On Wed, 9 Aug 2006, Matt Benson wrote:
--- Henri Yandell <[EMAIL PROTECTED]> wrote:
On Tue, 8 Aug 2006, Matt Benson wrote:
Henri, out of sheer curiosity, where is it
documented
that a commons committer doesn't have a binding
vote?
The only thing I could find in the charter [1] was
a
link to the Jakarta guidelines [2], which in turn
links to a "Decision Making" page [3], which
states
that "the only binding votes are those cast by a
Committer" (the next sentence is pretty
interesting as
well, though unrelated).
http://www.apache.org/foundation/how-it-works.html#roles
Our Decision Making page is dated in that respect,
probably worth deleting
and making sure that the foundation pages have
something with the
core important parts of it in them.
Since your link was Apache-specific, and my link was
Jakarta-specific, my assumption would be that Jakarta
trumps the ASF in its grant of a vote to committers
rather than PMC only. Ant does this as well; actually
some changes are a vote of committers and some are PMC
only, the specifics being outlined in the project
bylaws. The alternative interpretation is that no
project has the right to grant a binding vote beyond
the PMC, thereby overruling the ASF guidelines you
linked. But this interpretation would indicate a
large upheaval: a quick look shows that the ASF
"flagship" (HTTP server) grants committers a vote.
Gump takes an Ant-like approach (not surprising since
Gump's originators had worked on Ant). The XML TLP
also says committers have a vote. IMHO it still looks
as though Jakarta committers do have a binding vote
unless the privilege granted at
http://jakarta.apache.org/site/decisions.html is
explicitly rescinded.
In theory I think you're right. Subproject process overrides foundation
process as long as it doesn't clash with foundation process. ie) We can't
have a sub-pmc for Jakarta ECS etc.
Many of the pages on voting out there predate the informal nudging from
the board that only PMC votes are binding on PMC issues, and probably also
predate the foundation page I mentioned. The Jakarta page definitely does
and clashes with the foundation process so needs a change.
The general interpretation seems to be that there are issues which
committers can vote on, and issues that have to be the PMC. Code issues
are often placed in the former, release and process in the latter. It's
only issues of a legal concern (I think) that have to be in the latter.
Very nice tables on the Ant bylaws btw :)
Hen
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
