Hi Jon, Em Tue, 29 Mar 2022 09:31:43 -0600 Jonathan Corbet <corbet@xxxxxxx> escreveu: > Akira Yokosawa <akiyks@xxxxxxxxx> writes: > > > jinja2 release 3.1.0 (March 24, 2022) broke Sphinx<4.0. > > This looks like the result of deprecating Python 3.6. > > It has been tested against Sphinx 4.3.0 and later later. > > *Sigh*. I wish this stuff didn't feel like such a house of cards > sometimes... > > > Setting an upper limit of <3.1 to junja2 can unbreak Sphinx<4.0 > > including Sphinx 2.4.4. > > > > Signed-off-by: Akira Yokosawa <akiyks@xxxxxxxxx> > > Cc: Jonathan Corbet <corbet@xxxxxxx> > > Cc: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx> > > Cc: linux-doc@xxxxxxxxxxxxxxx > > --- > > Or we can bump the requirement to Sphinx>=4.0. > > Thoughts? > > It's probably time to consider a bump there, but that is a bigger one > than I would want to do at this point. Doing a bump at Documentation/sphinx/requirements.txt doesn't mean that people using have older versions would need to upgrade, as we don't need to bump the minimal requirement at Documentation/conf.py, nor the versions suggested by scripts/sphinx-pre-install. So, I don't see a problem on setting it to use sphinx==4.3.2 (or some other version). Yet, I would keep using a known-to-be-good version, instead of letting pip to just get the latest one. We need to verify both PDF and html generation, though, as I remember that some 4.x versions had/(have?) issues with the C domain and duplicate symbols detection. Also, it could be worth to check the build time with 2.4.4 and with whatever newer version we stick. > So I'll just fast-track this > patch in; dropping it into the stable updates probably makes sense too. Agreed. Thanks, Mauro
