Post B2BssSCo7C1hzW9XTU by khm@hj.9fs.net
 (DIR) More posts by khm@hj.9fs.net
 (DIR) Post #B22iW9p32EXgdq86me by lanodan@queer.hacktivis.me
       2026-01-07T13:25:26.483099Z
       
       0 likes, 0 repeats
       
       @dalias Reminds me that typically I want at least a 2nd implementation for data formats, protocols, …And even more when it comes from ones like Google or Facebook as they are known for horrible technical choices (like anti-packaging buildsystems and languages which don't bootstrap) but also for dropping projects with little to no announcements.
       
 (DIR) Post #B23L88KEqTcjGGFp5s by lanodan@queer.hacktivis.me
       2026-01-07T20:38:04.999633Z
       
       0 likes, 0 repeats
       
       @dalias Yeah, combination of both would be ideal, specially if the standard can end up with a decent amount of clout so doing things differently would look questionable to a lot of people.Also reminds me that zopfli allows to get higher-compression while still using deflate/gzip, which while much slower can be worth it for long-term files.
       
 (DIR) Post #B23SKZBPVxwJiG06ds by lanodan@queer.hacktivis.me
       2026-01-07T21:58:45.126992Z
       
       0 likes, 0 repeats
       
       @ska @dalias I meant to exclude de-facto ones, otherwise it would mean having to effectively clone systemd.Like for example it's not in https://www.freedesktop.org/wiki/Specifications/(Not that I'm a fan of most of those, yikes)
       
 (DIR) Post #B28ALaFODUW1uzzrvs by lanodan@queer.hacktivis.me
       2026-01-10T04:30:49.473299Z
       
       0 likes, 0 repeats
       
       @dalias @alwayscurious @lunareclipse Also you don't need to change curl to be offline-first, you change what's calling curl.Like distros use curl/wget all the time to fetch source tarballs and packages, and caching of artifacts is something that was solved decades ago.By the way for npm: Yarn 1.x has an offline mode where you can drop pre-fetched tarballs into a directory.
       
 (DIR) Post #B2BssPFB7v8mo0EtFo by luatic@mastodon.social
       2026-01-07T22:27:44Z
       
       0 likes, 0 repeats
       
       @dalias No, we should not.Open formats with high quality open implementations should be judged by their merit. This includes concerns over future maintenance and direction of projects, as well as the cost of switching standards - regardless of whether something is by big tech or not.Open source should not be stuck on suboptimal standards just because, quite literally, they were "not invented here [in our community]".
       
 (DIR) Post #B2BssQQYj5q2TaZUdE by khm@hj.9fs.net
       2026-01-08T00:16:26Z
       
       0 likes, 0 repeats
       
       oh see we can specify "open" but other qualities have to be "judged by their merit" eh1. this is arbitrary line-drawing, unless you're asserting openness is required for merit2. if one societal characteristic is required for merit, why are you asserting that contribution to or inhibition of human quality of life is a lesser or irrelevant factor in merit3. without a comprehensive definition of 'merit' your declaration boils down to "should be judged by stuff I care about" which isn't much use to anyoneCC: @dalias@hachyderm.io
       
 (DIR) Post #B2BssROp722nUUlcZ6 by luatic@mastodon.social
       2026-01-08T13:13:03Z
       
       0 likes, 0 repeats
       
       @khm you're missing my point and choosing to talk about points of your own volition instead.you can have that discussion, but i'm not interested in partaking in it.
       
 (DIR) Post #B2BssSCo7C1hzW9XTU by khm@hj.9fs.net
       2026-01-08T16:28:02Z
       
       0 likes, 0 repeats
       
       your point isn't valid
       
 (DIR) Post #B2BssTQfZ8i1mne7ii by luatic@mastodon.social
       2026-01-09T00:34:45Z
       
       0 likes, 0 repeats
       
       @khm I might have been interested in your opinion of my point had I gotten the impression that you made an attempt at understanding it before deciding to write a critique.
       
 (DIR) Post #B2BssUFMWfG6K1Mbjc by khm@hj.9fs.net
       2026-01-09T01:11:33Z
       
       0 likes, 0 repeats
       
       I do not believe you
       
 (DIR) Post #B2BssUooOrcC5xwuzQ by luatic@mastodon.social
       2026-01-11T22:04:39Z
       
       0 likes, 0 repeats
       
       @khm I think you forgot to read the part where I already told you that I don't care.
       
 (DIR) Post #B2BssVLmQHzDkDNFNQ by khm@hj.9fs.net
       2026-01-11T22:06:33Z
       
       1 likes, 0 repeats
       
       don't believe this either