[OSM-dev] 0.6 api - one more time

[Christopher Schmidt]

On Mon, May 05, 2008 at 03:33:39PM +0200, bvh wrote:
> From my POV, here is how the process can be improved concretly
> 
> - don't hold low-key meetings. Sure, invite only a few key
> driving people to improve efficiency but communicate about it
> openly. I'd go even as far as saying the more intimate you
> make the meeting, to more public you should be about it.

I agree with this. We screwed it up in the OpenLayers community the last
time we did it, and the forward motion of the tasks suffered as a
result.

> - take notes as you go and publish them as widely as possible

Honestly, it seems to me that this was done, within a reasonable limit. 

> - announce important stuff (for developers) on the dev mailing list
> as early as possible (certainly before linking them on the wiki)

Huh? That contradicts the above explicitly: One says publish as
early/often, the other says publish only after sending to the mailing
list? That doesn't make sense.

> - think twice before casually setting a date on changes that will
> break current setup for everyone

I think this one can be reworded/redirected as "Don't take any date
written in a wikipage as set in stone." 

> - use the wiki as a reservoir of information, not as a notification
> board

And it seems to me like that's what was done, except you expected
communication before there were complete notes -- and we've seen where
that gets us, since a single misplaced sentence caused a dozen or more
mailing list messages... 

Regards (and adding one more to the pile),
-- 
Christopher Schmidt
MetaCarta