mrcom has quit [Remote host closed the connection]
trumae has quit [Ping timeout: 260 seconds]
mrcom has joined #commonlisp
trumae has joined #commonlisp
masinter has quit [Ping timeout: 246 seconds]
Madsy has quit [Quit: Leaving]
ec has quit [Quit: ec]
Mandus has quit [Ping timeout: 248 seconds]
mrcom has quit [Remote host closed the connection]
mrcom has joined #commonlisp
Mandus has joined #commonlisp
mrcom has quit [Remote host closed the connection]
yottabyte has quit [Quit: Connection closed for inactivity]
mrcom has joined #commonlisp
random-nick has quit [Ping timeout: 248 seconds]
notzmv has quit [Ping timeout: 260 seconds]
mrcom has quit [Remote host closed the connection]
mrcom has joined #commonlisp
makomo has quit [Ping timeout: 246 seconds]
molson_ has quit [Quit: Leaving]
molson has joined #commonlisp
notzmv has joined #commonlisp
igemnace has quit [Remote host closed the connection]
waleee has quit [Ping timeout: 260 seconds]
dre has joined #commonlisp
notzmv has quit [Ping timeout: 258 seconds]
mixotricha has quit [Quit: Client closed]
dre has quit [Quit: Leaving]
masinter has joined #commonlisp
notzmv has joined #commonlisp
causal has joined #commonlisp
Bike has quit [Quit: Lost terminal]
<beach>
Good morning everyone!
yauhsien has quit [Remote host closed the connection]
mixotricha has joined #commonlisp
Bike has joined #commonlisp
Guest85 has joined #commonlisp
Guest85 has quit [Quit: Client closed]
kongo has joined #commonlisp
kongo has quit [Client Quit]
ec has joined #commonlisp
yauhsien has joined #commonlisp
leeb has joined #commonlisp
yauhsien has quit [Ping timeout: 258 seconds]
yauhsien has joined #commonlisp
<mfiano>
Good morning! I just published a release candidate of my first Common Lisp application. I need to gather some feedback from users and work on documentation before I give it a 1.0 release, but it has been working well for me for about a month of heavy use.
<mfiano>
Feel free to offer any criticism. I am aware I have some off style conventions and the project has all comments offline in a separate file until it is polished enough to convert into code comments and documentation strings, so it may be hard to understand :)
<mfiano>
s/off/odd/
saura has quit [Ping timeout: 260 seconds]
<CodeBitCookie[m]>
mfiano: Okay, I have an idea of BTRFS but not ZFS. However, I don't understand why you kept so little code in so many files. I see many files containing only 1 or 2 small methods. I like to learn from these projects, so that is why I am asking.
tyson2 has joined #commonlisp
<mfiano>
Some files only establish protocols and not implementations.
<mfiano>
It is organized into conceptual pieces of the problem domain.
yauhsien has quit [Remote host closed the connection]
yauhsien has joined #commonlisp
<hayley>
mfiano: I think you should topo-sort the code, as that guy yesterday suggested. :^)
<mfiano>
:/
tyson2 has quit [Remote host closed the connection]
<hayley>
):
<mfiano>
I think I will objectively object
yauhsien has quit [Ping timeout: 256 seconds]
mixotricha has quit [Quit: Client closed]
<CodeBitCookie[m]>
lol
<CodeBitCookie[m]>
What's the point of topo-sort? It sounds like the reverse of modularity to me. Although to be fair, I don't understand much about it.
<beach>
CodeBitCookie[m]: To implement one module, I use one directory, one ASDF system definition and one (perhaps more) package definition. The number of files in such a system is not important, nor is the order between them as far as modularity is concerned.
livoreno has quit [Ping timeout: 260 seconds]
livoreno has joined #commonlisp
<CodeBitCookie[m]>
beach: So what does that have to do with topo-sort?
<CodeBitCookie[m]>
Isn't everything in topo-sort interconnected? You gave an argument as to why manual memory management is bad, and you told me that since pointers have to point to other things in code, we don't get modularity, as changing one thing has to change the other and so on.
leeb has quit [Ping timeout: 258 seconds]
<beach>
If you sort the functions inside a module, you don't violate modularity.
leeb has joined #commonlisp
<beach>
It is not because pointers point to "other things" that modularity becomes a problem. The problem is knowing how many references there are to an object, something you don't need with automatic memory management (because it is done for you) and you must know in languages without automatic memory management, or else you have memory leaks and other problems.
bilegeek has joined #commonlisp
<CodeBitCookie[m]>
It looks like I am loop in circles, however you didn't mention how memory leaks can violate modularity.
<CodeBitCookie[m]>
looping*
<hayley>
They don't, but memory leaks are very undesirable.
<CodeBitCookie[m]>
Oh, I see.
<CodeBitCookie[m]>
Then what makes manual memory management unmodular
<hayley>
Because tracking who-frees-what needs to be part of the interface, which can restrict how a module can be used.
<CodeBitCookie[m]>
Ohhhh, okay.
<CodeBitCookie[m]>
Thanks *no-defun-allowed*!
<CodeBitCookie[m]>
Also, why is your username that?
<hayley>
Why it's no-defun-allowed? Because you should use defgeneric.
<CodeBitCookie[m]>
Why? Like Always? Even if your not using CLOS?
<hayley>
You probably shouldn't, as it's a joke.
<CodeBitCookie[m]>
Oh.
<CodeBitCookie[m]>
:)
prokhor_ has joined #commonlisp
hisacro has quit [Ping timeout: 258 seconds]
hashfunc1813 has joined #commonlisp
molson has quit [Ping timeout: 246 seconds]
yauhsien has joined #commonlisp
hisacro has joined #commonlisp
icer has joined #commonlisp
<utis>
is there a line-wrapping facility?
<Shinmera>
in what sense
<CodeBitCookie[m]>
utis: In format?
<utis>
in format or anywhere
<Shinmera>
but in what sense. wrap what and when?
<utis>
Shinmera: that will limit the number of columns and break at whitespace
<CodeBitCookie[m]>
Wow, Shinmera out here writing every library possible.
<Shinmera>
If only
yauhsien has quit [Ping timeout: 276 seconds]
<utis>
thanks!
ttree has quit [Ping timeout: 258 seconds]
<icer>
Hi, I'm trying to use croatoan for ncurses and set up emacs/slime to a cmdline sbcl/swank instance, which works, except connecting causes redifne warnings (WARNING: redefining SWANK::SYMBOL-INDENTATION in DEFUN) to stay on the terminal screen and apparently push the ncurses screen down. I've tried putting (sb-ext:muffle-conditions cl:warning) to suppress it, but they still show. Is there a
<icer>
better way?
<icer>
I've also updated my quicklisp and emacs packages, which resolved a mismatch error message on the slime side, but still get the warnings..
<icer>
slime/swank version mismatch error, that is
<jackdaniel>
the best way would be to track the redefinition bit and prevent it in the first place (if possible)
<jackdaniel>
other than that, perhaps you could bind the stream that is printed to to (make-broadcast-stream)
Fare has joined #commonlisp
<jackdaniel>
then the message will end up in a black hole
<icer>
I'm still somewhat new to lisp, and I'm not sure how to catch it, other than maybe start with a swank only stub to remove croatoan from the debugging.
<icer>
And is the result of (make-broadcast-stream) the stream that's a black hole? I guess warnings go to standard error?
<jackdaniel>
make-broadcast-stream accepts any number of streams that it forwards the message to
<jackdaniel>
if there are no streams, then it doesn't forward it
attila_lendvai has joined #commonlisp
pve has joined #commonlisp
hashfunc1813 has quit [Ping timeout: 258 seconds]
Fare has quit [Ping timeout: 240 seconds]
yauhsien has joined #commonlisp
<icer>
@jackdaniel Thanks, I'll try it. A "(let ((*standard-output* (make-broadcast-stream)) ..." is where I'm headed
<jackdaniel>
icer: I'd try first *error-output* if I were you
<jackdaniel>
n.b the convention on irc is "foo: dubi dobi dom", and not "@foo dubi dobi dom"
<icer>
jackdaniel: thanks
<jackdaniel>
sure
yauhsien has quit [Ping timeout: 246 seconds]
<icer>
Ok, let didn't work, but a top-level setf did. Thanks!
MajorBiscuit has joined #commonlisp
<jackdaniel>
sure, remember to return error-output to previous value after you successfully muffled the warning. also keep in mind that it is a kludge, not a solution
Colere has quit [Remote host closed the connection]
<jackdaniel>
lisp123: thanks, I was looking for something in this spirit
<lisp123>
So how would you interface ECL with the browser? Does it have access to DOM (thought WebAssembly didn't have that yet)
<jackdaniel>
if you want to interact with dom, you'll need to write a glue code to go through javascript
<mfiano>
Also unlike jscl, it doesn't crash the browser tab with 1d999
MajorBiscuit has quit [Ping timeout: 248 seconds]
<lisp123>
is there a standard interface between WebAssembly and JS?
aartaka has quit [Ping timeout: 260 seconds]
<lisp123>
as in calling JS from WA
<lisp123>
I guess there must be given how you had the msgbox
<jackdaniel>
there is a section in emscripten manual how to call js functions and vice versa
<lisp123>
oh its via emscripten
Everything has joined #commonlisp
<lisp123>
not direct
<jackdaniel>
what do you mean by "direct"?
<lisp123>
no emscripten
<lisp123>
so just ECL + own's code
<jackdaniel>
interesting definition of direct
<lisp123>
i just looked up emscripten
<jackdaniel>
as I've mentioned, I plan to experiment a little with wasi too
Everything has left #commonlisp [#commonlisp]
<jackdaniel>
nomen omen, ecl won't work without some kind of libc
<lisp123>
didn't know it was a compiler
<lisp123>
thought it was like emacs lol
MajorBiscuit has joined #commonlisp
<jackdaniel>
it is built without asdf, because that would only blow the executable size without any gain (there is no networking to pull extra libraries)
<lisp123>
this is pretty groundbreaking
<Nilby>
hmmm, how to wrap in xterm.js
<jackdaniel>
well, not really, you can preload libraries with your wasm and then asdf will make sense
<lisp123>
no I meant having ECL in the browser is ground breaking
<jackdaniel>
ah
Oddity has joined #commonlisp
<jackdaniel>
Nilby: thanks for the hint, I'll look into that too
aartaka has joined #commonlisp
<jackdaniel>
btw, if you want something less annoying (i.e not the input box), then downlaod ecl.wasm and ecl.js in the same directory and run them as `node ecl.js'
<jackdaniel>
it works™
<Nilby>
jackdaniel: this is cool. then maybe we could have more usable repl
<lisp123>
jackdaniel: have a look at codemirror too
<jackdaniel>
yes, the ultimate goal is to add official support to ecl with necessary documentation
<jackdaniel>
for a showcase a proper repl is needed
<lisp123>
its not a terminal, but if you want buffers etc, it would be one of your best bets
<jackdaniel>
I see
<lisp123>
author is ex-CLisper too (think he did postmodern?)
Oddity has quit [Ping timeout: 248 seconds]
<jackdaniel>
either way, I have other stuff to do too - it is prone to issues due to a bad glue between read/select and js reading, so if you encounter problems with that please ignore it for now; if you encounter other issues then please let me know - I'm leaving the website unless it crashes my server ;)
<jackdaniel>
s/unless/until/
<lisp123>
if I want to add lisp code to the image
<lisp123>
how would that work? does one precompile an ECL image and use that?
<jackdaniel>
the easiest method will be to statically link the library with ecl (so precomipile), alternatively you could preload a set of files during compilation
<jackdaniel>
there are other more sophisticated scenarios possible, but they would need to be implemented - i.e dropping a wasm module in a js widget - then it could be stored in the virtual fs and you could call dlopen
rodicm has joined #commonlisp
<lisp123>
precompile seems to make most sense
<jackdaniel>
also emscripten does present some ways to access the network, so it would be also possible with enough effort
dra_ has joined #commonlisp
Bike has joined #commonlisp
aartaka has quit [Ping timeout: 260 seconds]
<semz>
<jackdaniel> it is built without asdf <- Did you mean Quicklisp? Or do I misunderstand how you're supposed to use libs
<jackdaniel>
I've meant what I've told
<jackdaniel>
asdf is just a library in common lisp that happens to be often preloaded in the implementation
<jackdaniel>
quicklisp depends on asdf
<semz>
Oh wait, I get it now
azimut has quit [Remote host closed the connection]
aartaka has joined #commonlisp
Krystof has joined #commonlisp
attila_lendvai has joined #commonlisp
azimut has joined #commonlisp
<lisp123>
jackdaniel: does asdf really add that much size?
<jackdaniel>
let me see
<jackdaniel>
(rebuilding from scratch)
<lisp123>
o.o
<jackdaniel>
to compare resulting wasm file size
<jackdaniel>
because llvm wasm backend produces binaries of a different size than my native compiler
<jackdaniel>
erm, wrong compilation flags :)
<lisp123>
ah well :)
<jackdaniel>
so, with asdf 7.5MB vs 5.8MB without
<jackdaniel>
if we add gzip to the mix (because that's probably how that would be shipped on a real production server) it is 1.9MB vs 1.5MB
<jackdaniel>
not that bad
<lisp123>
I'm suprised asdf adds that much
<jackdaniel>
otoh asdf introduces extra startup time (that's the case also on linux)
<jackdaniel>
when ecl learns how to dump images the startup time will be greatly mitigated I hope
<lisp123>
its very promising already
szkl has joined #commonlisp
mixotricha has quit [Quit: Client closed]
<Nilby>
it's interesting that (find-package :ffi) doesn't return nil. i wonder if it can dlsym some fake system calls?
<jackdaniel>
dlopen etc won't work with this version; the package ffi has features like conversion between lisp and c types - if some c libraries are linked in wasm, and calls to them are also precompiled (i.e library compiled along with ecl), then you should be able to do ffi
<jackdaniel>
dlopen may eventually work, I did not investigate that yet but there are some resources on the internet how to use it
<jackdaniel>
as of a dynamic ffi - in ecl it is handled with libffi, and that library does not work on wasm currently; there is a limited hackish port made by a haskeller though
<jackdaniel>
so that may be a fine starting point for eager dynamic ffi calls hackers
lisp123 has quit [Remote host closed the connection]
<Nilby>
jackdaniel: it's theoretical at this point. i'm just wondering how i would port my code that uses o/s level functions.
<Nilby>
jackdaniel: thanks
<jackdaniel>
in principle WASI may act as an intermediate between the wasm runtime and the operating system, that depends on the runtime though
<jackdaniel>
I don't know whether there is a functionality to escape from the wasm runtime to call a function from a system shared object library
<jackdaniel>
I would be surprised if there were
<jackdaniel>
but if your code uses stuff like posix functions (fread and such), then they are implemented as part of either WASI or emscripten - the libc is available (although it is a bit esoteric at points)
<jackdaniel>
there is even working sdl port to emscripten that opens gl context in the browser
<Nilby>
ah, interesting
<jackdaniel>
i.e this
<jackdaniel>
(SDL2*) btw, that's how McCLIM will work in browser in 2033 (given my current time constraints!)
<Nilby>
browser is the new installer
<jackdaniel>
I'm personally very sceptical about web development, but hacking a new runtime is fun
<Nilby>
there's something enjoyable about making one's code run on everything
<Nilby>
not that one would want to use it there
<hayley>
I was thinking that some interesting kinds of interactive literate programming can be done, if one's favourite language can be run in a web browser.
<jackdaniel>
well, it surely helps with onbording - hey check this out http://evil-lisp.js
<jackdaniel>
instead of: download emacs
<jackdaniel>
done? ok, then here's slime, you need to add it to emacs.el
<hayley>
To keep the weirdness up, one demo I have in mind is a compiler book, which culminates in generating a native executable...that is then run in another emulator written in JS.
<jackdaniel>
ah, and download fnsCL
<jackdaniel>
well, you could output wasm code directly
<jackdaniel>
without an emulator
<hayley>
You can say you bootstrapped your compiler from a web browser, which few people can say.
<Nilby>
hayley: you've seen the SCIP with live in page execises
<hayley>
Sure, but I am not sure how good of a target WASM is yet. I think unwinding and GC roots might still be missing; haven't really paid attention to it.
<jackdaniel>
when you write a runtime that is managed by you, then you don't need a "platform gc", just like unix does not run gc for its proceses
<hayley>
I do need to be able to find roots in order to run GC. In Unix I can walk my own stacks to find pointers; I don't know how you do that in WASM.
<jackdaniel>
well, I'm walking my own stack of course! :)
<jackdaniel>
or at least, libgc does
<hayley>
I read somewhere that Emscripten has its own Boehm-compatible but not Boehm GC, which presumably avoids the problem somehow.
<jackdaniel>
I've compiled boehm for emscripten (from pre-release 8.2)
<jackdaniel>
it works well. the release before that doesn't work so well because, indeed, it can't scan the stack
<jackdaniel>
s/that//
<jackdaniel>
well, it compiled, but I had to disable the gc for the time of running a function
<jackdaniel>
otherwise it collected things that were stack-allocated
<hayley>
Then I'm not sure how Emscripten lets one walk the stack. I don't believe you can do that in WASM.
<hayley>
I could maintain a shadow stack, but that's nasty.
<jackdaniel>
what can I say to you? libgc works just fine and it scans the stack
<jackdaniel>
I did not investigate emscripten internals to learn how
<jackdaniel>
you may look for the function emscripten_scan_stack
Fare has joined #commonlisp
<jackdaniel>
comments in headers claim, that it is api that gives access to introspecting the wasm data stack
<hayley>
emscripten_scan_registers requires spilling registers to linear memory, by the description there. I was not aware of any way to introspect the data stack, after having skimmed the WASM spec some time ago.
<jackdaniel>
either way, good that they allow that; otherwise ecl wouldn't be able to run its gc :)
Fare has quit [Client Quit]
<hayley>
Sure. It affects how I'd approach code generation still, if it is necessary to spill to somewhere in linear memory, in order to scan the "stack".
<jackdaniel>
I imagine that this still would be cheaper than adding another layer of emulation
<hayley>
Quite possible, though I'm not sure if performance in the browser is that pressing. It is important, but the emphasis is not so much on using the finished implementation, let's say :)
roklein has joined #commonlisp
Fare has joined #commonlisp
Fare has quit [Remote host closed the connection]
Fare has joined #commonlisp
Fare has quit [Remote host closed the connection]
livoreno has joined #commonlisp
yauhsien has quit [Remote host closed the connection]
Guest74 has joined #commonlisp
Fare has joined #commonlisp
<Guest74>
any thoughts on Degrees VS Radians in vector graphics?
<Guest74>
I'm adding the ability to specify relative points by angle and length, but that requires a concrete orientation besides the format. Compass rose style clockwise, 0 at top? math like from the x axis +- 180, clockwise? counterclockwise?
<White_Flame>
doesn't really matter. you could do 0.0-1.0 for a full circle's angle if you want
<White_Flame>
no reason to lock into a single representation either
<White_Flame>
doesn't really matter. you could do 0.0-1.0 for a full circle's angle if you want
<White_Flame>
but it would probabl compile down to radians ccw fom +x in math
<Guest74>
except when you have the ability to specify paths in a string format you need a single representation.
<White_Flame>
the string can specify units, too
<Guest74>
dispatching on 1 char
<Guest74>
like a glorified svg path.
<Guest74>
also, if I'm making a generic functions if I don't go for one representation then I'll have to wrap numbers to be able to dispatch to degrees/radiansetc...
<Guest74>
I usually only think in relative degrees so I'm not sure what orientation is intuitive.
ldb has joined #commonlisp
<ldb>
morning everyone
MajorBiscuit has quit [Quit: WeeChat 3.5]
igemnace has quit [Remote host closed the connection]
MajorBiscuit has joined #commonlisp
random-nick has quit [Ping timeout: 246 seconds]
livoreno has quit [Ping timeout: 272 seconds]
rodicm has quit [Quit: Leaving]
molson has joined #commonlisp
<CodeBitCookie[m]>
morning ldb
Fare has quit [Remote host closed the connection]
orestarod has joined #commonlisp
artchad has joined #commonlisp
Fare has joined #commonlisp
<artchad>
Hi, I need some help. AllegroServe has #'publish-file. I want to loop over all files in a given directory an call publish-file on those. The tricky part is to get the list of directories. My first idea was to call out to the `ls' command and map over all the listed files. I just came back from a programming break, so I wouldn't know how to actually retrieve all those file names.
<artchad>
wait
<artchad>
actually I think there is a publish-directory function.
attila_lendvai has quit [Ping timeout: 258 seconds]
pranavats has left #commonlisp [Disconnected: Replaced by new connection]
pranavats has joined #commonlisp
<artchad>
but out of curiosity, how would you call `ls' and map over the result?
<Guest74>
You wouldn't call ls. You would probably use DIRECTORY with some combination of wilds to do what you want, depending on your implementation.
<artchad>
oh, #'directory is a build in. Thanks for the pointer. I didn't really work a lot with fileIO yet. Libraries did that stuff for me. :^D
<ldb>
(directory "*.png")
<ldb>
sometimes I also use lparallel if the tasks are independent
<artchad>
I'd guess lparallel is asynchronous is some way?
<artchad>
or parallel I guess
sterni has quit [Quit: WeeChat 3.3]
<artchad>
(directory "./banners/*.png") gives a neat list of paths. Thanks.
<ldb>
lparallel is taking advantages of multi threads
<Guest74>
just note that they are paths, and somethings prefer NAMESTRINGs
livoreno has joined #commonlisp
<ldb>
I think AllegroServe can handle that
sterni has joined #commonlisp
rodicm has joined #commonlisp
<artchad>
I'm currently working on a .onion site, which was a lot easier than anticipated. Tor makes it's proxy totally tranparent. So AllegroServe worked out of the box with a non-root user. Just had to make sure to bind a rather high port.
<ldb>
right low port numbers are for root user
<artchad>
I wish more people used Tor. No need to have a static IP address, so no need to rent a VPS, no need to rent DNS from shady organizations. It makes web hosting easy and fun. But people are scared or something. xD
<ldb>
well tor is more likely to be monitored, I concerned.
<Guest74>
if you can self host
<ldb>
not intented to be advertising, but sdf.org offers website hosting for a very low rate
<ldb>
and you can meet lots of people there
ldb has quit [Quit: ERC 5.4 (IRC client for GNU Emacs 28.1)]
<artchad>
Sounds interesting. But can I want my Lisp REPL. > . <
<semz>
key=address is really nice, but requiring custom software is probably the killer for most
<semz>
did you write your own SOCKS5 lib for the lisp side?
<artchad>
I was surprised that I didn't need to.
<semz>
or wait, that'd be for the client side I think
yauhsien has joined #commonlisp
<artchad>
for the client side I use torsocks
Fare has quit [Remote host closed the connection]
thuna` has joined #commonlisp
rodicm has quit [Ping timeout: 240 seconds]
ttree has joined #commonlisp
azimut has quit [Ping timeout: 240 seconds]
tyson2 has joined #commonlisp
Fare has joined #commonlisp
kpoeck has joined #commonlisp
karlosz has joined #commonlisp
dra_ has quit [Ping timeout: 240 seconds]
kpoeck has quit [Quit: Client closed]
kpoeck has joined #commonlisp
karlosz has quit [Ping timeout: 240 seconds]
kpoeck has quit [Quit: Client closed]
MajorBiscuit has quit [Ping timeout: 246 seconds]
kpoeck has joined #commonlisp
leeb has quit [Ping timeout: 246 seconds]
Dynom has joined #commonlisp
prokhor__ has quit [Remote host closed the connection]
kpoeck has quit [Ping timeout: 252 seconds]
yauhsien has quit [Remote host closed the connection]
yauhsien has joined #commonlisp
yauhsien has quit [Ping timeout: 240 seconds]
kpoeck has joined #commonlisp
<ecraven>
gabc: thanks, that's one of the reasons I asked, I was wondering whether there was anything newer ;)
yauhsien has joined #commonlisp
<gabc>
ecraven: that is fair, I haven't looked much but I have hard time resisting showing that I know stuff sometimes :P
<ecraven>
gabc: thanks for pointing it out though, others will be interested too!
Guest74 has quit [Quit: Connection closed]
Fare has quit [Remote host closed the connection]
phadthai has quit [Ping timeout: 240 seconds]
tyson2 has quit [Remote host closed the connection]
tyson2 has joined #commonlisp
random-nick has joined #commonlisp
roklein has quit [Quit: Leaving]
rodicm has joined #commonlisp
tyson2 has quit [Remote host closed the connection]
icer has quit [Ping timeout: 276 seconds]
jmdaemon has joined #commonlisp
Oddity has joined #commonlisp
phadthai has joined #commonlisp
skyl4rk` is now known as skyl4rk
scymtym has quit [Read error: Connection reset by peer]
Brucio-61 has quit [Remote host closed the connection]