absc
honked back 13 Apr 2026 16:15 +0200
in reply to: https://tech.lgbt/users/Natasha_Jay/statuses/116395868381628955
Me: I'm here, today, trying to "work", coughing the shit out. Happy Shitty Monday!
absc
honked back 13 Apr 2026 16:15 +0200
in reply to: https://tech.lgbt/users/Natasha_Jay/statuses/116395868381628955
Me: I'm here, today, trying to "work", coughing the shit out. Happy Shitty Monday!
absc
honked back 13 Apr 2026 16:12 +0200
in reply to: https://x.keinpfusch.net/users/uriel/statuses/01KP3G656XTS1E1GPXEN8Y1DVE
I haven't tried such a project myself, so I can't judge, of course. If i get it right: you must push hard on the agent to generate code kind of like you want it but this pushes the cost way up? How do you feel about reviewing the generated code? Is it doable after convincing the agent to create EAL5+ compliant code? I have the feeling that in the end, given some level of expected quality, everything will balance out and the speed-up will disappear under other things. This doesn't remove the fact that, probably, what really matters is that the kind of effort changed on your side, making the experience better anyway.
absc
bonked 13 Apr 2026 11:17 +0200
original: AaronDavid@mastodon.world

absc
bonked 13 Apr 2026 10:28 +0200
original: uriel@x.keinpfusch.net
absc
bonked 13 Apr 2026 10:18 +0200
original: pitrh@mastodon.social
pfsync(4) Packet Header Field Renamed to Avoid AI Bug Report Noise https://www.undeadly.org/cgi?action=article;sid=20260413055845 #openbsd #pfsync #networking #redundancy #carp #pf #packetfilter #development #libresoftware #freesoftware
absc
bonked 13 Apr 2026 10:10 +0200
original: stefano@mastodon.bsd.cafe
My week starts with a request: "I need a server to deploy to production, but the devs have no idea how to do it. They don't know how to use the terminal, they don’t know how to handle certificates, nothing. They need to be able to click a few buttons and deploy directly to production. They're Vibe Coding experts." Welcome to 2026.