On Jul 16, 2007, at 7:18 PM, J Aaron Farr wrote:
"Roy T. Fielding" <[EMAIL PROTECTED]> writes:
1. Somebody else has to vote on creation of a new lab.
I just don't see the point. I can understand having retroactive
votes wherein an inappropriate experiment is forcibly moved
elsewhere. I can also understand a documented list of things
that must be true for a lab to be started by a committer.
But I think we could easily replace the pre-proposal and vote
with a commit-then-review policy
Has this really stopped anyone from proposing a lab?
I'm fine with switching to commit-then-review, but I'm not sure it
will really make a difference in the number of type of proposed labs.
I don't know about numbers (how can we count the numbers of people
who thought about it, hit an entry barrier, and put it off to later?).
In any case, I don't think we need to encourage any more. It was more
a thought about group dynamics and how requiring a PI to write a full
proposal is a bit like asking each person to build their own campfire.
The result is neither efficient nor social.
Actually, this bullet really doesn't apply to webarch since it is
different enough to need a proposal anyway. I was thinking of the
various code projects in which it was obvious the vote would pass
and it was merely adding a few days delay and some uncertainty
regarding who calls the result. *shrug*
2. There is only one dev mailing list and one commits mailing list.
Having only one list has the effect of removing opt-in
conversation.
In other words, I self-censor my chatter on this list about my
lab
because I don't know if the others on the list are even remotely
interested...
I wouldn't mind chatter. It's way too quiet here as it is. If a
particular lab gets too noisy then we can see about new list.
It isn't bad now because there is no chatter. Start chatter about
every project in the lab and I will have to unsub just to keep my
mail load within my current irrationally optimistic reading time.
[Note the timeliness in which I responded to Yoav's original message.]
I know the practical reasons for both of those constraints,
since I am one of those infrastructure guys who occasionally
creates mailing lists and subversion repos. But the current
setup doesn't really work for me. I think it is important that
webarch be publicly archived in a way that is usable by others,
which means a topic-limited mailing list. Should I create such
lists on on my own ISP to talk about an Apache lab? Or should
I move my lab to my ISP so there isn't any concern about dual
use of Apache resources? Or should I widen the httpd umbrella
into something that would include webarch docs?
Any of the above? ;-)
I would prefer if we could find a way for Apache Labs to work for you.
I mean, isn't that the point? That is, unless you'd rather move the
work into httpd.
Yes, that's the point. I hate umbrella projects, even when I am
the chair of one. A solution would be to let a lab create their own
mailing lists if the PI (or one of the committers) is an infra
volunteer with apmail ability. I'm not sure if the board would buy
that one, but at least it would satisfy the original concern that
labs would be creating work for others.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
