On 11/29/23 15:05, Niall Douglas via Boost wrote:
On 29/11/2023 05:37, Jeff Garland via Boost wrote:
But you have to ask - how many have done the deep dive we typically see in boost reviews? Have you even read the paper before giving an LEWG review comment? I fear the bigger issue here is that tiktok attention span defines most people these days. Having a proving ground of real user experience allays those concerns because it's people that are invested in the details of how it works.
Last few meetings at LEWG I find myself repeatedly thinking "this person hasn't read the paper". And not just for my own papers, for a majority of papers. R0's and R1's get read. R14's do not in my experience.
Something I noticed about Titus when he was chair was he always seemed to have read the revision of the paper being discussed that day in detail, and had a good on-the-day knowledge of where things were and how forward progress could be maximised.
I know some felt as a result papers got pushed through too quickly by Titus, and a more reflective slower process would have produced higher quality results with fewer missing parts and footguns. However if a majority of the room does not read the latest revision of a paper, it's hard to produce higher quality results no matter how slow a process you use. There is an argument therefore to keep revisions well below ten, and either push stuff through faster or reject entirely much earlier.
I'm not a committee member, just an observer from the outside, but it looks to me that people that aren't familiar with the proposal should abstain from the discussion. And if those people constitute the majority of the room then either the proposal discussion should be rescheduled to a later meeting or the room should be dismissed to do their homework. I realize that most of these people are likely busy and there are lots of proposals going, but discussing something that not everyone is familiar with is simply not productive. And if that means that the committee is a bottleneck, then split the committee into more parallel groups that focus on separate more localized domains.