SUMMARY: Fast releases demand binary updates..
jrhett at svcolo.com
Thu Jan 12 00:11:48 PST 2006
FreeBSD-current: please accept this posting to close off the thread that
your list saw half of.
I'm going to kill this topic. Results of my trolling to see if we could
get any committer interest in this topic are:
17 enterprises e-mailing me privately to agree that it sucks, but that
they doubt it will ever change. Some notes about how we handle certain
situations today exchanged.
2 e-mails from people working at the bsdupdates project. I explain again
to them why we can't run GENERIC, and they agree that without core
operation system support there's no easy way to handle this.
3 e-mails from people who've been burned trying to raise this topic before,
suggesting that I run and hide after raising an issue like this. They
don't believe it will happen.
12 or so honest queries on the mailing list about why "make buildworld"
or the freebsd-update mechanism doesn't work for me. Which I try to answer
in detail, even when the questioner was insulting me.
7 or so suggestions on the mailing list that I deliver a working solution
before asking for consensus.
16-something people telling me that I clearly don't understand the problem,
and that if I wasn't an idiot I could solve it using this or that. Most of
which are tools we are using today, and know the limitations of pretty well.
(none of whom actually delivered information about how to improve on this)
20-something people telling me that my shorthand for -core proves that I'm
an idiot. It wasn't the topic, and it doesn't change the real question one
bit, but it's a great chance for everyone to call me an idiot.
1,215,545 or thereabouts people writing me to call me an idiot, without
much justification whatsoever.
In short, the situation remains as before. A lot of need, but no solution.
A lot of interest, but no belief that it will be accepted into the core
operating system. No interest from any of the developers who could make
Without any interest, you can't come up with a starting definition or goals
for consensus. Thus there's no point in trying to build a team to do it.
Kind of like invading Iraq without a clear objective and a plan.
(sorry for the off-topic political reference, but it is apt)
For now we'll have to see if Colin can find ways around the problems
without the tools he needs to do it right.
SVcolo : Silicon Valley Colocation
More information about the freebsd-current