14:41 <freemint> beneroth: https://www.youtube.com/watch?v=o_AIw9bGogo The tragedy of systemd worth a view - really excellent commentary?

14:42 <beneroth> Thanks freemint, I'll check it out!

14:42 <beneroth> freemint, btw. I would be interested in a pil wrapper for telegram.

14:43 <freemint> beneroth: i proposed several levels of implementatino, i would like your feedback on that. Do you need a refresher?

14:46 <freemint> my ideas are for the telegram bot are 1. dumb interface (message "bla") '*Group1) 2. an http_server like setup (text is parsed in to commands triggering functions) 3. 2 but with addtional times/conditions/IFTT(if this then that) to make use out the "push" side of messaging and domain specific language to manage all that.

14:46 <freemint> or to express in existing picolisp analoges: 1. 'client 2. @lib/http.l 3. @lib/form.l

14:48 <beneroth> yeah. I'm interested in telegram.l :)

15:04 <freemint> beneroth: how does that help me with my decision. I have 3 ideas how to do it.

15:04 <freemint> And i can not decide.

15:06 <beneroth> ah. well I understood as one idea. I would split the stuff up.

15:07 <beneroth> 1) telegram.l which does the wrapping, so which handles the telegram API. 2) telegram-hooks.l (or however you name it) to do your text parsing to trigger something 3). telegram-trigger.l for the IFTT features.

15:08 <freemint> good idea.

15:08 <freemint> I might start in a month or so earliest.

15:08 <beneroth> different layer for different tasks/logic. 1) is just encoding/decoding, pil API for telegram.

15:09 <beneroth> no pressure. :)

15:12 <freemint> beneroth: that really helped me -

15:12 <freemint> thanks

15:15 <beneroth> you're welcome

15:17 <beneroth> make an effort of trying not to build unnecessary restrictions on the lower level functions (e.g. type/number of arguments or so). I mean: try to only restrict what is necessary to be restricted to gain the intended functionality, but try to no have any other restrictions which are a result of "it will never be used that way anyway".

15:18 <beneroth> this is one of the things I learned and try to make a habit thanks to picolisp (applies also to other languages)

15:18 <beneroth> that way you get great re-usability and maintainability you can't plan for.

16:48 <beneroth> freemint, I watched the full talk. Thank you very much for pointing this out! Great talk!

16:50 <beneroth> I wish I would have found such an explanation of systemd a year ago. I researched a bit but not enough, as it appears. the notion of systemd being intended as a universal process-watchdog I kinda missed until this talk, and I can see the attractiveness for that concept.

16:51 <beneroth> humm.. I still suspect the systemd implementation is rather flawed, but I can see the concepts are meaningful and valuable.

16:51 <beneroth> I will forward the talk to other people. Thanks freemint !

21:26 <rick42> <beneroth> and in a way it serves as a good filter to keep the people away who are not really interested in understanding programming stuff, who would probably attempt to pollute picolisp with comfort.

21:26 <rick42> Read: pythonifying the language. lol :)

21:45 <beneroth> I don't know python enough to say so :)

21:45 <beneroth> One thing is having a tons of libraries, that must not be bad. But compromises on language principles is another thing.