07:19 <tankf33der> My next task will be

07:19 <tankf33der> https://rosettacode.org/wiki/Canonicalize_CIDR

07:19 <tankf33der> completely do not understand

07:27 <tankf33der> the context

07:34 <abu[m]> interesting

07:50 <abu[m]> This is a use case for 'rev'

07:51 <abu[m]> (hex (rev 32 (dec (** 2 12))))

07:51 <abu[m]> -> "FFF00000"

08:03 <abu[m]> OK, solved the task

08:11 <abu[m]> Tell me if I should post it here

08:55 <tankf33der> http://ix.io/46NJ

08:55 <tankf33der> my implementation

08:56 <abu[m]> Very good!

08:56 <tankf33der> printing will do later

08:57 <abu[m]> and chop and split of string is needed I think

08:59 <tankf33der> i can do printing on lists or strings, do not what is better

09:03 <abu[m]> I stole your (sum >> (-24 -16 -8 0) L)

09:03 <abu[m]> Then I get http://ix.io/46NN

09:03 <tankf33der> i see

12:33 <abu[m]> @tankf33der , your (sum >> ..) is a good candidate for a #PicoLispRepl tweet. OK to post it?

12:34 <tankf33der> sure.

12:34 <abu[m]> ☺

12:34 <tankf33der> http://ix.io/46OD

12:35 <tankf33der> full printing task

12:36 <Hunar> Hello :) quick question how to turn the resulting string from (argv) into a number

12:36 <abu[m]> Good. Just the indentation of the (for A loop looks strange in the browser.

12:37 <abu[m]> Hi @Hunar! 'format' is to convert to/from number

12:37 <abu[m]> But usually (format (opt)) is better

12:38 <tankf33der> http://ix.io/46OE

12:38 <tankf33der> fixed.

12:38 <Hunar> Great :D thanks

12:38 <abu[m]> @tankf33der: Perfect!

12:42 <tankf33der> ~160th rosetta task solved.

12:43 <abu[m]> \😃/

12:46 <tankf33der> afk.

13:10 <Hunar> I saw a benchmark at http://github.com/unipickle/Benchmarks I ran some examples from array-reverse, I got these different results (C 0.001s, Java 2.932s, Go 4.225s, Python3 builtin 3.279s, pil 11.266) here is my pil code ix.io/46OP I don't know how python is faster than java for me :/ does my picolisp code contain major mistakes? or the speed is expected

13:10 <Hunar> I think my github link is broken

13:10 <Hunar> http://github.com/unixpickle/Benchmarks

13:11 <Hunar> I ran 100000 items and 100000 iterations like he did

13:13 <abu[m]> So N is 100000x ?

13:13 <Hunar> yes, both are 100,000

13:16 <abu[m]> Hmm, this finishes in 0.00 sec

13:16 <abu[m]> oops, forget that ;)

13:18 <abu[m]> I think your pil code is optimal

13:18 <abu[m]> I did not really check the task, but the other implementations used arrays, right?

13:34 <Hunar> sorry I suddenly got very busy at work..

13:34 <Hunar> yes the used arrays

13:35 <abu[m]> No problem!

13:35 <abu[m]> OK, but then the benches cannot really be compared

13:35 <abu[m]> Traversing a linked list is much more expensive, and cache hits are worse too

13:38 <Hunar> hmm, you're right..

13:39 <Hunar> Now that I think about it, who needs to reverse an array in real life 😅

13:40 <Hunar> especially 100,000 times

13:40 <abu[m]> haha, true, but it happens

13:40 <abu[m]> In case of Pil, the efficiency of the 'flip' primitive is measured

13:41 <Hunar> The benchmark also measured built-in functions seperately .. what is an easy way to manually reverse a list in pil?

13:42 <abu[m]> The 'flip' function in Pil is a little slowed down as it checks for circularity on each cell

13:44 <Hunar> oh

13:46 <abu[m]> Wait, I'm wrong. Not really, but it gracefully handles circular lists not to loop infinitely