Post B6u5A2WiN5xBvKMrZ2 by sinbad@mastodon.gamedev.place
 (DIR) More posts by sinbad@mastodon.gamedev.place
 (DIR) Post #B6u5A1fBa78TFJK784 by sinbad@mastodon.gamedev.place
       0 likes, 0 repeats
       
       It’s looking increasingly likely that if you need stable systems for the next couple of years, forking important dependencies just before December 2025 and only cherry picking fixes since (and reviewing them) is the only sane choice, unless those projects have sworn off LLMs. Far too many developers have totally lost the plot in favouring velocity over everything else. I’d like to think that engineering principles will eventually re-emerge but for the moment the cocaine is winning
       
 (DIR) Post #B6u5A2WiN5xBvKMrZ2 by sinbad@mastodon.gamedev.place
       0 likes, 0 repeats
       
       I’m sure part of it is cases where open source maintainers are completely over-worked and under-compensated (been there), are burning out and when something comes along that seemingly provides more output for less effort, maybe it looks like a better outcome than simply stopping. Personally I disagree - I think stopping is a better outcome to moving backwards by slowly eroding trust in what you have. Those by the sidelines who might maintain it properly are then empowered to do so.