freenode/#lisp - IRC Chatlog
Search
14:48:35
fiddlerwoaroof
I've also tried atom-slime and it works, but I've never heard of anyone that uses it.
18:07:21
jmercouris
hey everyone, indiegogo campaign finished, we raised 1850$ for working on Next, thank you to everyone that contributed
18:07:58
jmercouris
I hope to deliver something that will work nicely on macOS and Linux, I'll post any new major releases in this channel, but so far, things are going well
18:40:15
jmercouris
deprecation warnings in general are annoying, as are deprecations and any system behavior changes
18:49:13
fiddlerwoaroof
Yeah, but adding arbitrary rules to a central part of a programming ecosystem should only be done for really serious reasons
18:50:21
fiddlerwoaroof
Plus, as far as I can tell these changes were pushed through without community buy-in
18:51:44
|3b|
as far as i can tell, nobody wants to help maintain asdf, so they get what they get :/
19:08:46
jackdaniel
last time I've tried to contribute to asdf I had a very draining discussion about merge request, but there is a chance that the problem was an incompatibility of my personality with the peer reviewer (so I gave up)
19:32:46
fiddlerwoaroof
Eventually I'll get around to writing lisp bindings to Linux's cgroups stuff
19:59:58
jmercouris
fiddlerwoaroof: how is the obj-c bridge going via CFFI? did you get the SBCL app to actually work?
20:00:57
jmercouris
s/compilese/compiles - too much time in a foreign country, my language skills are really messed up
20:01:20
jmercouris
though if you wait a little bit, like maybe a month or so, there should be a new Linux release
20:02:16
mercourisj
with almost no dependencies, possibly distributed via your package manager or flatpak, undecided
20:22:06
fiddlerwoaroof
jmercouris: it works pretty well for me, but the code isn't in a great state for someone else to understand what's going on :)
20:22:31
fiddlerwoaroof
ccl is more reliable than sbcl as far as producing a working app bundle goes, though
20:23:36
mercourisj
that's where I figured a lot of the CCL code spent time, making a compatibility layer with Cocoa itself, which you'd have to reinvent and manually invoke
20:23:40
fiddlerwoaroof
Nah, as long as you disable sbcl's attempts to catch floating point exceptions, the two runtimes get along pretty nicely
20:23:58
mercourisj
interesting, so it is a separate cocoa process you are invoking and interoping with?
20:24:50
fiddlerwoaroof
The part I haven't had time to figure out is things like turning objective-c errors into lisp exceptions
20:25:45
mercourisj
and leaves lots of space for all sorts of features to make their way into production code
20:26:07
fiddlerwoaroof
ccl has an advantage here in that it was actually developed by Apple/Next for a while, as I understand it, so they've already done a lot of the hard work implementing the bridge
20:26:40
mercourisj
I remember talking to RME about the lineage and it was a little bit different than what I thought
20:27:25
fiddlerwoaroof
"Well anyway, after that some stuff happened, and Apple ended up acquiring Coral Software because Apple at the time was using Lisp and Smalltalk and other dynamic languages to do their research. So that happened"
20:29:05
fiddlerwoaroof
It went, as I understand Coral Comon Lisp -> MCL/OpenMCL -> Clozure Common Lisp
20:31:04
fiddlerwoaroof
Anyways, my bridge is really useful for the sorts of things one might use Applescript for already
20:31:28
fiddlerwoaroof
E.g., here's a little demo of getting the current itunes track info: https://github.com/fiddlerwoaroof/objc-lisp-bridge/blob/master/scripting-bridge.lisp
20:32:12
fiddlerwoaroof
And here's a little bit of code I wrote for converting Safari's reading list into an org-mode file: https://github.com/fiddlerwoaroof/objc-lisp-bridge/blob/master/reading-list-reader.lisp
20:33:40
fiddlerwoaroof
(I actually run that one in a cronjob so I can refile my reading list stuff into my bookmarks.org file)
20:35:15
fiddlerwoaroof
But, my current project (slowly progressing project) is to figure out Mcclim so I can either ressurect the beagle backend using my system or write a new backend
20:38:25
mercourisj
fiddlerwoaroof: what's that license on the code you posted? Can I include it in Next?
21:10:57
fiddlerwoaroof
And, I don't notice any direct dependencies that are incompatible with such a license
21:41:37
sjl
You can clone it into local-projects and update to whatever version you want, and that will take precedence over whatever's in the quicklisp dist
21:45:24
mercourisj
ivy-imenu-anywhere doesn't seem to work so well with the system, only with open files
21:46:23
mercourisj
I guess I could open all the files in the system, but there must be an easier way :D
23:37:06
aeth
When I want to find a defun by name I use M-. in the SLIME REPL as if it was M-. in a source file
23:58:52
fouric
("technically" being that it's not a Hard Problem to write a compiler that supports the thing, but the CL spec wasn't written to allow it)
0:22:37
fiddlerwoaroof
fouric: the issue I've had is that they don't have access to the type inference information that some implementations compute)
0:24:26
fiddlerwoaroof
(of course type inference is an implementation detail in CL, so you'd need to standardize that too)
1:54:38
p_l
emaczen: while not exactly the answer, any good old description for making "uberjar" manually should work