libera/#commonlisp - IRC Chatlog

7:22:54 boeg Anyone have experience with caveman and mito? I'm not quite sure how to connect the two properly. Is all that needs to be edited found in the db.lisp file, or is there other places where I need to do something so caveman2 is configured properly with mito?

8:21:20 Cymew It's Fukamachiware, which usually means it's not that well documented, and I get the impression you need to have done web development in ruby to understand what the code is trying to achieve.

8:21:56 Cymew I don't even understand what mito is used for. There seems to be some kind of db support in caveman these days, but maybe you're interested in doing something more complex.

8:30:26 Zakkor_ ** NICK Zakkor

8:37:17 jmercouris what are monads? can someone explain them in a way I can understand?

8:37:30 jmercouris I've been working with this question since university...

8:38:10 jmercouris is it just a chain of funcall results???

8:40:49 aeth jmercouris: https://stackoverflow.com/questions/3870088/a-monad-is-just-a-monoid-in-the-category-of-endofunctors-whats-the-problem

8:40:51 phoe Lisp is a poor language to use monads in because Lisp has mutable state - see https://marijnhaverbeke.nl/monad.html

8:41:09 aeth jmercouris: It's basically a thing that satisfies algebraic properties

8:42:21 TMA jmercouris: generally, in order to understand monads you need to understand monads first. people who understand monads lose the ability to explain them to people that dont

8:43:20 phoe I'll allow myself to post https://en.wikibooks.org/wiki/Haskell/Understanding_monads and route the discussion to #lispcafe since it clearly isn't about Lisp (since Lisp doesn't force you to understand monads in order to use the language in a sensible way)

8:45:28 jmercouris thanks for the links, maybe it will help to read marijn's post

8:46:10 pjb jmercouris: In CL, method-combinations are monads.

8:46:13 phoe or join #lispcafe

8:46:41 jmercouris I left lispcafe sometime ago, because the discussion was not so friendly at times :-(

8:47:24 phoe you can always escalate to the mods

8:50:02 phoe like, e.g., me

8:50:26 jmercouris just read the Marijn article

8:50:34 jmercouris very fascinating stuff, I'm glad we don't have that nonsense in CL

8:50:48 phoe we don't need to, since CL allows imperative programming

8:50:51 jmercouris this obsession with "purity" makes things unusable

8:51:08 jmercouris what I like about CL is its pragmatism

8:51:28 jmercouris it is not just a toy language, restricted to academic ivory towers

8:51:49 phoe neither is Haskell, but that's already off-topic

8:51:49 jmercouris yet it is fully suitable for academic research as well due to some wonderful properties

8:52:01 phoe ACTION goes back to writing

8:52:02 jmercouris I didn't bring up haskell, that was you :-D

8:52:10 jmercouris I'm just preaching to the choir here :-D

9:07:29 pjb jmercouris: if you need more links about monads: http://haskell.readscheme.org/monads.html http://dorophone.blogspot.com/2011/04/deep-emacs-part-1.html http://dorophone.blogspot.com/2011/05/hyperturtle-monad-makes-pretty-pictures.html

9:07:29 pjb

9:07:55 pjb jmercouris: notably the deep-emacs page explains monad to emacs lisp programmers, so it might be particularly understandable.

9:54:04 Lord_of_Life_ ** NICK Lord_of_Life

10:35:24 Shinmera Just a note on yesterday's discussion :) https://shinmera.github.io/dns-client/

10:35:55 phoe holy cow you're fast

10:36:23 jackdaniel and furious

10:38:59 phoe when Shinmera gets angry he constructs a dns client in a day

10:39:19 Shinmera I ain't even mad

10:39:57 phoe at this rate you gotta get an animal groomer license

10:40:21 phoe or the police might come knocking at your door with questions about all the yaks you have shaved

10:46:48 phoe I gotta finish being angry myself

10:47:09 phoe my angry is currently at 143 pages and 6 ASDF systems

10:47:47 jackdaniel and what exactly are you angry at? perceived inconsistence between handler-case and handler-bind?

10:53:46 phoe nope, not really

10:54:14 phoe I was originally angry at the fact that there's no "good" tutorial to the Lisp condition system that shows how to construct it from scratch

10:55:07 phoe and now I'm angry at the fact that my tutorial turned into a book

11:07:11 Xach a book is ok

11:07:43 _death ACTION is concerned, not yet angry, about alexandria :/

11:08:30 phoe _death: which parts exactly?

11:11:40 _death not sure I want to go into it right now.. maybe it's better to ignore and hope for the best

11:13:13 phoe the only hopes that are valid for alexandria are hopes that it'll become as ossified as ANSI CL

11:13:23 phoe ;; at least in my opinion

11:28:02 pjb Shinmera: told you…

11:28:12 Shinmera I didn't doubt you!

11:30:13 pjb DNS-PORT should not be a constant…

11:32:47 lemoinem ** NICK Guest65992

11:39:56 Xach ACTION feels a quicklisp dist shaping up

11:44:11 Shinmera !

11:55:39 phoe yes, you feel something. I must know what you are feeling

12:28:42 Xach the auspices and omens all tell the same tale

12:28:55 Xach "nothing major is broken and it is almost the end of the month"

12:29:07 phoe quick, time to break something!

12:31:57 Xach don't you dare

12:32:08 phoe okay

12:40:03 MichaelRaskin Hm, looks like some of the disappeared packages are now reuploaded by other people (or at least clx-truetype is…)

12:42:25 Bike beach: I didn't see anyone mention it - in a lexical environment with flet bindings for "f", the global setf expander for "f" doesn't apply

12:42:55 beach Bike: I see. Thanks.

12:43:24 beach Very important.

12:47:00 Bike also, if it does end up being a macro form, you have to use macroexpand-1 rather than macroexpand, so that if the thing immediately expanded into has a setf expander, you use that instead of macroexpanding

12:47:16 pjb (defun (setf f) (nv x) (print (list '(setf f) nv x)) x) (flet ((f (x) (print (list 'f x)) x)) (incf (f 2))) #|

12:47:16 pjb (f 2)

12:47:16 pjb ((setf f) 3 2) --> 2 |#

12:47:27 pjb Bike: why do you say so?

12:47:44 beach Bike: Another good point, yes.

12:47:56 pjb Bike: nothing is said in clhs about shadowing (setf f) by (flet ((f )) …)!

12:48:28 pjb (defun (setf f) (nv x) (print (list '(setf f) nv x)) x) (flet ((f (x) (print (list 'f x)) x) ((setf f) (nv x) (print (list '(new setf f) nv x)) x)) (incf (f 2))) #|

12:48:28 pjb (f 2)

12:48:28 pjb ((new setf f) 3 2) --> 2 |#

12:48:40 pjb if you want to shadow it, you need to bind it in flet yourself!

13:02:27 lieven 5.1.2.9 seems to be applicable

14:44:15 Bike Shinmera: by the way, you could define a default single-float-bits and double-float-bits using integer-decode-float, assuming they're supposed to return the IEEE754 formats... it'll work, though it'll be slower than what the implementation can do, and I don't think you can do it the other way precisely

14:45:02 Shinmera Hmm.

14:46:12 Shinmera Not sure that's a good idea, since you said /assuming/ they return the IEEE formats.

14:46:29 Shinmera So it would require testing per implementation, anyway.

14:51:07 Bike well, i meant, assuming that's what you want returned. you shoudl be able to tell from standard functions whether a single-float will fit.

14:51:35 Bike like, check float-radix is 2, float-digits is 24 or less for a single, and the exponent is in range.

14:51:50 Bike and if not you can do a not supported error.

14:52:45 Bike there's also a standard feature for ieee754 conformance

16:17:28 [X-Scale] ** NICK X-Scale

18:00:59 anlsh can anyone see thing?

18:01:03 anlsh this*

18:01:34 Bike hi.

18:02:20 anlsh oh thank god, I've been trying to figure out the issue with erc for an hour :|

18:03:13 anlsh Turns out I wasn't actually authenticating correctly, which prevented me from being able to see others' messages

18:30:58 phoe anlsh: I see you, yes

18:40:01 anlsh I started a small generics implementation, any comments? https://gist.github.com/anlsh/653cb92552063e9d209b6415b1bc01b1

18:40:03 jackdaniel what do I spy, with my little eye? I spy anlsh with my little eye!

18:41:04 anlsh Haha thanks guys, I can see you too :D