Sebastiano Vigna <sebastiano.vi...@unimi.it> writes: > On June 30, 2025 2:19:16 PM GMT+02:00, Simon Albrecht > <simon.albre...@mail.de> wrote: >>On 30.06.25 07:00, Sebastiano VIGNA wrote: >>> LSR has been useful, I'm sure, for a lot of people, but by now you >>> can more easily ask to any AI information like that provided by the >>> LSR. >> >>Given all the stories of AI providing complete nonsense I cannot >>imagine how that’s a serious suggestion. > > I think that rather than repeating some fourth-hand information > randomly gathered on the internet,
Uh, isn't that the modus operandi of AI? > you'd better try firsthand stuff like Augment. The main problem I see overall is that a snippet repository is not providing solutions but examples. In the LaTeX universe, style files are used for encapsulating functionality. Openlilylib was partly intended to provide a framework for such work but it hasn't gained critical mass and it is not as integrated into the LilyPond engine as LaTeX is with the TeX engine. AI and LSR both don't provide guarantees and may be hit and miss, to various degrees. But abandoning LSR does not seem like a step forward without anything slated to succeed it. -- David Kastrup
