Drafting a policy? [was: Re: Qt 3.2 requirement]

Daniel Molkentin molkentin at kde.org
Wed Jul 30 01:06:45 BST 2003

Hi Guillaume, thread-participants,

On Wednesday 30 July 2003 00:04, Guillaume Laurent wrote:
>No. Ask a user which one he prefers :
>- get a better, more stable desktop in a shorter time, at the expense of
>upgrading some basic libs
>- get a buggier, less stable desktop in a longer time, but you won't have to
>update some basic libs.

Nice set of questions ;) Why do you ask them if you choose them in a way they 
are bogus to ask? Sounds like you missed a carrier as a politican... 

Could we please get back to a factual discussion? This thread has seen way to 
much of philosophical and polemical posts. I don't say a bit of philosophy is 
wrong from time to time, but when it starts filling a deep, deep thread it's 
time to hold your breath and summarize:

- People's discipline is difficult to control
- Still is desireable to have layering and other techniques employed even it 
requires said discipline

This all sounds like somebody (or some group) should sit down and think about 
how such a thing could be realistically turned into practice for KDE. Next 
step would be to post the results to kde-policies where they should be 
discussed. If the result finds a consensus, we should try to deploy it. We 
don't really have anything to loose at that point, except for maybe some 
time. But I think we can afford that. We should test it a bit in the 3.x 
phase and - if it is useful - deploy it for 4.0. This is not meant as a 
policy that should demotivate developers. It's in contrary something that 
should enable the developers to employ certain techniques that help with the 
issue. People were able to learn about and adapt d-pointers and other BC-
keeping measures, why not this?

Marc, would you be interested in drafting something as soon as you find time 
again? Guillaume, Martijn, Aaron, Martin, Ralf and Matthias, I'd like to see 
input from you, too (and from everyone else). Maybe this is something to sit 
down and talk about in n7y? I guess we're not going to be short of beer 
there ;)

Oh, one last thing: Please refrain from answers about how pointless this idea 
is, thanks!


More information about the kde-core-devel mailing list