<void_pointer>artyom-poptsov: that is a major usability improvement (seems like a pain to have to type manually more than once and such a thing shouldn't require doing it manually anyways) <paroneayea>just had the activitypub test suite freeze up while doing a request to something that needs to make a request back to it <paroneayea>(web client) blocks, so 8sync/suspendable ports prevents the thing on the other end from requesting back <paroneayea>wingo: would there be any harm in having open-socket-for-uri in (web client) be nonblocking? <paroneayea>though! I guess I could set it nonblocking myself, if I pass in #:port... <catonano>sneek later tell paroneayea some time ago you pointed to a talk by Gerrald Sussmann about his adaptive computation concept. Could you point me again to it ? I'd lie to suggest it to someone but I can't find it again <MaliRemorker>It's again me and the build systems. It seems that the one safe-ish way to detect where the heck Guile modules reside is to use ${pkgdatadir}/$GUILE_EFFECTIVE_VERSION (after a call to GUILE_PROGS macro in configure.ac). <MaliRemorker>Btw, i just cloned the GNU Shepherd repo, it sets guilemoduledir to ... "${datarootdir}/guile/site/$GUILE_EFFECTIVE_VERSION" which doesn't exist, at least not in my installs of Guile <MaliRemorker>of course, this may not matter if Shepherds creates those subdirs :) <nalaginrut>ArneBab_: good to read (although I dislike a language use integer rather than int) <ArneBab_>nalaginrut: note also the really weak data on this: 22 robust papers in 35 years… <ArneBab_>nalaginrut: I asked for the full results to be able to do sort of an evaluation of Scheme with that. <civodul>CustosLimen: this is Gnulib code imported from the GNU libc <civodul>when building libc, the _LIBC macro is defined <civodul>"when building libc", not "when building with libc" <civodul>essentially it's code that's meant to compiled within glibc, or outside glibc <CustosLimen>see I'm trying to figure out if the memory leak I found is particular to solaris (that does not use gnu libc) - but if _LIBC is never set it should happen on linux also <CustosLimen>or well -it should happen always if RE_ENABLE_I18N is defined <CustosLimen>still present in latest git version also - just used 2.0.11 because that is where we observed the problem <CustosLimen>really dissapointed by the trash alternatives to this on linux ***dsmith-w` is now known as dsmith-work
<nalaginrut>ArneBab: unless they're doing very carefully and slow research, anyway, computer science is a rapid changing area, a 10 years book/paper maybe outdated soon <sneek>paroneayea, you have 1 message. <sneek>paroneayea, catonano says: some time ago you pointed to a talk by Gerrald Sussmann about his adaptive computation concept. Could you point me again to it ? I'd lie to suggest it to someone but I can't find it again <dsmith-work>paroneayea: That message you left for catonano looks failry long. Might be truncated. <dsmith-work>paroneayea: Note that the bot does replay messages in order, so you can split a long message into several messages and they will be delieverd properly. <ArneBab>nalaginrut: the question which follows is: how to become a classic? ☺ <lottin>how do can i execute a procedure step by step? <lottin>when i call the procedure, it triggers the debugger <lottin>but the commands ,s and ,n don't enter the procedure <ijp>,bt doesn't set a breakpoint, it gives a backtrace <ijp>assuming that was a typo, I'm going to need an example <ijp>so what's the procedure like? <lottin>it takes 2 arguments and the body is a named let, but i can't post it here <lottin>Step into #<frame 7ffdd82f5b40 #<procedure process-file-names (lst num-ext)>> <ijp>how do you know it isn't stepping in? <ijp>what doesn't move forward <lottin>i want to execute the instructions in the procedure <ijp>e.g. are you not seeing the frame # change? that could just indicate the frame is being reused due to tail recursion <ijp>are the ,locals not changing? that could be a mistake in your procedure <lottin>no ,locals says no local variables <ijp>like I say, the frame might not change if you have a loop ***michaniskin_ is now known as michaniskin
<Muto>Hello, hello! I'm having some trouble generating a site with Haunt. Actually, I'm having trouble using Haunt in general, I get a "nq code for module (haunt ui)" error. Can anyone help? <Muto>("no code for module", not "nq code") <davexunit>Muto: sounds like your guile load path is not configured properly <Muto>davexunit: I installed Guile 2.0 with the APT command. I'm not really used to using the load-path prefix, how can I configure it, if you don't mind me asking? <davexunit>GUILE_LOAD_PATH and GUILE_LOAD_COMPILED_PATH need to be set <OrangeShark>Muto: how did you install haunt? With `./configure && make && sudo make install`? <Muto>davexunit: I installed Haunt from source with make, yes. <Muto>OrangeShark: Yes that's how I built Haunt. <Muto>davexunit: I installed Haunt in /usr/local/bin/haunt <OrangeShark>by default, that doesn't install it in the normal guile load paths. <Muto>OrangeShark: Ah, I see, so should I just use the "cp" command to copy the haunt file to my module path? <davexunit>Muto: that's the executable, but where are the modules? <davexunit>sounds like you didn't specify an installation prefix <Muto>davexunit: Ah, hold on, I have them here, let me just find the module folder. Also ah I see <Muto>davexunit: Okay I think my module path is: /usr/local/lib/guile/2.0/ccache/haunt <davexunit>export GUILE_LOAD_COMPILED_PATH=/usr/local/lib/guile/2.0/ccache <davexunit>there's your first env var that you need to set <Muto>davexunit: Okay, /usr/local/share/guile/site/2.0/haunt. These look like some source files. <davexunit>export GUILE_LOAD_PATH=/usr/local/share/guile/site/2.0 <Muto>export GUILE_LOAD_PATH=/usr/local/share/guile/site/2.0 <Muto>This is... Not a terminal. <Muto>Alright, I'll try it out, thanks a bunch! <ijp>paroneayea: I *think* my rewrite has been successful <ijp>I was able to reuse some functions for computing dominance in (language cps utils), and that made getting the nesting straightforward <ijp>A side effect of this is that the inliner is no longer necessary for compilation to work <ijp>The code is still pretty fugly, but I can clean it up <ijp>hopefully I can change tack in the next day or so <paroneayea>ijp: happy to see the progress. Look forward to hearing more throughout the week :) ***evhan` is now known as evhan
<ArneBab>ijp: nice! What is the scheme-variable?