[meta] additional input during public reviews


(Chéyo Jiménez) #1

Hi there,

I have noticed that on the most controversial reviews there has been a trend of people arguing their specific stance on the proposal beyond the formal response. I am referring to the instances when somebody formally reviews a proposal but keeps responding to other peoples proposals in order to reiterate their position.
I'm all for debate and bike shedding but I don't think a review thread is the right place:
1) it makes it difficult for review managers to gather all the input
2) it distracts from other reviewer's input
3) turns reviews into heated discussions that discourage actual formal reviews.

I recently saw this in the sealed by default review but I have seen it before.

Perhaps the review manager can state that every person should only respond once or twice to the public formal review thread?

Another thing I have noticed is people giving .5 , +0 reviews. I know we are not voting here per se but I fail to see how these reviews help the review manager or the community make a decision for or against the proposal; perhaps this could be addressed in a different thread. I just feel that when somebody gives a 0, .5 review, it encourages people to respond to the reviewer in trying to convince them into a side. My recommendation is that people who are not sure should wait for other people to give concrete input to inform a concrete response.

Thanks!


(Chris Lattner) #2

Hi there,

I have noticed that on the most controversial reviews there has been a trend of people arguing their specific stance on the proposal beyond the formal response. I am referring to the instances when somebody formally reviews a proposal but keeps responding to other peoples proposals in order to reiterate their position.

The important thing here is that it doesn’t make sense to keep arguing points that were covered before, but *new* points are always welcome. For example, the core team plans to re-discuss SE-0111 in light of the ongoing discussion that was sparked after the proposal was accepted.

That said, I completely agree that it is clearly the case that not everyone will like the resolution of every proposal, and also agree that it is pointless to argue already well explained positions after a proposal is accepted.

-Chris

···

On Jul 11, 2016, at 10:47 AM, Jose Cheyo Jimenez via swift-evolution <swift-evolution@swift.org> wrote:

I'm all for debate and bike shedding but I don't think a review thread is the right place:
1) it makes it difficult for review managers to gather all the input
2) it distracts from other reviewer's input
3) turns reviews into heated discussions that discourage actual formal reviews.

I recently saw this in the sealed by default review but I have seen it before.

Perhaps the review manager can state that every person should only respond once or twice to the public formal review thread?

Another thing I have noticed is people giving .5 , +0 reviews. I know we are not voting here per se but I fail to see how these reviews help the review manager or the community make a decision for or against the proposal; perhaps this could be addressed in a different thread. I just feel that when somebody gives a 0, .5 review, it encourages people to respond to the reviewer in trying to convince them into a side. My recommendation is that people who are not sure should wait for other people to give concrete input to inform a concrete response.

Thanks!

_______________________________________________
swift-evolution mailing list
swift-evolution@swift.org
https://lists.swift.org/mailman/listinfo/swift-evolution


(Xiaodi Wu) #3

It'd certainly be nice to have more formal reviews, but I would disagree
with the rest of your proposal. The Swift Evolution process guidelines
explicitly tell authors to be responsive to comments during the review
period. This naturally lends itself to back-and-forth conversation, and in
the past the conversation has (sometimes) resulted in improvements to
proposals. I would hate to see that reduced to formal postings only.

Naturally, it is important that all participants maintain a professional
approach to debate, but let's not pretend controversy doesn't exist in the
community when it does. Among other things, it only lends itself to new
threads with requests for re-consideration because argument X or Y wasn't
brought up during review.

···

On Mon, Jul 11, 2016 at 12:47 Jose Cheyo Jimenez via swift-evolution < swift-evolution@swift.org> wrote:

Hi there,

I have noticed that on the most controversial reviews there has been a
trend of people arguing their specific stance on the proposal beyond the
formal response. I am referring to the instances when somebody formally
reviews a proposal but keeps responding to other peoples proposals in order
to reiterate their position.
I'm all for debate and bike shedding but I don't think a review thread is
the right place:
1) it makes it difficult for review managers to gather all the input
2) it distracts from other reviewer's input
3) turns reviews into heated discussions that discourage actual formal
reviews.

I recently saw this in the sealed by default review but I have seen it
before.

Perhaps the review manager can state that every person should only respond
once or twice to the public formal review thread?

Another thing I have noticed is people giving .5 , +0 reviews. I know we
are not voting here per se but I fail to see how these reviews help the
review manager or the community make a decision for or against the
proposal; perhaps this could be addressed in a different thread. I just
feel that when somebody gives a 0, .5 review, it encourages people to
respond to the reviewer in trying to convince them into a side. My
recommendation is that people who are not sure should wait for other people
to give concrete input to inform a concrete response.

Thanks!

_______________________________________________
swift-evolution mailing list
swift-evolution@swift.org
https://lists.swift.org/mailman/listinfo/swift-evolution