[clug] A question of regression. [KU]buntu

Randall Crook rcrook at vtown.com.au
Wed Nov 18 03:10:25 MST 2009


Just trying to get a feel for the policies Canonical have in regards to 
bleeding edge. But it goes beyond that to how the OSS development teams 
run their projects. But most importantly, how laymen perceive Linux 
because of the policies and project management.

Two examples to show what I mean.

Amarok.

The way Canonical have handled the KDE teams management of their project 
has not been a good advertisement for GNU/Linux in the layman's eyes..

My wife and both my daughters have used Amarok 1.X extensively. The 
girls iPods and the wifes iRiver have been exclusively managed by amarok 
for a couple of years now. So when I upgraded to Jaunty all hell broke 
loose. Amarok 2 was a serious and major regression. So much so that my 
family refused to use it at all.

I was then forced into a decision, Downgrade or go searching for a 
replacement or some nice person who has built a 1.4 package, or build 
the now abandoned code myself. For me not a problem, as I am not a 
layman. I quickly found someone who had 1.4 packages and updated the 
apt.sources and installed. But that's not the point... A normal end user 
would have been stuck, and most likely would have gone back to windows 
stating Linux sucks!

Now there is two layers at work that made this bad move come into play. 
The development teams valid decision to drop work on the old code so 
they could focus on the new, and Canonical's policy to drop 
un-maintained code.

Both are valid and sensible policies. But it does nothing for the 
perceptions of those who Canonical are trying to attract to their products.

GDM

Once more stung by regression. I upgraded to the Koala just recently and 
found two things are broken because of the seemingly relentless need to 
have the very latest on the system, even if it loses functionality.

The new version of gdm is smaller cleaner and has at least one bug and 
no easy way of configuration. The quickly knocked together gui canonical 
threw in is almost useless. Once again, not a major problem for a geek, 
but a total disappointment to even a power user.

It seems to me something has to give if Canonical want to actually break 
the desktop hold MS has. I think a compromising stance and a call to the 
community for a little more support is needed.

General Perception from a geek.

I was a huge fan of KDE until KDE4 was thrust upon me by canonical and 
the KDE team when kubuntu went full KDE4 and dropped 3.5.X. Through 
necessity (Needed a working environment that delivered everything I 
used) I went from KDE to Gnome. And in the almost 10 months i have been 
using Ubuntu I had grown to love the gnome environment. But now I am 
finding the same policies and attitudes creeping into the gnome world too.

So what does it all mean? I think the old habits die hard. The old idea 
that if it isn't right you can fix it yourself is still permeating the 
wrong parts of the OSS universe. Yes, that is the case for a lot of 
Linux users, but if a company like Canonical want to survive and 
prosper, they need to focus a little more on consistency and usability. 
Replacing a fully functional tool like amarok 1.4 with a barely usable 
Amarok 2 with less than 40% of its predecessors functionality only makes 
them look like pure idiots....

And even worse, it made me look like an idiot to my family, as I then 
had to teach them how to use a totally different piece of software to 
manage their mp3 players...

But the real kicker is, how do we fix this? I am not smart enough to 
program and help the development teams. I do provide as much feedback as 
I can, though I am usually there after a whole bunch have already 
provided feedback. I do love Ubuntu/Kubuntu and I will keep using it and 
doing what I can to contribute, but I sometimes despair at some of the 
silliness I see from Canonical.

So what do you guys/gals think can be done to try and prevent these sort 
of issues from]\ turning people away from [ku]buntu and Linux in general?

Randall




More information about the linux mailing list