[blt]

The author's expanded definition of "weak accept" is perfect.

I'm in a different area of CS. I've had reviews where the reviewer clearly skimmed over the math details. Still, the totally blank reviews in this article would not be accepted by program chairs in my area. They would find a new reviewer or ask the original reviewer to try again.

[bradleyjg]

Peer reviewing is considered duty to the academy but isn't tracked, publicized, or taken into account in e.g. lateral transfers, summer grants, or other things professors might want, right?

Is it harder to get a paper published in a given journal if you've blown off requests from them to peer review or done a crap job?

[blt]

Suppose postdoc X submits a terrible review to a conference. Conference committee member Y requested the review from X, and thus learns that X is a bad reviewer. Later on, Y's colleague Z is considering hiring X for a faculty position and asks Y about X. Y tells the story about the terrible review. X doesn't get the job.

[jmgrosen]

In my experience, reviews are typically anonymous.

[bo1024]

In my experience with CS conferences, reviewers are rarely anonymous to anyone involved except the authors.

[blt]

Yes for the paper authors, but what about the committee member who allocates reviewers to papers? Are they converted into "Reviewer 1", etc.? I wouldn't know, I have never been on a conference/journal committee.