Ejecting things more aggressively out of 'seen'?
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Nov 2, 2026, 07:31 UTC
- Message-ID
- <xmqq1p93ofjx.fsf@gitster.g>
- In-Reply-To
- <47bd0302-fc52-4df0-98a0-6fad7eb0fb05@gmail.com>
Phillip Wood <phillip.wood123@gmail.com> writes:
Show 15 quoted lines
> On 29/07/2026 18:48, Junio C Hamano wrote: >> >> Perhaps the sensible thing for me to do is to stop taking any new >> topics into 'seen', even if I've spotted them, until I see somebody >> give them a real review. >> >> Otherwise, it becomes too tempting for me to jump in, give them a >> superficial read after seeing them linger in the "What's Cooking" >> draft in the "Needs review" state for too long, and, believing I've >> seen enough, mark them for 'next'. If I don't queue a patch that >> nobody seems to have read carefully, I won't succumb to such >> temptation. > > Maybe, though I do find having the patches in seen makes it easier to do > an in-depth review as it means I don't have to apply them myself.
I have operated in "aggressively accept many things into 'seen'" mode for some time since this exchange. We ended up with many "Needs review" topics, which nobody was interested in, sitting in 'seen' for weeks. Having these and other topics in 'seen' may have helped other topics get reviewed, which might have been missed by you and others if they were not in my tree in a readily available form. Thus, it is not entirely fair to declare this a failed experiment.
Still, we need to decide, and better yet, establish a policy on, how to deal with topics that remain unreviewed even after lingering in 'seen' for quite some time.
A mechanically applicable set of policies makes it easier to manage a large volume of topics. Some of these are currently written down, while others may have been guessed by those who read a few "What's cooking" reports in a row:
- A topic that has spent a week in 'next' graduates to 'master', unless there is a strong reason to do otherwise.
- A topic marked as "Will merge to 'next'" is merged 60 hours after such a label is applied (in other words, people can stop me if they act soon enough).
- A topic goes into the "Waiting for response" state once reviewed, where it remains until the author responds. After staying in this state for a month or so, it is demoted to the "Stalled" state, and eventually dropped.
We should probably have a rule that says "a topic that did not get reviewed for N weeks is ejected from 'seen'".
Tonight, I am sending a notice for each topic that has been in that state for more than a month (the oldest one dating back to mid-July).