Post #1733388
2026-04-27 19:23 UTC
Replies (3)
-
@dalias@hachyderm.io 2026-04-27 19:27
My idea of a successful advancement in software/computing is something that lets me forget about an awful layer of hacks we had to do in the past BY NO LONGER HAVING IT BE A THING* and free up the associated cognitive space. (*) NOT by having it outsourced to someone else who will do it wrong, or another abstraction layer. That does not let me forget about it. It just makes it worse.
-
@ska@social.treehouse.systems 2026-04-27 20:16
@dalias@hachyderm.io I had taken a look at DNS-01 and discarded it because it was easier for me to grant the ACME client write access to a place in my served HTTP data (just a directory in the filesystem made writable to the acme user) than to my served DNS data (need to edit or assemble text files then compile a constant database, need root privileges to install the new cdb). I took a look at DNS-PERSIST-01 and yes, that looks much better on the ACME client side, but it looks like it does require more persistent state on the ACME server side. It's a tradeoff I'd be happy to make as a client, but does it really reduce the overall complexity? (Edit: typo)
-
@lispi314@udongein.xyz 2026-04-27 23:25
@dalias@hachyderm.io It does look fairly convenient.