[Ksummit-2011-discuss] Discussion topic proposal: trivial tree -- worth it?

[Jiri Kosina]

Title: trivial tree -- worth it?

Abstract: 

One of the trees I maintain is trivial.git. Assorted things are merged 
through it -- documentation updates, comment fixes, redundant code 
removal, duplicate code removal, superfluous condition testing, etc. 
Having some larger/tree-wide patch is not an exception, and breaking it up 
per subsystem will cause even more overhead.

It also sometimes serves as Andrew's offload/backup (trivial tree in 
re-transmission mode, for patches authored/acked by appropriate code 
maintainer).

I avoid purely coding-style/whitespace fixes.

It takes me some time and energy to maintain it; mostly because many 
people are attracted in some perverted way by this tree, and try to send 
weird stuff through it (which I keep rejecting); sometimes I feel the time 
and energy could be invested in better and more productive way.

Personally I think there is value even in merging things like typo fixes, 
as it makes code easily greppable (and I am trying to merge only such typo 
fixes for which this actually holds). But I am very well aware of the fact 
that some people consider this to be unnecessary churn, bringing more 
(simple) merge conflicts than average tree does. Some may also think that 
it pollutes the history unnecessarily, etc.

I am fine with going on maintaining this tree, but I'd like to see whether 
there is more acceptance/love than hate towards it in general.

-- 
Jiri Kosina
SUSE Labs