Gnome Article on UI Design on /.

Nick Papadonis nick at coelacanth.com
Mon Apr 22 21:03:52 BST 2002


Thomas Zander <zander at planescape.com> writes:

> On Mon, Apr 22, 2002 at 01:40:56PM -0400, George Staikos wrote:
>> On April 22, 2002 12:39, Thomas Zander wrote:
>> > On Mon, Apr 22, 2002 at 08:56:24AM -0700, Neil Stevens wrote:
>> > > On Sunday April 21, 2002 05:43, Nick Papadonis wrote:
>> > > >   - You can't make EVERYONE happy with a design decision, however you
>> > > > can please most people and others can hack a configuration file for
>> > > > happiness.
>> > >
>> > > Hard to do that when we have people who complain at every new
>> > > configuration option and who try to reduce the configurability, though.
>> >
>> > Why?
>> > not providing a config like he said seems to 'reduce the configurability'
>> > IMO. If you 'hack a configuration file' then that has nothing to do with
>> > the configurability. But a lot more with coding/checking overhead (since
>> > you have to check for it in the software anyway). The GUI (o.a. me) people
>> > will not complain about that!
>> 
>>    I challenge you to edit the KSSL config files by hand.
>> 
>>    Editting config files is basically out of the question with KDE.  What we 
>> could do is have an "advanced" section in KControl.  However it seems that 
>> most people use the defaults, so having good defaults is probably better than 
>> anything.
>> 
> The point was not to allow settings to be edited with an editor; the point was
> to cut down on configuration options without hurting certain developers that 
> _really_ want that feature.
>
> The latter does imply the former; but the reason why you are editing the file
> are different.

This was exactly my point.  I claim there is some subset of features
that 90% of users want and another subset that 10% want.

Could kcontrolcenter be in favor of the 90% of users?


In addition to the above, we haven't talked about UI quality 
assurance yet.

The more options added in kcontrolcenter, the more permuatations of
configuration must undergoe test before a 'reliable' KDE release.
Am I right?

- Nick

  




More information about the kde-core-devel mailing list