Conversation

Notices

1. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 20:01:51 EST clacke
   Listening to SPJ, man this guy has so much joy and energy every time!
   
   Now he's going into "the life of a language designer" and presents a graph of the most common uptake rate of a new language.
   
   https://youtu.be/re96UgMk6GQ https://social.heldscal.la/attachment/1012211
   1. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 20:20:04 EST clacke

      in reply to
      He goes on to describe how the "successful research language" dies the "slow death" after a decade and after having peaked with a couple of hundred users, and that the third road is when you cross the boundary of immortality, at which point you become the new COBOL, like Java, C, Python, Perl and other things that humanity will now never be rid of, held up by the network effect of legacy code.
      
      Haskell, interestingly, plateaued in an intermediate stage for a decade, but now seems destined to become one of the immortal legacy languages.
      1. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 20:51:59 EST clacke

         in reply to
         Lots of references, mostly verbal, to seminal papers, including "Cons should not evaluate its argument", "Lambda the ultimate <blah>" and "Why functional programming matters". I hope this is all written down somewhere or I'll have to do it.
         1. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 21:01:21 EST clacke

            in reply to
            > F# calls them "workflows" rather than "monads", which is a very sensible idea.
            1. Nobody [Linux Walt Alt (@lnxw37a2)] (lnxw37a2@pleroma.soykaf.com)'s status on Saturday, 25-Nov-2017 21:31:42 EST Nobody [Linux Walt Alt (@lnxw37a2)]

               in reply to
               @clacke Is that to avoid the reputation of monads (they attract or create spittle-dribbling fanatics)?
               1. clacke (clacke@social.heldscal.la)'s status on Sunday, 26-Nov-2017 00:42:29 EST clacke

                  in reply to
                  @lnxw37a2 That's what he is implying.
            2. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 21:35:46 EST clacke

               in reply to
               80% of the slides in this talk are masterpieces of self-deprecating humor and love. https://social.heldscal.la/attachment/1012391
               1. clacke (clacke@social.heldscal.la)'s status on Saturday, 25-Nov-2017 21:36:08 EST clacke

                  in reply to
                  > When people first start to use Coq or Isabel, you don't hear from them for two years, they sort of go dark. And then they emerge, with all their pain receptors destroyed, and they say "it's the only way to live".
                  1. ay (ayy@gs.smuglo.li)'s status on Saturday, 25-Nov-2017 21:37:45 EST ay

                     in reply to
                     @clacke I tried to learn coq it seems so amazing! But I never put in enough time.
                     1. clacke (clacke@social.heldscal.la)'s status on Sunday, 26-Nov-2017 09:01:41 EST clacke

                        in reply to
                        @ayy Yeah it's on my todo as well. I have experienced some other model proof language in a uni course, but now the name escapes me.
                        1. clacke (clacke@social.heldscal.la)'s status on Sunday, 26-Nov-2017 11:48:47 EST clacke

                           in reply to
                           @ayy Hah! I found it!
                           
                           Took me a while. I finally had to use the right terms to find the name of my formal methods professor and then found one of her lecture schedules that mentioned it. :-D
                           
                           http://spinroot.com/spin/whatispin.html
                           
                           It's pretty cool. If I remember correctly, you provide it with an implementation (in the tool's own special language) of a multi-threaded application, and it exhausts the combination space and shows you a sequence of operations that leads up to one of your asserts breaking.
                           
                           It also has some interactive modes that I forgot how they work or maybe we never went into it. The thing that stuck rather than the tool itself was the concept of Kripke structures for modeling states through time.
                           
                           There was also a number of formal logic models of control structures for pen-and-paper proofs of invariants that *should* have stuck. :-)
                           1. clacke (clacke@social.heldscal.la)'s status on Sunday, 26-Nov-2017 12:04:51 EST clacke

                              in reply to
                              Also cool: The iSpin GUI wrapper around the CLI spin is in !tcltk :-)
                  2. clacke (clacke@social.heldscal.la)'s status on Sunday, 26-Nov-2017 12:21:25 EST clacke

                     in reply to
                     #Albatross seems a bit cool.
                     
                     http://albatross-lang.sourceforge.net
                     
                     Not super active community, but latest commit at https://github.com/hbr/albatross is only a few weeks ago.
                     1. Zordon (Linux Walt) (zordon@mastodon.social)'s status on Sunday, 26-Nov-2017 12:26:54 EST Zordon (Linux Walt)

                        in reply to

                        @clacke Hmm. Are they not aware of the "albatross around your neck" terminology? https://www.phrases.org.uk/meanings/30800.html