Guillaume Emont

> Would you mind sending a few problematic anonymized mails our way? Ideally in the form of a PR that adds (failing) unit tests. I'll see if I find time...

Ok, I did some more trying with things, and it turns out that my minimal test is already in the tree: it is https://github.com/pazz/alot/blob/master/tests/static/mail/utf8.eml Here are a few screencaps of...

A couple additional notes on tests: - If I remove the mock decorator from `test_simple_utf8_file()`, the test fails, which matches my experience in alot. I do not have a `~/.mailcap`...

FWIW, I have the following `text/plain` lines in my `/etc/mailcap` (unmodified by me, fresh ubuntu docker image): ``` text/plain; more %s; needsterminal text/plain; vim %s; needsterminal text/plain; view %s; edit=vim...

Just updated to latest master with #1526 merged. It does seem to fix the main issue for me (non-ascii utf-8 characters are displayed correctly), though `togglesource` still shows me inaccurate...

> But this is to be expected: togglesource will result in alot displaying the email's source text verbatim, including not yet decoded quoted-printables. In this case, the problem is that...

I was hitting a segfault all the time with the same backtrace, using 424669b4e08dd841f499981bbce10c15a5c48463. I have now updated to 486df70dbaf31eaefd441f1f2f47c93d8a057fa6, using python3-notmuch2 from ubuntu 20.10. I get much less segfault,...

After running my optimization script on 2.38 I found the same values as for upstream, so I think this patch can be relevant on this branch as well.

> Closing this, this patch landed in #3991 I think you mean landed in ae23b36653931c33218d916ccc3b98181105f288

The main issue why this is still a draft, is that the templated test generator has some issues. In particular, we want to use it with templates such as: ```javascript...