Skip to content


Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Browse files


  • Loading branch information...
commit e8aa413fd8298a644d64aa2f5eb558313a34e52a 1 parent e2caea3
Zheng Li authored
Showing with 328 additions and 0 deletions.
  1. +328 −0 2011-11-23-irc.log
328 2011-11-23-irc.log
@@ -0,0 +1,328 @@
+00:07:55 <rgrinberg> I can't seem to install fftw through godi. Here is the error: in case anybody knows what could be wrong.
+00:07:55 <rgrinberg>
+00:07:55 <rgrinberg>
+00:07:55 <rgrinberg>
+00:07:55 <rgrinberg>
+00:07:56 <rgrinberg>
+00:07:56 <rgrinberg>
+00:07:57 <rgrinberg>
+00:07:57 <rgrinberg>
+00:07:58 <rgrinberg>
+00:19:50 <gildor> does someone can open a test feature request here
+00:20:46 <gildor>
+00:21:02 <gildor> (need to test that the restore is ok)
+00:49:56 <thelema> gildor: trying now
+00:50:46 <thelema> gildor: created ID 1060
+02:50:09 <oriba> rgrinberg, fftw-float-type.c:5:18: fatal error: fftw.h: No such file or directory
+02:50:20 <oriba> looks like the header-files are missing
+02:50:50 <oriba> if on Linux (debian/ubuntu...) maybe install the fftw-dev packages
+06:35:55 <superbobry> so, any type-conv users here? can't get it to work -- "... is not a supported type generator"
+06:40:36 <tomprince> superbobry: I don't use type-conv, but you should probably post some more detail, ideally pastebin a small example that demonstrates the issue.
+06:47:37 <superbobry> I wan't to make a minimal working example of type-conv usage. The extension ( compiles successfully, but the file using it ( fails with a weird error --
+06:53:45 <sgnb> superbobry: did you compare with working type-conv stuff? (e.g. ocaml-data-notation)
+06:54:28 <superbobry> yeah, none of them seem to include any examples nor instructions
+07:05:46 <superbobry> funny thing is, the above example actually compiles, when i do it manually with "camlp4 -I +camlp4 -parser o -parser op -printer o pa_type_conv.cma _build/syntax/pa_bson.cmo examples/"
+07:05:58 <superbobry> seems like i'm missing some _tags tricks
+07:08:33 <sgnb> must be built before computing dependencies of
+07:08:53 <sgnb> the error message is understandable
+07:10:47 <superbobry> sgnb: it is built before
+07:12:07 <sgnb> your situation really looks like ocaml-data-notation (with its test suite)... I would have a look there
+08:22:57 <superbobry> jessicah: right, i wonder why is that
+08:23:57 <jessicah> you need to tell ocamlbuild that it's a dependency of
+08:24:41 <superbobry> E: Command '/Users/bobry/.homebrew/bin/ocamlbuild syntax/pa_bson.cma src/cobson.cma src/cobson.cmxa src/cobson.a examples//world.native tests//tests.native -tag debug' terminated with error code 10
+08:24:49 <superbobry> now it is passed, but the error remains the same
+08:24:50 <jessicah> i never seen that wrapper thing around ocamlbuild before
+08:24:58 <superbobry> oasis you mean?
+08:25:03 <jessicah> i guess so
+08:25:10 <jessicah> so i dunno how to help
+08:25:43 <jessicah> can you pass -verbose to see the actual command?
+08:27:32 <superbobry> yup, here it is: 'ocamlfind ocamldep -package type-conv -package camlp4.quotations -package camlp4.lib -package camlp4 -package calendar -package binary -syntax camlp4o -modules examples/ > examples/'
+08:27:46 <jessicah> you probably want smth like ocamlbuild -pp syntax/pa_bson.cma,src/cobson.cma ....
+08:28:21 <jessicah> that tells ocamlbuild that needs to get passed to camlp4; your ocamlfind line above doesn't know about it in there
+08:28:29 <jessicah> so the -pp might fix that
+08:31:01 <jessicah> what does oasis do that ocamlbuld can't do?
+08:31:36 <jessicah> ocamlbuild has its challenges for me without adding another level of complexity into the equation :p
+08:33:05 <superbobry> jessicah: oasis serves a different purpose, it's a replacement for all that autoconf / Makefile mess, usually used in OCaml projects
+08:33:13 <adrien> oasis makes it simpler ;-)
+08:33:37 <jessicah> people use autoconf with ocaml? :o
+08:33:40 <superbobry> and -pp doesn't help, i think, i'll have to wait for gildor to show up :)
+08:33:48 <superbobry> jessicah: unfortunately yes
+08:34:08 <adrien> well, what's the issue with it? ;-)
+08:34:30 <jessicah> with ocamlbuild?
+08:38:56 <jessicah> superbobry: if you have smth i can git clone, i could have a look for you
+08:39:36 <jessicah> i've abused ocamlbuild a bit, so simpler stuff shouldn't be too bad :)
+08:40:20 <superbobry> jessicah: that would be nice! :)
+08:43:09 <jessicah> boo, unbuntu is always so out of date for ocaml stuff
+08:44:32 <jessicah> hmm, it matches version on jane street's site for type-conv
+08:44:56 <jessicah> is it hosted somewhere else?
+08:44:56 <superbobry> you can probably use odb, to install the latest version
+08:45:18 <superbobry> also,
+08:45:56 <jessicah> is that a diff project? or just migrated from jane street?
+08:46:32 <superbobry> migrated, afaik
+08:47:09 <jessicah> odb is like godi?
+08:48:38 <superbobry> i guess, i've never used godi myself
+08:49:28 <jessicah> neither
+08:49:39 <jessicah> where do i get odb from?
+08:53:19 <superbobry> jessicah:
+08:58:11 <jessicah> odb isn't working for me
+08:58:38 <jessicah> ocaml was easier when i could just get a tarball & make, sudo make install =/
+08:58:52 <jessicah> if it wasn't in apt
+08:59:08 <superbobry> well, everything is easier with make install :) that's probably the only good thing about it
+08:59:12 <superbobry> what isn't working?
+08:59:20 <superbobry> have you followed the instructions in the readme?
+08:59:24 <adrien> except when it doesn't install properly ;-)
+08:59:32 <superbobry> right :)
+08:59:36 <jessicah> i have for odb
+08:59:45 <jessicah> some its packages don't download
+08:59:51 <superbobry> okay, which ones?
+08:59:56 <jessicah> curl complains about smth wrong with -o
+08:59:59 <adrien> jessicah: odb is a bit young but should beat configure, make, make install in every regard quite soon
+09:00:02 <adrien> jessicah: which OS?
+09:00:05 <jessicah> bin_prot.syntax
+09:00:07 <jessicah> unbuntu
+09:00:15 <superbobry> hrm
+09:00:37 <jessicah> jessica@AirUnbuntu:~$ ocaml bin_prot.syntax odn.with.syntax odn.without.syntax --force-all
+09:00:46 <jessicah> curl: option -o: requires parameter
+09:00:54 <jessicah> Failure "Curl failed to get".
+09:00:56 <superbobry> ah! just install bin-prot
+09:01:42 <jessicah> hmm, seems i already have
+09:01:46 <superbobry> odb --repo unstable bin_prot this should work
+09:02:06 <jessicah> i'd rather not use odb
+09:02:22 <superbobry> odb has a concept of repos, with "stable" being default -- but 'bin_prot' is not available in it
+09:03:24 <jessicah> what library is 'binary'?
+09:03:34 <jessicah> and are all these deps on ocamlforge?
+09:03:44 <superbobry> yup, they are
+09:03:45 <jessicah> ocamlcore*
+09:04:05 <jessicah> i'll try build from source
+09:04:10 <superbobry> except for binary -- :)
+09:10:39 * jessicah sighs
+09:12:36 <jessicah> kaputt is another example of a project with broken make install target
+09:12:38 <jessicah> =/
+09:13:03 <superbobry> you can comment out test target in _oasis
+09:13:12 <superbobry> kaputt is only needed for tests
+09:13:40 <superbobry> and yeah, it's another example on "why you should try oasis" :)
+09:14:15 <jessicah> it's an example of use `ocamlc -where` to install your stuff
+09:14:21 <jessicah> people overcomplicate things
+09:14:58 <superbobry> if i recall correctly, kaputt issue is different
+09:15:05 <superbobry> it doesn't install kaputt.cmi
+09:15:37 <jessicah> no, kaputt doesn't install in a location where my system ocamlfind knows about
+09:16:01 <superbobry> it has install-findlib target i think
+09:16:54 <jessicah> things in ocaml-land have gotten way out of hand these days
+09:16:59 <jessicah> seriously =/
+09:17:20 <superbobry> :)
+09:17:31 <jessicah> i never had issues like this in the past
+09:17:47 <superbobry> broken makefiles?
+09:18:03 <jessicah> people simply doing stuff weirdly
+09:19:08 <jessicah> inventing systems to fix problems that are b/c of people, not the tools
+09:19:24 <jessicah> causing more problems
+09:20:12 <superbobry> not really, building things in OCaml world is just too comples, all that ocaml{dep,build,c,...} are a bit confusing, esp. for the newcomers
+09:20:20 <superbobry> for me oasis worked perfectly
+09:20:33 <superbobry> until the moment i tried writing a syntax extension :D
+09:20:59 <superbobry> s/comples/complicated/
+09:20:59 <jessicah> ocamlbuild is still somewhat new, but it's not really that hard to use :|
+09:21:14 <gildor> superbobry: what the problem with oasis ?
+09:21:57 <jessicah> my only problem with ocamlbuild was having a .cmx generated with invalid crc
+09:22:15 <jessicah> and only because i have special issues :)
+09:22:16 <superbobry> gildor: i have a syntax extension in my project, and an example using it -- for some reason example fails too build, and the error indicates, that my syntax extension wasn't loaded:
+09:22:39 <gildor> jessicah: oasis use ocamlbuild, so it can do everything ocamlbuild does
+09:23:13 <jessicah> yeah i get that
+09:23:15 <gildor> jessicah: this is not another layer of complexity, but a human-readable description of your project (depends, source repository, META)
+09:24:21 <jessicah> how is it different to/better than GODI and ocamlbuild/findlib?
+09:24:47 <gildor> GODI is a traditional packaging system
+09:25:13 <gildor> jessicah: if you want a package for GODI you have to create a separate set of files that are only GODI related
+09:25:32 <jessicah> if a project comes with oasis support, do i have to use it to build it?
+09:25:33 <gildor> jessicah: whereas _oasis file is made to be useful for dev and for GODI people
+09:25:35 <jessicah> or is it optional?
+09:25:57 <gildor> i.e. you use _oasis when developping your library/software and people at GODI use it for packaging
+09:26:27 <gildor> jessicah: no oasis generates a self contained that does everything
+09:26:39 <gildor> (a little bit like autoconf/configure)
+09:27:16 <gildor> jessicah: as an example, there is oasis2debian that convert an _oasis file into debian/* files
+09:28:27 <gildor> superbobry: using a generated syntax extension need some works, see;a=headblob;f=/tests/
+09:30:32 <gildor> superbobry: maybe adding something like Byte/NativeOpt: --ppopt _build/syntax/pa_bson.cma can do
+09:30:41 <gildor> to Executable hello
+09:32:07 <gildor> superbobry: and don't forget to add "examples/": syntax_camlp4o to _tags
+09:33:13 * gildor @work
+09:45:18 <superbobry> hm, this seems to work, when executed manuall (as in odn tests)
+09:45:54 <superbobry> but NativeOpts (ByteOpts) aren't being passed to ocamlbuild for some reason
+09:46:31 <superbobry> making 'ocamlfind ocamldep ...' fail
+09:57:50 <jessicah> how do i get ocamlbuild to include a directory, but not compile files in it?
+09:59:56 <jessicah> i currently use -ignore, but the line gets pretty long....
+10:14:13 <avsm> jessicah: you mean to pick up .cmx/.cmi files, but not part of the dependency checking?
+10:21:05 <jessicah> avsm: yes
+10:22:41 <avsm> jessicah: hm, i actually have rules to move all the .cmx files into a stdlib directory, and compile programs against that (with -nostdlib)
+10:23:10 <avsm> makes life a lot easier, especially since its not safe to build a pack .cmx as the submodules will be picked up if you add that directory to -I for someone else
+10:25:39 <jessicah> hmm
+10:26:06 <jessicah> actually, the tags I have from my myocamlbuild actually work
+10:26:11 <jessicah> just never tried them before
+10:26:14 <jessicah> minus one folder
+10:26:42 <avsm> i guess you can add a flag ["ocaml"; "build"] to add the -I
+10:26:48 <avsm> which should skip ocamldep
+10:31:47 <jessicah> hmm
+10:35:35 <jessicah> oh, i found an easy way
+10:35:49 <jessicah> just added a flag rule to my snowflake_lib function
+10:41:34 <jessicah> flag ["ocaml";"compile";"plugin"] to add the -I worked
+10:41:40 <jessicah> then just -tag plugin to ocamlbuild
+10:41:51 <jessicah> thanks for inspiration :)
+10:42:24 <jessicah> can't believe how easy that was -_-
+11:18:31 <gildor> superbobry: submit a bug if they are not passed, this is not normal
+11:20:23 <gildor> superbobry: bug + tarball of your problem
+11:21:06 <superbobry> sure, but i think i'll do some more checking first
+11:54:43 <pierrc> Could someone tell how to load I've just installed batteries.
+11:55:10 <pierrc> I would like to have a taste of it. I don't where to begin.
+11:59:45 <superbobry> well, the easiest way would be to load it with finlib
+11:59:48 <superbobry> #use "topfind";;
+11:59:52 <superbobry> #require "batteries";;
+12:05:43 <pierrc> Yes, batteries is loaded, thank you! :)
+12:10:38 <f[x]> using global NativeOpt to pass ppopt is very fragile
+12:11:03 <f[x]> syntax manipulation should be per-file precise
+14:57:14 <Drakken> What does the pattern "_ as x" mean?
+14:58:50 <thomasga> Drakken: in this case it means "x"
+14:59:06 <thomasga> but you can use _ deep inside a pattern
+14:59:18 <thomasga> and x will be the name of the whole pattern
+14:59:32 <thomasga> as in "(_,true) as x"
+15:00:54 <Drakken> thomasga thank you but I'm not asking about a specific case. I'm just wondering why anyone would write "_ as x" instead of just "x".
+15:01:20 <thomasga> (or "Foo _ as x")
+15:01:36 <thomasga> well I don't see any reason why you would want to write that
+15:01:43 <Drakken> Neither do I.
+15:02:12 <f[x]> (a,b,_) as x
+15:02:19 <f[x]> (a,b,_ as x)
+15:02:44 <Drakken> (a,b,x)
+15:02:50 <f[x]> not
+15:03:02 <Drakken> What's the difference?
+15:03:17 <f[x]> priority
+15:03:20 <f[x]> read the manual
+15:03:31 <f[x]> or test is manual
+15:04:00 <rixed> a,b,_ as x -> x is (a,b,_) :-)
+15:04:41 <f[x]> * or test in toplevel
+15:06:02 <Drakken> This is why Lispers love their parens so much.
+15:06:09 <thomasga> well this case is weird
+15:06:16 <thomasga> and confusing
+15:14:09 <Drakken> agreed
+15:14:59 <Drakken> f[x] thank you
+15:21:51 <thelema> thomasga: I occasionally use (Foo _ as x)
+15:26:17 <reynir> Hello
+15:27:07 <thelema> reynir: hi
+15:28:01 <reynir> I'm supposed to truncate an int32 to a byte, but still have it as an Int32. That is, make 128l into -1l, for example
+15:28:23 <reynir> Any ideas?
+15:28:23 <thelema> an unsigned byte... hmmm
+15:28:35 <thelema> signed is easy - just and with 0xffl
+15:28:36 <reynir> no, it's supposed to be a signed byte
+15:28:45 <thelema> err, unsigned is easy
+15:29:00 <thelema> for signed, you'll have to extend the hgh bit
+15:29:04 <reynir> Yea
+15:29:28 <reynir> Do you know how to do that?
+15:30:54 <thelema> open Int32 let to_byte x = if x log_and 0x80 then x log_or 0xffffff00l else x log_and 0xffl
+15:32:38 <thelema> oops, no _ after log
+15:35:07 <reynir> Thanks thelema, that worked!
+16:01:56 <reynir> Thanks for your help, thelema :-)
+16:02:02 <thelema> you're welcome
+16:13:58 <thelema> I wonder if I'm having problems with this:
+16:14:11 <thelema> neither ocaml-buddy nor cairo2 will compile for me because of undefined references
+16:15:05 <thelema> despite me being quite sure of having the right libraries
+16:26:17 <thelema> hmm, I wonder if ocaml is calling gcc wrong...
+16:26:34 <thelema> gcc -o 'test.byte' '-L/home/thelema/.odb/lib/oUnit' '-L/usr/local/lib/ocaml' -L/usr/local/lib '/tmp/camlprim0f3ebf.c' '-lunix' '-lbdd' 'libbuddy_stubs.a' '-lcamlrun' -I'/usr/local/lib/ocaml' -lm -ldl -lcurses -lpthread
+16:26:46 <adrien> the undef refs being?
+16:26:47 <thelema> I need to move the '-lbdd' after the 'libbuddy_stubs.a'
+16:26:54 <thelema> libbuddy_stubs.c:(.text+0x22f): undefined reference to `bdd_nodecount'
+16:27:00 <thelema> plus many more
+16:27:27 <adrien> yup, -lbdd should be right to -libbuddy_stubs
+16:28:55 <thelema> hmm, -ccopt options go before file arguments
+16:29:26 <f[x]> thelema, you may be interested in
+16:30:08 <f[x]> the behaviour changed in 3.12, and I am not sure that for better
+16:30:34 <adrien> thelema: but I think the order is stored during ocamlmklib's invocation
+16:31:22 <thelema> f[x]: it seems to have been made worse by upgrading to ubuntu 11.10 ; I could compile cairo2 before the upgrade, but can't now.
+16:31:44 <f[x]> that's probably because of ld/binutils
+16:31:49 <f[x]> or the combination of these two changes
+16:31:55 <thelema> f[x]: yes, that's what the gentoo page seems to indicate
+16:31:59 <adrien> pass -cclib --no-as-needed
+16:31:59 <f[x]> I've seen builds breaking because of this
+16:32:33 <thelema> hmm, maybe not - gcc: error: unrecognized option ‘--no-as-needed’
+16:32:56 <thelema> hmm, maybe need -Wl
+16:33:42 <thelema> -ccopt "-Wl,--no-as-needed" - this isn't working either
+16:35:00 <adrien> it won't use ccopt's argument at link time
+16:35:21 <adrien> I think it uses ld and not gcc to do the linking
+16:36:35 <thelema> my error is coming from a gcc invocation
+16:37:13 <thelema> same error without the -ccopt "-Wl,--no-as-needed"
+16:41:09 <adrien> does the .c file from ocaml's gcc invocation still exist?
+16:45:19 <thelema> adrien: no, it's deleted
+16:49:12 <f[x]> gdb ; b unlink ; then look it up in /tmp :)
+16:49:24 <f[x]> not sure how it will help here though
+16:50:57 <thelema>
+16:54:00 <thelema> hey, using `-cclib -lbdd` instead of `-ccopt -lbdd` seems to have worked
+16:55:33 <thelema> except it needs to be at the end of the command, not in the middle, where ocamlbuild puts it
+16:56:47 <thelema>
+17:11:45 <reynir> Is it possible to use ocaml for scripting?
+17:14:28 <hcarty> reynir: Yes, although the ease depends on the scripting you are trying to do
+17:15:27 <hcarty> reynir: See odb as an example, or you can use ocamlscript for automatically compiled scripts/programs
+17:16:16 <hcarty> reynir: Combined with the appropriate libraries it can be very useful.
+17:17:43 <reynir> hcarty: thanks, that sounds interesting :)
+18:14:41 <hcarty> thelema: I have a few list splitting/grouping functions which may be useful -
+18:15:19 <hcarty> thelema: I think that is close enough to break that they are effectively the same
+18:15:42 <hcarty> thelema: I don't know if [group ~by l] would be general enough to include in Batteries
+18:26:58 <adrien> does this patch come from darcs?
+18:47:37 <gildor> adrien: it is extracted from JaneStreet version of oasis
+18:48:38 <adrien> I've applied it manually but I'm getting a build error I haven't been able to understand yet
+18:49:14 <adrien> (rather: I understand the error but don't understand why it wasn't failing before :P )
+18:52:49 <hcarty> Is it possible to tell oasis to install .cmx files?
+18:53:46 <adrien> I think I've failed to apply the patch properly (it has no context lines)
+19:02:28 <gildor> hcarty: as far as I know it should be automatically installed
+19:02:42 <gildor> adrien: darcs apply should
+19:02:50 <gildor> adrien: "darcs apply" should do
+19:04:53 <adrien> ghc can't be bootstrapped and is annoying to get on slackware (and I had to reinstall recently)
+19:05:17 <adrien> anyway, the patch won't work on 0.2.1, should be a small change from 0.2.0
+19:06:00 <gildor> hcarty: err, I plan to do but have only worked on a proto a long time ago
+19:06:04 <gildor> hcarty:;a=headblob;f=/src/tools/
+19:06:21 <hcarty> gildor: Ah, ok
+19:08:03 <hcarty> gildor: Is there a way to manually indicate which .cmx files should be installed?
+19:09:53 <ijp> Hello, is there a less verbose way to do (function x -> x#method) ls;; ?
+19:10:10 <ijp> I was hoping I could just use (#method) or #method
+19:10:17 <hcarty> ijp: (fun x -> x#method) ls
+19:10:33 <hcarty> #method means something else I think...
+19:10:45 <hcarty> I may be wrong about that.
+19:10:46 <ijp> I guess that's better than nothing
+19:11:17 <hcarty> There may be a syntax extension somewhere that provides shorthand syntax like that
+19:15:55 <gildor> hcarty: not sure
+19:16:20 <gildor> hcarty: you'll have to patch to do that
+19:16:49 <hcarty> gildor: Ok thanks, I'll take a look.
+19:18:17 <gildor> hcarty: I recommend you to have a look at how it is done for .cmi
+19:18:45 <gildor> hcarty: and complement it with InternalModules
+19:18:51 <gildor> and send me the patch
+19:18:53 <gildor> ;-)
+19:19:36 <hcarty> gildor: :-)
+19:20:16 <gildor> hcarty: stop
+19:20:20 <gildor> hcarty:
+19:23:47 <hcarty> gildor: Do you expect that patch to go into the next release?
+19:27:01 <gildor> yep in 0.2.1 release
+19:42:28 <tomprince> I am having issues using oasis to compile a module with C stubs:
+19:44:12 <tomprince> I have managed to succesfully build by hand with: ocamlopt -ccopt '-I/usr/include/python2.7 -fPIC' -cclib '-lpython2.7' -a -o pycaml.cmxa pycaml.mli pycaml_stubs.c
+20:08:40 <adrien> gildor: the patch neds some slight changes before it can apply to 0.2.1, I'll put my updated patch whn when I get on a network that has less than 30 minutes of lag
+20:57:22 <adrien> tomprince: I don't remember well but do you have .clib files?
+21:52:44 <adrien> would it be possible that Dynlink returned a first-class module?
+21:56:49 <tomprince> adrien: oasis is geenrating an empty clib file
+21:57:20 <gildor> tomprince: yeah, std error with c binding
+21:57:57 <gildor> replace Library pycaml by Library pyocaml, e.g.
+21:59:24 <gildor> wait, maybe it is not what I think
+21:59:33 <gildor> could you give me the content of _build
+21:59:37 <gildor> (ls _build)
+21:59:41 <gildor> tomprince: ^^^
+22:01:45 <tomprince> gildor:
+22:01:46 <gildor> tomprince: and it seems that there is a line missing in your build log
+22:02:06 <gildor> tomprince: it should build the pycaml_stubs.o at the beginning
+22:02:32 <tomprince> _build/_log:
+22:04:37 <gildor> could you locate a generated .clib file ?
+22:04:59 <gildor> (you can run oasis setup to generate it)
+22:05:45 <tomprince> I have one in both . and _build, and the just contain comment (#) lines
+22:05:53 <gildor> strange
+22:06:17 <gildor> tomprince: can I have access to your project ?
+22:06:52 <tomprince> with the oasis file I posted.
+22:13:04 <gildor> tomprince: ok, got it
+22:13:20 <gildor> tomprince: put a comma between pycaml_stubs.c and pycaml.h
+22:14:17 <gildor> tomprince: and open a feature request to output a warning when 2 files separated by space exists whereas one file with space in it doesn't exist
+22:15:30 <gildor> e.g. W: file "pycaml_stubs.c pycaml.h" doesn't exist but "pycaml_stubs.c" and "pycaml.h" exists, maybe you should write "pycaml_stubs.c, pycaml.h"
+22:18:27 <tomprince> gildor: thanks
+22:18:58 <adrien> heheh :P
+22:19:19 <adrien> and maybe more generally warn about spaces in filenames?
+22:19:59 <gildor> if you have a good scheme to warn about them, please comment to feature request that tomprince will open
+22:20:07 <gildor> adrien: ^^^
+22:20:18 <gildor> adrien: this is the kind of stuff that can go into 0.2.1
+22:43:26 <tomprince>
Please sign in to comment.
Something went wrong with that request. Please try again.