Divorce the 3 tiers [was Re: My resignation]

Hui Zhou zhouhui at wam.umd.edu
Tue Jul 13 13:06:11 PDT 2004

On Tue, Jul 13, 2004 at 09:26:39PM +0200, Matthias B. wrote:
>On Tue, 13 Jul 2004 11:32:42 -0400 Hui Zhou <zhouhui at wam.umd.edu> wrote:
>> How about divorcing the 3-tiers (which was in my mind all along)? 
>I thought about this, too, and I think it would be a good idea. This would
>be a development model more like that of Mozilla/Firefox. They share the
>same repository, they share most of the code and fixes/changes may be
>merged in the one or other direction, but Mozilla is not meant to become
>Firefox at some time or vice versa.
>In the LFS case, committers for unstable would be a superset of committers
>for testing which would be a superset of committers for stable.

I even believe we should destroy this superset structures. The 
different branch in LFS has intrisic conflicts in policy and 
interests. It's quite difficult for a committer to commit in one 
branch without the influence of the other branch. For example, the one 
supports belfs gets annoyed on any accusation of including certain 
packages nomatter it is being accused on the unstable branch or the 
testing branch. And the one who supports discussion will accuse the 
inclusion of certain packages no matter it is included in the unstable 
branch or the testing branch. And they argued and hurt feelings.
The things are they should not be conflict. There are developers like 
certain packages in, and there are developers do not; more 
importantly, there are audience/users who like to see certain 
developments directions, and there are users would like to see the 
other. So make it seperate and both the developers and users will be 

It's just seperate in committers. They are still equal in mailinglist 
discussions but different people will the final say on different 

Hui Zhou

More information about the lfs-dev mailing list