01:43 <nisstyre> > (for/list ([_ (in-range 10)]) 4)

01:44 <nisstyre> or just (build-list 10 (const 4))

01:45 <nisstyre> probably that's the right way

01:45 <nisstyre> but neither of those will do the same thing as [0]*n in python

01:45 <nisstyre> because it's creating copies, not references to the same value

04:12 <aaaaaa> I want to write "(foldl or #f l)", but bad syntax. Have to "(foldl (lambda (x y) (or x y)) #f l)". Why not?

06:15 <dzoe> samth: http://pasterack.org/pastes/24542

06:16 <dzoe> samth: of course I am not sure about how the set! hack affects everything, but that's how I get the disassembly without much hassle there

06:16 <dzoe> samth: for the count1 function, everything gets unboxed and the loop does not allocate memory at all

06:17 <dzoe> samth: with count2, where I am adding step values to u and v variables, everything is boxed and unboxed and allocated in each and every iteration

06:17 <dzoe> I am talking about latest CS master, of course ...

07:00 <samth> aaaaaa: or is a macro, not a function

07:03 <samth> dzoe: I would expect that the set! is potentially problematic

07:25 <dzoe> samth: but without that I am unable to disassemble the loop

07:28 <dzoe> And measurements suggest, that no unboxing happens really :)

07:29 <dzoe> I must probably look into extending your disassemble to support "follow call" feature

07:54 <phoe> hello, #racket

07:56 <phoe> I am organizing a series of Online Lisp Meets that started after this year's electronic European Lisp Symposium - the announcement mailing is at https://mailman.common-lisp.net/pipermail/online-lisp-meets/ and the talk videos so far have been posted at https://www.youtube.com/channel/UCymtXMj1M7cKiV9TKLoTtEg

08:00 <phoe> so far the talks have been mostly utilizing Common Lisp for the practical parts (I guess because I am a CL programmer and most of the Lisp people I know are CL people), but obviously that isn't the idea behind ELS nor I want that to be the idea behind this post-ELS online series - I've wanted this place to be a meeting grounds for people doing more Lisp dialects than just CL, and the people who've been

08:00 <phoe> joining us on Twitch so far share the same wishes

08:02 <phoe> so, since I'm already doing a wall of text - please let me know if you're working on something related to Lisp in any any way and you consider it interesting enough to record a video about it, anywhere from a few minutes to ~1h. I've decided to have a go at the meeting series to provide a place where the wider Lisp community can meet and talk and where the ideas can mix and breed, and so far it's been

08:02 <phoe> well-received on #lisp - it's just that it's comfortable to post stuff there because it's my online home, much unlike e.g. here on #racket

08:03 <phoe> please feel free to throw your ideas and videos at me, and I'll be happy to host them on Twitch for everyone to see - thanks a lot. ;; wall of text ends here

09:46 <dzoe> phoe: I'd address racket-users@ ML with this - not everyone who might want to be involved is here and I can think of at least few people that might be interested and read the ML

09:47 <phoe> dzoe: OK - what is the mailing list server?

09:50 <phoe> I found users@racket-lang.org over at https://lists.racket-lang.org/

09:54 <dzoe> Oh

09:55 <dzoe> it is racket-users@googlegroups.com

09:55 <phoe> gasp

09:55 <dzoe> But I am not a Google user, so I am subscribed the old-fashioned way.

09:55 <phoe> OK - I'll use that one. Thanks!

11:35 <samth> dzoe: did you try just lifting it out to the top level?

11:43 <phoe> dzoe: posted, thank you.

12:20 <dzoe> samth: making the loop a top level procedure with tail-call

12:21 <dzoe> ?

12:21 <dzoe> Because just lifting the loop does not allow me to disassemble it

12:21 <samth> right

12:22 <dzoe> Btw, those safe-fl+ and other tricks are Matthew's suggestions and they seem to match the np-unbox-fp-vars! pass

12:22 <dzoe> Although it seems to me that let loop will be treated differently than the tail-recursive function.

12:23 <dzoe> Although there are cases for calls as well ... alright - I'll re-test.

12:23 <dzoe> That's a good idea anyway.

12:24 <samth> dzoe: let loop just expands to exactly the code you would write

12:25 <samth> the only difference is passing in the free variables vs referencing them from the outer environment

12:26 <dzoe> samth: I know, but it looks like the np-unbox-fp-vars! stage is somewhere in between, really I am not fully accustomed to the whole cpnanopass.ss details (but I am working on that)