Closed kestlund closed 8 years ago
:+1:
:+1:
5 normal days, or 5 business days?
5 normal days. What is a business day, anyway. I always tell people, PCDM is the fun in work/life balance.
Then let's give it 7 days in case folks are away and want to be able to weigh in on something.
I think that's the distinguishing question. Do we wait to give all a chance; is that what is required? Or is there a number/time where we have enough that is good enough and their expertise is considered sufficient?
I don't suppose PCDM has emergencies though, so I'm fine with 7 days.
I'm fine with 7 days — I agree we don't have emergencies, and I think it's nice to give people ample time to weigh in.
Then we create an existential emergency ontological extension :smile:
Jokes aside, I'm just always aware of the numbers of Hydra folks vs Islandora folks. There could be a case where Hydra folks could push something through and Islandora folks never have a chance to comment.
Good point @ruebot !
Suggestion Rev. 1 Votes on code should be held by posting comments to the relevant Github pull request or issue. To pass, a vote needs a :+1:/+1 from three or more committers (excluding the author of the pull request), with no :-1:/-1's within 7 days of original pull request or issue.
:+1: to 7 days.
I can't see any reason why we'd want to rush ontology publishing.
Though this discussion has me thinking: do we have any procedures around releases or versioning?
We have been following the same practice as Fedora: the namespace URI redirect to a versioned page (e.g., http://pcdm.org/models# redirects to http://pcdm.org/2015/09/28/models#).
@escowles okay. that seems fine. If there's any cause to discuss it more, we can open a separate ticket.
:+1: to the wording: https://github.com/duraspace/pcdm/issues/44#issuecomment-186667381
:+1: for 7 days.
Revisiting this, often times there is considerable back-and-forth on issues and pull-requests. As a result, it does not seem reasonable to require a veto vote to come in during the first 7 days of the initial PR... because the initial PR is likely different from the evolved, post-debate, PR. I would suggest an updated policy that requires vetos to come in after the point where there is consensus on a "final PR". In my opinion, a "final PR" is one that has 3 or more +1 votes.
@awoods :+1:
@awoods I agree with the substance of what you're saying, but I'm not sure how that translates into a voting process. Is the new guideline something like:
To pass, a vote needs a :+1:/+1 from three or more committers (excluding the author of the pull request), with no :-1:/-1's. Discussion should be left open for at least 7 days, and at least 1 day after any comment that isn't a :+1:/+1, to allow ample opportunity to raise concerns.
@escowles, The text you are offering seems reasonable. I could also see a process like this:
How about:
In order to ensure that there is consensus when making a change to content maintained by the PCDM community, a successful poll of the committers is required. Votes on requested changes are submitted by posting comments to the relevant Github pull request or issue. To be successful, a poll needs :+1:/+1 votes from three or more committers (excluding the author of the pull request), with no :-1:/-1 votes within 7 days of the most recent commit or substantial edit to the requested change.
+1 @azaroth42
(I fixed 2 minor typos in comment above)
:+1: @azaroth42 's changes in https://github.com/duraspace/pcdm/issues/44#issuecomment-187264684
If we get one more +1 and no -1's by tomorrow (day 8; 7 days having passed) and ignore the self-referential conditions of such a vote, then I think this passes!
:+1:
Moving voting criteria discussion out of pull request discussion
Wiki: voting
Suggestion Votes on code should be held by posting comments to the relevant Github pull request or issue. To pass, a vote needs a :+1:/+1 from three or more committers (excluding the author of the pull request), with no :-1:/-1's within 5 days of original pull request or issue.