Re: lei-managed pseudo mailing lists

[Eric Wong <e@80x24.org>]

Konstantin Ryabitsev <konstantin@linuxfoundation.org> wrote:
> On Mon, Apr 26, 2021 at 05:37:26PM +0000, Eric Wong wrote:
> > > The latter is specifically something I think would be of interest to kernel
> > > folks, so I envision that we'd have something like the following:
> > > 
> > > - a maintainer publishes a configuration file we can pass to lei
> > 
> > The command-line might be enough, the pathname of the current
> > state/config file is a bit tricky and tied to its output.
> > I suppose "lei import-search" can be a command, though...
> 
> Excellent, excellent. How well does it deal with the situation when the search
> parameters change?

"lei edit-search" can be used to zero the maxuid parameters;
and normal v2 deduplication will prevent duplicates from showing up.
It's not automatic, though; though that probably seems like a good
idea to keep manual, anyways, given the step 2. below.

> > > - our backend lei process uses all of lore.kernel.org sources to create and
> > >   continuously update a new public-inbox repository with matching search
> > >   results
> > 
> > There's already some accomodations for that in LeiSavedSearch
> > which can present itself as a PublicInbox::Inbox-ish object to
> > PublicInbox::WWW (untested).
> > 
> > Searching an within LSS isn't implemented, yet, but I think it's
> > doable w/o extra Xapian storage.
> > 
> > However, git object storage isn't duplicated, which is nice for
> > local use (instaweb-like), but supporting clone/fetch isn't as
> > natural...
> 
> I'm thinking we need the ability to make it a real clonable repository --
> perhaps without its own xapian index? Actual git repositories aren't large,
> especially if they are only used for direct git operations. Disk space is
> cheap, it's the IO that's expensive. :)

True, though cache overheads hurt a bit.  I also wonder if lei
can increase traffic to public-inbox-<imapd|nntpd> to reduce
the need/use of "git clone".

> If these are real clonable repositories, then it would be easy for people to
> set up replication for just the curated content people want.

Understood.  Using --output v2publicinbox:... w/o --shared is
totally doable.

> > Perhaps supporting a v2 inbox as an lei q output destination
> > is in order:
> > 
> > 	lei q --output v2publicinbox:/path/to/v2 --shared SEARCH_TERMS
> > 
> > --shared would be "git clone --shared", the new v2 inbox can
> > use ~/.cache/lei/all_locals_ever.git/ as an alternate and not
> > duplicate space for blobs.
> 
> Not really worried about deduping blobs, but I'm wondering how to make it work
> well when search parameters change (see above). E.g.:
> 
> 1. we create the repo with one set of parameters
> 2. maintainer then broadens it up to include something else
> 3. maintainer then decides that it's now *way* too much and narrows it down again
> 
> We don't really want step 2 to lead to a permanent ballooning of the
> repository, so perhaps all query changes should force-append a dt: with the
> open-ended datetime of the change? Or do you already have a way to deal with
> this situation?

The aforementioned maxuid prevents stuff that's too old from
being seen.  Otherwise, there's always "public-inbox-learn rm".

> > > - we set up a mlmmj list that doesn't receive any direct mail but is only fed
> > >   from saved search results; people can subscribe/unsubscribe as they would
> > >   with any other mlmmj list
> > > 
> > > Any particular reason this wouldn't work?
> > 
> > Nope :)  As long as all the data formats can interoperate
> > (mostly RFC5322/2822).  "lei convert" is nice, too :)
> 
> Great! I believe this will help untangle the current situation with "where
> should I send this kernel patch". 
> 
> I want "just send it to linux-kernel@vger.kernel.org" to be a valid option
> again. Participating subsystems can then define what patches they want to see
> by setting up pseudo-lists and letting participating reviewers/maintainers
> subscribe to them via their preferred mail delivery mechanism.

Yup, that seems easiest for new contributors.