A style of reviewing written work used in the community that flourished around adapting Christopher Alexander's idea of "patterns" to software.
Richard P. Gabriel 
(who imported the style from the poetry community) provided a deeper introduction .
--- I think these workshops work well and are useful. My summary of the process goes like this:
1. You have an author of a text and a bunch of people who've read it. They are typically arranged in a circle.
2. The author may do a brief presentation of the work (conventionally, reading some snippet that represents it well, to suggest what matters to her or to set the mood). I don't find this part very valuable, but ¯\\\_(ツ)\_/¯.
3. The author then becomes a "fly on the wall" and backs out of the circle. (Physically moves into the background.) The reviewers are talking to each other; the author just listens. (This is astoundingly hard for some people. They follow the letter of the rules, but violate the spirit with ostentatious facial expressions that convey their disagreement.)
4. The first segment is devoted to people saying what they liked about the text. This feels awkward for people in professions that emphasize critique rather than support (like software). For that reason, when introducing the style, I've characterized this as the reviewers telling the author "When you revise the work based on our feedback, be sure not to discard this good thing."
5. The next section is devoted to what's often called "suggestions for improvement." The attitude is *not* critiquing what the author is trying to accomplish, but one of highlighting how they have fallen short of their goal.
When a moderator is needed, this is when the moderator needs to step in and keep reviewers from drifting into "that thing you want? you shouldn't want that." Also something software people are prone to.
I have a bit of a bias against *requiring* that every "this doesn't work" be paired with "and here's how you might make it work," but I think on balance it's good to require a suggestion for improvement. If one critique gets lost because the reviewer couldn't think of a suggestion for improvement, well wotthelll archy .
6. The author rejoins the circle.
7. The author should *want* to thank the reviewers. In any case, it's convention that she do so.
8. The author may comment on what she's heard. Often, that's a defensive "what I meant was...", which is not useful. (What you *meant* isn't at issue; how it was *received* is.) More useful is asking for clarification about some suggestions for improvement.