Re: Problems with unicode characters under emacs and Xorg

[dm-list-email-notmuch@scs.stanford.edu]

David Edmondson <dme@dme.org> writes:

> I haven't seen this. Threads with a lot of complex HTML content (lots of
> nested tables, for example) can take a long time to render for me, but
> that is generally interruptable.
>
> Could you share one of these messages, or a sufficiently similar test
> case?

Thanks for the reply.  I can send one of these emails to you privately
if necessary, as it might contain semi-sensitive information.  However,
I think all you need is the subject line.  For example:

Subject: =?UTF-8?B?RGF2aWQ6IEhvdy1UbyBIaXJlIHRoZSBCZXN0IFJlbW9kZWxpbmcgQ29udHJhY3RvciDwn5Od?=

That subject line alone triggers the problem, because any search
returning that thread triggers the problem.  When decoded, the subject
line ends with unicode code point 0x1F4DD (MEMO).  Indeed, if I open up
a fresh emacs, and, independent of notmuch, type "C-x 8 RET memo RET",
it causes the emacs to hang for a minute or so.

Arguably this is a limitation of emacs or fontconfig, or I've installed
too many fonts on my system, or I've installed too few fonts (because
after all that computation it just renders a box with hext digits 01F4DD
in it instead of showing the MEMO icon).  However, the problem only
happens with notmuch, because notmuch is the only emacs functionality I
need that renders anything other than a very limited set of unicode
characters.  So if there's any way either to workaround the problem, or
to copy whatever other notmuch users are doing (is there some particular
unicode font I should just install on my system?), I would be very
happy.

Thanks,
David