10

When reviewing posts on Stackoverflow, I miss the ability to communicate with other reviewers. I think that I, as a reviewer, do not improve myself, because I don't learn from other reviewers. I'm allowed to review just because of an amount of reputation, which I got from asking and answering. I got feedback on questions and answers and improve myself. But when I start reviewing, I don't get any feedback anymore.

I should have some good example here, where the reviews are hard to judge and I think that communication would improve the reviewing quality a lot. Unfortunately I don't have them at hand. I only remember that question tend to be closed as "not constructive", just because nobody knows the answer. When someone starts closing, it gets closed for sure.

  • The review process should be something like a thread, where people vote and may comment. (Mandatory, hand-written comments would probably already improve the review quality, because it scares batch-hunters away. It is not necessarily a good idea in every case, because it slows down the review process.)
  • Reviewers should see what others do.
  • Reviewers should get messages in the inbox. E.g. the review did not end according to my vote (which is nothing bad at all, it is a democracy, but may be still important for me to know). So I get feedback if I for instance permanently use the wrong reasons or do other mistakes.
  • There should also be a way for the community to flag/remove votes e.g. because of obviously silly arguments.

There is a discussion (project Honeypot) which is about the same topic.

9
  • 1
  • Chat. Create new room titled "Reviewers Lounge" or something like that and wait for people to join. With time people will notice and come and you'll be able to communicate with them. Mar 20 '13 at 10:03
  • The problem is, while you want to communicate about items in the review queue others don't. There are too many people who are auto-reviewers and won't want to take the time to discuss the correct approach to items in the queues. I think by the time you get other's to comment/discuss the item in the review queue will have been approved, rejected, etc.
    – Taryn StaffMod
    Mar 20 '13 at 10:08
  • @Sha: It would be hard to talk about all the (hundreds) of current reviews in a single chat. And I want to talk to all the people which review the same as I do. Mar 20 '13 at 10:08
  • You mean inline chat in the review page itself?? Mar 20 '13 at 10:09
  • 1
    @Sha: not chat. I think it would be enough to comment your vote. You should have arguments to your vote. Mar 20 '13 at 10:21
  • Maybe it would be enough to see for which reasons others rejected a suggestion. This is especially for the "Other" reason, where at the moment you see that someone has rejected using the other reason, but you don't see the text... Mar 20 '13 at 10:26
  • 1
    @bluefeet: I don't think that we should support "auto-reviewing". That's exactly what I want to avoid by improving review quality. I don't understand you last sentence ("I think by the time...). Mar 20 '13 at 10:26
  • 2
    @Spontifixus: And you don't see for what reason someone approved an edit or voted to leave open a post. Actually, you don't see them at all. Until the review is finished. But nobody ever comes back to the review after voting... Mar 20 '13 at 10:30
3

I can see the intended benefits, but I don't think it would work this way in practice. First, you're making reviews too heavyweight. A majority of reviews don't require a specific comment. Second, for the ones that do require a specific comment, it often belongs on the post, notified to the poster — for example to explain why you're closing or reopening. Third, reviewers' identity are intentionally not exposed in many queues, and some of their actions (voting, flagging) are deliberately kept secret. To expose reviewers more risks retaliation against negative reviews.

Lastly, and most damning, communication between reviewers is intrinsically a bad thing. The point of having multiple reviewers is to have multiple independent points of view on a post. If you could see too easily what previous reviewers thought about a post, that would create a massive “follow the leader” effect where only the first reviewer's opinion matters.

Badge hunters would have it easier under your proposed system — they'd just skip whenever they were the first to review and agree with the previous reviewer otherwise.

3
  • I understand some of these points. But: I didn't say that there should be a mandatory comment. Just that there should be the possibility to add one. The independent points of view are not given today, because I can easily "follow the leader" in the title e.g. "Should this question be closed as off topic"? Sure I want. Or I click on "close" and see what others do. I don't even have to understand why. Most important to me is to be able to add a comment when I want to, to see when others vote to keep it and to be informed about the outcome when others disagree with me. Mar 20 '13 at 20:30
  • Quote: "To expose reviewers more risks retaliation against negative reviews." So why are the reviewers specified in "Closed as ... by ..."? Mar 20 '13 at 20:47
  • @StefanSteinegger Votes to close/reopen are exposed. For the votes not to close/reopen, you have to go digging. Up/down votes are never revealed. Yes, sure, occasionally I've wanted to write my opinion of the post in a comment, but in such cases it was almost always appropriate to leave a comment on the post (e.g. “I'm voting to close because …”). Mar 20 '13 at 21:15

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .