freenode/#lisp - IRC Chatlog
Search
21:31:47
k-hos
any ideas why calling (compile-file "test.lisp" :system-p t) in ECLs repl would emit a .fasc file instead of an .o
21:54:45
k-hos
their examples on the site say :system-p t is supposed to make it emit .o files by invoking gcc
22:10:51
k-hos
but I can't find any information about how to manually specify the compiler, or get it to emit just .c files
22:34:01
oni-on-ion
k-hos: and you've seen this? https://common-lisp.net/project/ecl/static/manual/ch26.html
22:38:11
k-hos
standalone program is what I am trying to get, preferably by compiling all files involved and not loading them at run time
22:39:56
oni-on-ion
k-hos: well. bulding programs and choosing a C compiler i think are very related concepts. you did say that you checked the manual and i did not intend to not believe you but i did a google search and thats where i ended up
3:50:06
SaganMan
It's fairly good morning today. Mangkhut has passed China and things are already back to normal.
3:53:33
SaganMan
yeah, I'm now talking to friend who lives in Shenzhen. He's safe now but he had hell of experience in typhoon.
4:12:25
no-defun-allowed
(ie: if i say "same license as emacs" is it what license was used when i published the software or does it change automagically when emacs changes?)
4:14:38
sabrac
I would interpret it as what the emacs license was when you published, but I am not a copyright lawyer in any jurisdiction.
4:17:35
sabrac
I have seen license language that adds "as amended" or "as amended from time to time", which I would interpret differently.
4:24:06
mfiano
I have a DEFUN over a LET over a DEFUN, so I can initialize and also reset the state of the closure. What can do to have the call-sites to the inner function not emit compiler warnings? Is using funcall on a quoted symbol instead of #' here the only thing I can do, because that sort of feels like a hack, and ?
4:28:34
mfiano
Sure, well the code is just a pretty simple memoization closure: https://gist.github.com/mfiano/79694e884f37779b9d1e31be78f8ea77
4:29:10
mfiano
and call-sites look like (load-texture ...) but will never hit that at runtime until the toplevel initialize function is called
4:29:25
mfiano
This is only a compile time warning because the load-texture function doesn't exist then
4:29:28
loke
beach: I've always been a bit confused about the symbol FMAKUNBOUND. It's 11 characters. What system had a length limit of 11 character so as to make it impossible to use the name FMAKEUNBOUND?
4:31:23
loke
beach: Didn't some system have a limit of 6 characters? ITS Filenames certainly did. Was it originally FMAKUN?
4:33:33
beach
Or `ls', though that was from Multics. But Multics had a long name and a short name for most commands, so `ls' was short for `list'.
4:33:48
mfiano
At first thought I was thinking that'd emit a redefinition warning at runtime when I call the toplevel initialize function. I guess that is not the case
4:35:23
loke
Someone on SO mentions radix50 which packs 6 characters in one word (that's what's used for filenames in ITS). But those use 36-bit words, and Unix was for PDP-11. Also, 6 characters would fit ‘create’. :-)
5:25:12
Fare
sometimes, a hackish name is a good way to signal that the function is a hack. FMAKUNBOUND sounds like it.
8:49:41
beach
flip214: When a chunk is in use, there is no reserved word at the end of it. There is only one word at the beginning indicating the size. The rest is user data. That's why there is an additional bit in the following chunk that indicates whether the preceding one is in use or not.
8:51:41
beach
flip214: Allocating a chunk involves either finding a free one with the right size or splitting a bigger one. There is no need to mention both cases separately. They are both considered to be allocation.
8:54:05
beach
flip214: I am not updating the ARM specification right now because it contains incorrect information, and I will very likely eliminate it entirely and instead use ARM 64.