techhub.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A hub primarily for passionate technologists, but everyone is welcome

Administered by:

Server stats:

4.6K
active users

#utf8

3 posts3 participants1 post today

It is 2025 and it has been 0 days since I wasted way too much time due to f"¿Quéucked up character encoding.

Continued thread

Turns out #sqlite3 does not have a built-in function to validate #utf8 - it just does GIGO (garbage in, garbage out): sqlite.org/invalidutf.html

So one quick hack of a utf8 validation loadable extension, I ran `select * from oc_filecache where isutf8(name) is null` and there was one (!!!) among the 3.5M entries (it was an external file).

`delete from oc_filecache where fileid=287791` and I'm ready to go again. Setting `cpupower frequency-set -g performance` and now it's just 20min left.

sqlite.orgInvalid UTF Policy

AH ! Dovecot a enfin un support de SMTPUTF-8 (les adresses mails internationalisées) si on le compile avec la bonne option (--enable-experimental-mail-utf8).

C'est le cas dans Debian (c'est sans doute le défaut de Dovecot).

L'option ("mail_utf8_extensions") n'est ensuite pas active par défaut.

Activons là et a nous les adresses mails en katakana inutilisable \o/

#Unicode is one of those little things in life that I can't help but smile about.

Is it perfect? No, of course not. Is it better than the alternative? Yes, so much so that every time I'm confronted with a long list of character encodings I can choose from, I feel a sense of relief when I find #UTF8 among them.

I wouldn't have thought it possible to standardize a single character encoding for everyone, and yet, somehow, there is just such a standard.