On Thursday 13 March 2008 05:12:47 Carlos López wrote:
> I have also found the next two links that you might be interested in:
>
> http://www.linuxforums.org/forum/yoper-linux-help/110774-i-got-yopered.html
> http://cutecomputer.wordpress.com/2007/06/29/howto-yoper-30-a-smart-error-f
>ixing/
>
> Things said in the above links must be taken into consideration before
> releasing Yoper 3.1. Everytime  I read this sort of stuff, I'm more
> convienced that we should pay more attention to polish this distro rather
> than build and build build lots and lots and lots for software that do not
> integrate well as a whole.... as said in previous email, better little and
> good and lots a bad...
>
> Definitely we must talk about this and everyone's opinion counts.
>
> Best regards! ;-)
> _________________________________________________________________
> Tecnología, moda, motor, viajes,…suscríbete a nuestros boletines para estar
> siempre a la última
> http://newsletters.msn.com/hm/maintenanceeses.asp?L=ES&C=ES&P=WCMaintenance
>&Brand=WL&RU=http%3a%2f%2fmail.live.com
> _______________________________________________
> yoper-dev mailing list
> yoper-dev@lists.yoper.com
> http://lists.frank-hosting.de/cgi-bin/mailman/listinfo/yoper-dev

Hi Carlos,

well.. I didn't read the links yet, though I know that specific story, because 
I read about it earlier. Your friends complain about the boot speed of 
Yoper.. ohm, are you kidding? Even with the current rather unoptimized init 
scripts, Yoper is among the fastest concerning boot time. Just take a look at 
Ubuntu or Suse and compare with those.
Of course you're right about the need to improve the user experience. I also 
agree about a repository lock and some way to enforce us a bit to work on 
user experience only. Currenlty, as you may know, I'm working on adding OSS4 
support to kmix, because I think KDE without proper kmix would in no way help 
to convince our users of OSS4.

I'd recommend creating a TODO list, specifically for user experience only. And 
here I start:

- add static ip setup support to the installer (e.g. properly integrate the 
redhat configuration tools, for now, later replace them with better ones - 
see lcc plans)
 [At least I try to give some information on how to make NetworkManager either 
ignore a device with static setup or use the provided values, I remember even 
the later should be possible]

- rearrange programs in the kmenu: I'd recommend a simple topological sort 
using the description tokens of each link and group together subgroups of 
size > 1, separated of course. Sort all items in a subgroup alphabetically.
[It should be possible to write such a script rather fast]

- a decent default theme (I'm already working on that... though I encounter 
tons of problems with a "dark " theme. Looks like general program support for 
such themes is weak. Some programs don't honour some very important colors 
at all (e.g. default font and background color at the same time, example, 
konqueror without overriden css). CSS remains to be done, make konqueror 
ignore the custom colors alltogether - anyone with CSS knowledge please step 
up. I'm currently trying to build a .kde template for the rest of the apps.
[I'm sorry to say it, but I just can't stand the current default. It doesn't 
look professional at all, with one exception: the kdm theme.]

- test functionality of the following programs again, because they seem not to 
work properly:
    - avahi/zeroconf: I remember ubuntu having an interface alias with a "zero 
conf" ip.. I guess I need to take a look at some documentation about that. 
Some package depencies are also incomplete. I report about that in detail 
tomorrow. For now, avahi doesn't seem to work at all here.
    - kde network awareness (I don't even entirely know what it is doing)
[I'll assign that one to me after kmix is done]

- properly setup kapabilities

- throw away rhgb for now until we got implemented e.g. splashy instead (least 
priority)

- track down all remaining error messages on boot up, because they give a very 
bad impression of a good system.

- properly force noisy daemons to use a specific log file in /var/log via 
syslogd-ng

- remove boot[x] logs entirely

Only a few of these are much work.. the other points are usually a matter of 
some hours only. It might be that this is only detail work, but the impact on 
user experience should not be underrated.

I also want to remember all devs about the planned meeting this friday 
13.03.08, GMT 8:00 pm.

Regards

apriori
_______________________________________________
yoper-dev mailing list
yoper-dev@lists.yoper.com
http://lists.frank-hosting.de/cgi-bin/mailman/listinfo/yoper-dev

Reply via email to