Java linux HTML qmail C++ PHP Debian MySQL ASP JavaScript perl Delphi NetBSD Solaris etc etc etc [...]






Page: 1 2 3 4 5 6 7


Comments:
<cods> nyef: yes, sorry
<cods> Try to start clisp with -E utf-8
<cods> or.. iso-8859-1 or whatever, matching the right encoding.
<waterpie> cods: =8-) i do have a file that shows in strange chars, but it didn't seem to bother it earlier. i tried loading clisp in a diff dir, but still the problem. (btw, why would it care about the files in the cwd? )
<cods> waterpie: I don't remember exactly the detail. Yes, it would be weird if it was scanning current working directory.
<waterpie> cods: it still seg faults with -E utf-8
<cods> but I had this sort of problem in the past, and I remember it was because of filename.
<cods> waterpie: does -E iso-8859-1 also segfault ?
<waterpie> cods: this works



<cods> waterpie: you could try to strace clisp when it segfault to check where it fails exactly
<waterpie> cods: a) thanks :-) b) wtf?! :-) c) (obviously i make a mistake) but I didn't change anything since noon! what could have happened?? d) yes, i'll try strace
<cods> waterpie: about c), it's hard to tell. One thing changed though: the time. But I don't think it's related ;)
<waterpie> cods: hmm. strace prints many pages. i suppose the last lines are the importand?
<cods> waterpie: right. You can filter output with -e followed by comma-separated system call name,
<cods> and you can also use -o to redirect output to a file.
<cods> Also, -f may be usefull to follow each fork/thread
<waterpie> cods: hmm. this is way over my head, but (scanning for open) last msg is: "open("/usr/lib/gconv/ISO8859-1.so", O_RDONLY) = 5"
<waterpie> also without scanning, last msg is :"read(0, "
<cods> waterpie: this file belong to the glibc. You compiled clisp yourself ?
<waterpie> no
<cods> waterpie: what's your linux distrubtion ? (Assuming it's linux..)
<waterpie> also last 2 msgs scanning for read: "read(5, "# /etc/inputrc - global inputrc "..., 1698) = 1698" and "read(0,"
<waterpie> latest kubuntu
<Pupeno> I created a 'system' with asdf, I linked the asd file from ~/.sbcl/systems, is that they way to do it ?
<cods> waterpie: Check if Kubuntu's bug tracking have something related to your problem.
<cods> Pupeno: usually yes
<Pupeno> When loading it I've got the error: The file "/home/pupeno/.sbcl/systems/packages.lisp" does not exist.
<cods> Pupeno: does the symbolic link ~/.sbcl/systems/packages.lisp point to the correct place ?
<nyef> No, no... Is it a symlink or a hardlink? (should be a symlink.)
<waterpie> cods: it dosn't. it has only one bug (that it can't load cl packages)
<Pupeno> nyef: it was a hard link, thanks. Solved.
<smokecfh> *why* are there so many gtk+ lisp bindings?
<cods> waterpie: ok. Well, paste the 20 last lines of the strace output on lisppaste, just to check if there is really nothing obvious.
<lisppaste> waterpie pasted "encoding problem seg faults clisp" at http://paste.lisp.org/display/24034
<cods> waterpie: but here it is not segfaulting, right ?
<cods> and the read(0, mean that clisp try to read from standard input (from the prompt most probably here)
<slyrus> I don't suppose any of you happened to lose a white rabbit in the berkeley area?
<waterpie> this is the output of "strace clisp" if i run "clisp", it enters, my only option is to abort, i abort, and says it seg faults, with: "- handle_fault error2 ! address = 0xb79a1384 not in [0x67ffdc10,0x680a8000) !"...
<pjb> waterpie: you could try to find a clisp -x something that segfaults, then strace clisp -x something
<waterpie> i try "clisp -x "(+ 1 2)"" and it says again: "- invalid byte sequence #xC3 #x2E in CHARSET:UTF-8 conversion" then it says "Bye" and exits without saying it segfaults
<pjb> waterpie: ok, so you have a mismatch between the encoding used on the terminal or files and the encoding of the terminal or file (like ~/.clisprc).
<pjb> What encoding do you use on the terminal and in your files?
<waterpie> btw, is there a way to see to what characters #xC3 #x2E corespond to? this may give a clue
<pjb> You can specify it with -E



<pjb> It really depends on the encoding...
<waterpie> pjb: cods told me about -E. if i use iso-1 it's ok. but few hours ago, i didn't have to do that
<cods> waterpie: but, as pjb said, find something which cause clisp to segfault, then strace it.
<pjb> AFAIK, the only file that's read when you run clisp -x "(+ 1 2)" is ~/.clisprc (or ~/.clisprc.lisp). Check the encoding of thesefiles.
<IPmonger> 0
<waterpie> pjb: i don't have those files
<slyrus> so what the heck am I supposed to do with this damn bunny now?
<nyef> slyrus: Rabbit stew?
<slyrus> the thought crossed my mind. I don't think the kids would forgive me however.
<slyrus> I suppose feeding it some of the rabbit pate that's in the fridge is not a good idea
<waterpie> i don't think i can segfault it with "-x" it seems it doesn't evaluate what i type (i tried garbage). it just aborts automatically
<nyef> They probably wouldn't forgive you for explaining about where eggs, ham, or bacon come from either.
<slyrus> well, they haven't met the pig that provided the bacon, unlike the bunny
<slyrus> and eggs, well, they're not nearly as cute as a rabbit
<nyef> The little baby ones are!
<slyrus> you need to scramble a lot of those little robin's eggs to make a good breakfast, I'd imagine
<pipeline> robin's eggs taste awful, too.
<nyef> Ostritch?
<pipeline> Most eggs taste awful. Chickens are a rare exception to a general rule.
<pipeline> nyef: IIRC, ostrich eggs smell like something died and rotted in the vicinity when you cook them.
<pipeline> nyef: great novelty item, no so great for actual eatin'.
<pipeline> nyef: or is that the emu egg? Some really damn big bird egg.
<Pupeno> damn, I hate errors on slime/sbcl... there's no indication on where they are happening.
<nyef> pupeno: How do you mean?
<Pupeno> paste?
<Pupeno> lisppaste: url?
<lisppaste> To use the lisppaste bot, visit http://paste.lisp.org/new/lisp and enter your paste.
<lisppaste> Pupeno pasted "Something wrong in models ? where ?" at http://paste.lisp.org/display/24036
<Pupeno> nyef: that, I don't understand where to start looking for the error.
<nyef> You're not using M-x slime-load-system?
<Pupeno> nyef: no, I didn't know about it.
<nyef> FWIW, I -still- haven't gotten around to reading the slime manual to find out what sort of goodies it has that I'm unaware of.
<pjb> waterpie: clisp loads ~/.clisprc.lisp before executing -x arguments.
<pjb> If you don't have it, then I'm at loss...
<waterpie> pjb: i dont' have it. but i can touch it :-) is there a command to set the encoding in .clisprc ?
<pjb> waterpie: perhaps strace clisp | grep open and see what file it opens (it opens a lot of libraries too).
<slyrus> dtrace in osx 10.5 sounds like it will be useful
<Pupeno> I could use the printer at the office... but that'd be a bit of abusing.
<rydis> Everyone who has looked into dtrace seems to love it, so I guess so.


Name:

Comments:

Please enter the result of the sum 63 + 46 (to avoid spam):






Return to #lisp
or
Go to some related logs:

gentoo 4x6.bdf command not found
totem load-modules
web
php
mysql
compile php GD no jpeg
UniChrome Pro IGP xorg ubuntu
what is report_zend_debug
linux
perl