Project Timelord Announced

[kubicle]

http://www.kubuntu.org/news/timelord

Specification (pdf):
http://people.ubuntu.com/~apachelogg...nouncement.pdf
Original brainsorming document (also pdf):
http://people.ubuntu.com/~apachelogg...ntu_Issues.pdf

Some earlier discussion here:
http://kubuntuforums.net/forums/inde...opic=3107135.0

[MoonRise]

Re: Project Timelord Announced

Thanks for the info. Saw the main Kubuntu site about this but needed more info, especially the Brainstorm doc. Thanks!

[kiraitachi]

Re: Project Timelord Announced

Nice i love it thats y i choose kubuntu hope its as good as karmic koala or better

[lmilano]

Re: Project Timelord Announced

Overall, I love to see the urge to improve Kubuntu and make it more polished. The one unsettling thing I read is about not using Python for config GUI's, which is a shame if you ask me. Developing in Python is so much faster than C++, and no less elegant or powerful, and getting folks help fix broken stuff is so much easier if you use python (they can change a line and try, as opposed to setting themselves up with a dev environment, recompiling, installing, etc.

I would love to see more interaction between devs and the rest of us. Many of us spend a lot of time helping others, and we have a sense of what's broken or not, generally speaking ... so I think it would be important to open that channel of communication.

[kubicle]

Re: Project Timelord Announced

Overall, I love to see the urge to improve Kubuntu and make it more polished. The one unsettling thing I read is about not using Python for config GUI's, which is a shame if you ask me. Developing in Python is so much faster than C++, and no less elegant or powerful, and getting folks help fix broken stuff is so much easier if you use python (they can change a line and try, as opposed to setting themselves up with a dev environment, recompiling, installing, etc.

I would love to see more interaction between devs and the rest of us. Many of us spend a lot of time helping others, and we have a sense of what's broken or not, generally speaking ... so I think it would be important to open that channel of communication.

Well I for one am all for using less python...and I've written some python code myself, so I don't think it's a bad language...it's also an excellent "first programming language" to learn.

Firstly, there is no way around the fact that python is not as efficient as C/C++ code, which is usually not a big issue with simple applets or config dialogs, but a number of python daemons can start to eat up your memory pretty quickly.

Secondly, (although clearly prejudicial) I tend trust C/C++ programmers more than python programmers to write good code/software, even though it's quite possible to write good and bad software with any language.

In short, I don't mind python in configuration applets or when something needs to be cooked up quickly (to meet a deadline, for example), but I really don't want my "operating system" written in python.

[lmilano]

Re: Project Timelord Announced

Sure, anyone using Python for kernel development is nuts :-) I was actually thinking of config dialogs and GUI's where you can use the KDE/Qt libs but quickly code in Python. Python also gives you garbage collection which is a blessing. For daemons, I thought all the low level stuff was decided for all *buntus at the same time. I am not sure about memory consumption bit (I just don't know how to quantify it, it is an interesting point and I am sure it must be well known/documented).

Incidentally, I love the direction Kubuntu has been taking lately, of reusing the same basic libs as Ubuntu as much as posible, like with KNetworkManager, KPackageKit, etc. I'd like to see the *buntu family grow organically as a desktop agnostic base of libraries and desktop/library aware GUI's

Cheers!

[kubicle]

Re: Project Timelord Announced

Incidentally, I love the direction Kubuntu has been taking lately, of reusing the same basic libs as Ubuntu as much as posible, like with KNetworkManager, KPackageKit, etc. I'd like to see the *buntu family grow organically as a desktop agnostic base of libraries and desktop/library aware GUI's

Though I generally agree that desktop agnostic backends with different GUI frontends is a good idea forward, this is sort of a double-edged sword for KDE users.

The problem is that all the "universal backends" seem to originate from the gnome camp (packagekit, policykit, devicekit etc. come from redhat, for example), which means they make the backend and a gnome frontend, and let others worry about making a kde frontend for it. That means the KDE frontends always lag behind the other parts. This also happens when changes to the backend are made, they make sure the gnome frontend works and let KDE people play catch-up, which has lead to breakage (knetworkmanager).

And it seems to me that the "KDE-frontends" (often written in a hurry) are often just simple ports of the gnome tools, which feel a bit alien in KDE...and I feel that all this is a bit retroactive. I'd also like gnome come up with some frontends for KDE backends, and not always the other way around.

Oh well, I'm drifting way Off-topic, just had to get that off my chest

[dequire]

Re: Project Timelord Announced

From the "Preaching to the Choir" department - My call for action, from a slightly downstream perspective.

Some Thoughts on Project Timelord

[lmilano]

Re: Project Timelord Announced

From the "Preaching to the Choir" department - My call for action, from a slightly downstream perspective.

Some Thoughts on Project Timelord

And not so downstream I think we are missing a space where informed people can share thoughts on the development of Kubuntu. To me, leveraging on the "default" Ubuntu is a must for very many reasons, one of which is that they have plenty of resources. But, obviously, when you ship a distribution (Jaunty) where the default package manager (KPackageKit) is plain broken, something needs to be reconsidered. Not in the decision itself, because you can't go back in time, but in the decision making process. It seems to me that decisions are made by a very small group of people. These are wonderful folks doing the bulk of the kubuntu specific work, but I think they could use some more perspective, which will also come if they provide a space of communication.

Incidentally, I would love to help more than what I do (currently, forum support). I know how to write and fix code, I write advanced c++ code all the time at work. But my 14 hours working day and busy family life make it a bit of an impossible mission!

[richard]

Re: Project Timelord Announced

Excellent! As a fairly undemanding user my experience is that kubuntu is not really a separate distro but Ubuntu with a KDE desktop grafted on in a not too elegant or functional manner. I love KDE.