6.2; 6.3; EOL; combustible discussions

[Robert Watson]

On Sat, 7 Jun 2008, Josh Carroll wrote:

> While it would be interesting to see the response here, it still doesn't 
> necessarily provide a solution. It will still involved developers' time to 
> QA the user-submitted patches, so it won't entirely eliminate the additional 
> workload for maintainers. There is also zero (enforceable) accountability. 
> If X people commit to this, what happens when only a fraction of them 
> actually do end up helping?

Just to be clear here, Adrian's claim that if someone else provided patches 
for 6.2, they would be committed, is incorrect.  The cost of committing the 
patch is almost zero -- the cost of QA'ing the patch, doing freebsd-update 
rebuilds, preparing security or errata notices, etc, is extremely real, and 
the reason that we carefully limit the number of releases we support at once. 
In fact, I'd argue that we have been supporting too many releases at once, as 
I think our latency for shipping errata notices and advisories is too high. By 
reducing the number of releases we support, we improve the speed and attention 
we can give each notice/advisory, which is an important consideration.

Robert N M Watson
Computer Laboratory
University of Cambridge