, and this ensure that Windows not Linux take the blame if something wrong happen in the (risky) repartition phase.
2) During Linux installation, you can still use Windows so the user doesn't have to wait..
You might want to check out Wubi. It was meant to get into Gutsy but didn't quite make it due to a couple of missing features, but I understand it's quite usable as it is.
http://wubi-installer.org/
It doesn't repartition (I believe that's one of the features that kept it out of Gutsy) but it installs Ubuntu on a loopback inside your NTFS partition and adds itself to the windows boot manager, so there's very little risk in trying a non-lived-cd install.









