[00:06:21] *** mdel has quit IRC
[00:08:25] <CIA-37> macsforever2000 at macports dot org * r89630 /trunk/dports/python/ (4 files in 4 dirs):
[00:08:25] <CIA-37> http://trac.macports.org/changeset/89630
[00:08:25] <CIA-37> py-parsing: unify
[00:22:24] *** dcolish has quit IRC
[00:35:47] <CIA-37> mww at macports dot org * r89631 /trunk/dports/lang/gcc47/Portfile:
[00:35:47] <CIA-37> http://trac.macports.org/changeset/89631
[00:35:47] <CIA-37> gcc47: version 4.7-20120204
[00:40:26] *** khindenburg_ has joined #macports
[00:43:39] *** julian9_ has joined #macports
[00:44:09] *** ndrs___ has joined #macports
[00:44:13] *** plavcik_ has joined #macports
[00:44:35] *** elbjorn_ has joined #macports
[00:45:08] *** catmistake has quit IRC
[00:45:08] *** kurthindenburg has quit IRC
[00:45:08] *** elbjorn has quit IRC
[00:45:09] *** Intensity has quit IRC
[00:45:09] *** ndrs__ has quit IRC
[00:45:10] *** \textit{Ionic} has quit IRC
[00:45:10] *** plavcik has quit IRC
[00:45:10] *** julian9 has quit IRC
[00:45:11] *** catmistake has joined #macports
[00:45:53] <CIA-37> and.damore at macports dot org * r89632 /trunk/dports/math/tiemu3/ (Portfile files/patch-gtk_disable_deprecated.diff):
[00:45:53] <CIA-37> http://trac.macports.org/changeset/89632
[00:45:53] <CIA-37> port tiemu3: fixing gtk deprecated error as per #30665, new patch file from ubuntu package, closing ticket
[00:47:12] *** Guest4862 has joined #macports
[00:47:58] <CIA-37> and.damore at macports dot org * r89633 /trunk/dports/math/tiemu3/Portfile:
[00:47:58] <CIA-37> http://trac.macports.org/changeset/89633
[00:47:58] <CIA-37> port tiemu3: changed descriptions
[00:48:57] <CIA-37> and.damore at macports dot org * r89634 /trunk/dports/math/tiemu3/Portfile:
[00:48:57] <CIA-37> http://trac.macports.org/changeset/89634
[00:48:57] <CIA-37> port tiemu3: whitespace changes
[00:52:50] *** anddam has quit IRC
[00:55:11] *** JaRoel|4d has joined #macports
[00:56:29] *** nedkelly has quit IRC
[01:01:08] *** winut has quit IRC
[01:06:45] *** winut has joined #macports
[01:12:29] *** Intensity has joined #macports
[01:14:29] *** TSK has quit IRC
[01:21:07] *** macrover has joined #macports
[01:21:15] *** Guest4862 has quit IRC
[01:22:41] *** Guest98758 has joined #macports
[01:22:57] <CIA-37> mmoll at macports dot org * r89635 /trunk/dports/math/slepc/Portfile:
[01:22:57] <CIA-37> http://trac.macports.org/changeset/89635
[01:22:57] <CIA-37> math/slepc: fix typo
[01:24:21] *** UnconventionalT has quit IRC
[01:26:27] *** Guest98758 is now known as Ionic
[01:57:29] *** macrover has quit IRC
[02:17:30] *** mdel has joined #macports
[02:40:54] <CIA-37> ryandesign at macports dot org * r89636 /trunk/dports/devel/git-core/ (Portfile files/patch-Makefile.diff):
[02:40:54] <CIA-37> http://trac.macports.org/changeset/89636
[02:40:54] <CIA-37> git-core: avoid deleting /dev/null during the build (!); see #33127
[03:10:08] <Bart-> hi guys
[03:10:20] <Bart-> i have some strange problems installing bwm-ng
[03:12:06] <Bart-> http://pastebin.com/GEmuWGhC
[03:12:08] <Bart-> it says
[03:12:26] <Bart-> info:build ld: symbol(s) not found for architecture x86_64
[03:12:26] <Bart-> :info:build clang: error: linker command failed with exit code 1 (use -v to see invocation)
[03:12:33] <Bart-> so that means its not for 64bit
[03:12:55] <Bart-> anything i could do?
[03:14:27] *** ndrs___ has quit IRC
[03:18:28] <neverpanic> this looks like the good old code-assumes-gnu89-inline-behavior problem
[03:18:56] <neverpanic> try port clean bwm-ng && port install bwm-ng configure.compiler=llvm-gcc-4.2 and if it succeeds, file a bug
[03:25:14] <Bart-> ah thanks neverpanic will try it out now
[03:25:19] <Bart-> i will file a bug then
[03:25:40] <Bart-> but i think will go on tomorrow i am so fscking tired
[03:25:46] <Bart-> my eyes falling down all the time ;/
[03:26:07] <Bart-> hehe i mean they close not falling down ;D
[03:31:37] *** aber has quit IRC
[03:47:09] *** lipsinV2 has joined #macports
[04:10:48] <CIA-37> nicos at macports dot org * r89637 /trunk/dports/kde/konsole/Portfile:
[04:10:48] <CIA-37> http://trac.macports.org/changeset/89637
[04:10:48] <CIA-37> konsole: adding missing dependency (ticket #33137)
[04:14:02] *** ovnicraft has joined #macports
[04:15:28] <ovnicraft> hello after install python and pygtk i run python from /opt/local/bin/python2.6 then try to use pygtk but tell there is not a module, macport set python path for new python ?
[04:26:14] *** robink has quit IRC
[04:37:07] *** jkyle has quit IRC
[04:45:20] <CIA-37> mmoll at macports dot org * r89638 /trunk/dports/games/xmj/Portfile:
[04:45:21] <CIA-37> http://trac.macports.org/changeset/89638
[04:45:21] <CIA-37> games/xmj: enable universal variant
[05:29:03] *** robink has joined #macports
[05:29:04] *** robink has joined #macports
[05:57:29] *** Beetle has joined #macports
[05:57:54] *** Beetle has left #macports
[06:06:19] *** winut_ has joined #macports
[06:08:13] *** winut has quit IRC
[06:08:13] *** winut_ is now known as winut
[06:20:39] *** winut has quit IRC
[06:47:09] <CIA-37> ryandesign at macports dot org * r89639 /trunk/dports/python/py-parsing/Portfile:
[06:47:09] <CIA-37> http://trac.macports.org/changeset/89639
[06:47:09] <CIA-37> py-parsing: rewrite master_sites to avoid redirects
[06:48:34] <CIA-37> ryandesign at macports dot org * r89640 /trunk/dports/python/py-parsing/Portfile:
[06:48:34] <CIA-37> http://trac.macports.org/changeset/89640
[06:48:34] <CIA-37> py-parsing: disable livecheck in subports
[06:50:19] <CIA-37> ryandesign at macports dot org * r89641 /trunk/dports/python/py-parsing/Portfile:
[06:50:19] <CIA-37> http://trac.macports.org/changeset/89641
[06:50:19] <CIA-37> py-parsing: adjust whitespace
[06:58:24] <CIA-37> ryandesign at macports dot org * r89642 /trunk/dports/python/py-parsing/Portfile:
[06:58:24] <CIA-37> http://trac.macports.org/changeset/89642
[06:58:24] <CIA-37> py-parsing: create docdir before trying to install things into it (#33138), and use $subport instead of $name so the subports don't conflict with one another
[07:03:49] *** macrover has joined #macports
[07:14:09] *** bgupta has quit IRC
[07:46:23] *** bgupta has joined #macports
[07:50:06] *** macrover has quit IRC
[09:03:40] *** mpbot_ has quit IRC
[09:03:59] *** mpbot has joined #macports
[10:57:11] *** elbjorn_ is now known as elbjorn
[10:58:23] *** drw has quit IRC
[10:59:01] *** drw has joined #macports
[11:27:07] <CIA-37> stromnov at macports dot org * r89643 /trunk/dports/python/py-blosc/ (. Portfile):
[11:27:07] <CIA-37> http://trac.macports.org/changeset/89643
[11:27:07] <CIA-37> py-blosc: new port
[11:35:30] *** Bemmu has quit IRC
[11:37:09] <CIA-37> stromnov at macports dot org * r89644 /trunk/dports/python/py-psutil/ (. Portfile):
[11:37:09] <CIA-37> http://trac.macports.org/changeset/89644
[11:37:09] <CIA-37> py-psutil: new port
[11:55:49] *** TheInfinity has joined #macports
[12:01:24] *** drw has quit IRC
[12:01:57] *** drw has joined #macports
[12:31:50] *** TheInfinity has quit IRC
[12:33:29] *** anddam has joined #macports
[12:33:30] <mpbot> anddam is and.damore and maintainer of zmq20, lua-luaexpat, py-mercurial_keyring, and repo (of 48 total)
[12:33:43] <anddam> hi
[12:45:38] *** GlobalOD has joined #macports
[12:46:47] <GlobalOD> hi
[12:47:42] <GlobalOD> anyone up this early?
[12:50:36] <GlobalOD> ...
[12:50:52] <anddam> GlobalOD: where are you from?
[12:50:57] <anddam> it's lunchtime in EU
[12:51:11] <GlobalOD> i better spell that out then
[12:51:15] <GlobalOD> kansas
[12:51:35] <anddam> Dorothy's home
[12:51:41] <GlobalOD> you got it
[12:51:46] <anddam> I read the book
[12:52:02] <GlobalOD> yes better than the movie
[12:52:08] <GlobalOD> and the movie's not bad
[12:52:17] <GlobalOD> anyway
[12:52:33] <GlobalOD> I don't know if you or anyone else could help me get macports setup
[12:52:55] <GlobalOD> the installer ran just fine but it keeps throwing up errors w/rsync
[12:53:17] <GlobalOD> well and not even errors really
[12:53:38] *** UnconventionalT has joined #macports
[12:53:40] <GlobalOD> it just can't connect to the server
[12:55:11] <GlobalOD> macports' install page mentioned a port 873
[12:55:29] *** UnconventionalT has quit IRC
[12:55:50] <GlobalOD> my os is wide open
[12:56:27] <GlobalOD> and even though i haven't had to do this in the past with macports, i opened the port on the router
[13:05:26] *** aber has joined #macports
[13:14:52] *** aber_ has joined #macports
[13:16:59] *** aber_ has quit IRC
[13:19:44] *** GlobalOD has quit IRC
[13:20:42] *** robottinosino has joined #macports
[13:20:43] <anddam> quick check
[13:20:44] <anddam> telnet rsync.macports.org 873
[13:21:06] <anddam> if you read    @RSYNCD: 30.0    your network is ok
[13:25:09] <CIA-37> mk at macports dot org * r89645 /trunk/dports/net/netpipes/Portfile:
[13:25:09] <CIA-37> http://trac.macports.org/changeset/89645
[13:25:09] <CIA-37> netpipes: add livecheck (closes issue #33135)
[13:39:02] *** wwii has joined #macports
[13:39:51] <CIA-37> and.damore at macports dot org * r89646 /trunk/dports/aqua/AquaLess/Portfile:
[13:39:51] <CIA-37> http://trac.macports.org/changeset/89646
[13:39:51] <CIA-37> port aqualess: removing unneeded post-destroot augmented phase, closing #30855
[13:41:11] <wwii> Hi
[13:41:49] <CIA-37> and.damore at macports dot org * r89647 /trunk/dports/aqua/AquaLess/Portfile:
[13:41:50] <CIA-37> http://trac.macports.org/changeset/89647
[13:41:50] <CIA-37> port aqualess: adding missing license keyword
[13:41:52] *** robottinosino has quit IRC
[13:42:12] <anddam> lunchtime
[13:42:13] *** anddam has quit IRC
[14:04:50] *** Protux has joined #macports
[14:23:17] *** anddam has joined #macports
[14:23:18] <mpbot> anddam is and.damore and maintainer of libtifiles2, vnstat, libticables2, and dineroIV (of 48 total)
[14:27:22] *** sixcorners has joined #macports
[14:28:16] <sixcorners> Why doesn't this function like I think it should? http://pastebin.com/jkwCkdud
[14:29:25] <anddam> sixcorners: port isn't a script parser
[14:29:47] <anddam> the shebang line should be of a shell
[14:30:05] <sixcorners> it accepts cmd files
[14:30:11] <anddam> sixcorners: btw that pastebin doesn't show what you think, it's left to reader's guessing
[14:30:17] <anddam> dunno, let me check
[14:33:05] <anddam> IMHO that won't work, idk the details but it's up to the shell how to execute a script
[14:33:13] <anddam> I mean the bash details
[14:33:52] <sixcorners> I tried to leave the shell out of it.. just have it do port -F cmdfile
[14:34:15] <Raim> sixcorners: hm, there seems to be a problem with the expansion of the pseudo-ports
[14:34:28] <anddam> Raim: so it's supposed to work?
[14:34:36] <Raim> anddam: yes
[14:34:44] <anddam> I get the actual script lines when running the file
[14:35:06] <anddam> ah right, there's no expansion
[14:35:14] <anddam> I'm reading the pseudo-ports selectors
[14:35:19] <anddam> my bad
[14:35:48] <Raim> sixcorners: as a workaround, you can also do that on a single command line: sudo port selfupdate \; upgrade outdated request \; ... \;
[14:36:29] <anddam> Raim: it's correctly expanding selectors using stdin
[14:36:50] <anddam> port -F - < mport.script
[14:37:42] <Raim> this works?
[14:37:48] <anddam> this works
[14:38:07] <sixcorners> It seems to work when I do port -F nameoffile
[14:38:13] <anddam> obviously I'm using a shorter command  "echo installed and name:lua"
[14:38:20] <anddam> sixcorners: that's the same
[14:38:36] <anddam> now I'm curious how bash parse scripts
[14:39:49] * anddam heads to #bash
[14:40:03] <Raim> this does not have anything to do with bash
[14:40:35] <Raim> the line #!/opt/local/bin/port -F tells your kernel to execute this script using port, there is no bash involved
[14:41:15] <sixcorners> For some reason when I execute this file sh or bash (not sure) is executing the file..
[14:41:31] <sixcorners> I am executing it directly.. ./portupdate
[14:41:39] <anddam> Raim: there has to be bash in the middle, I'm giving commands to a shell after all
[14:41:57] <anddam> maybe it passes control as soon it sees the first argument of a line is an executable file
[14:42:35] <sixcorners> It works with #!/usr/bin/env wakeonlan -f
[14:42:47] <sixcorners> I'm not sure the problem I am having is a macports problem..
[14:42:48] <anddam> and there's still room for my doubt, is the file we just called passed as an argument to the file specified in shebang?
[14:44:24] <Raim> uhm, I might have been wrong... somehow this is not executed with port, but with sh :-S
[14:44:36] <sixcorners> yes : \
[14:45:14] <Raim> yeah, it works with /usr/bin/env portf
[14:45:15] <sixcorners> I see how you thought it was an expansion problem.. You did something like 'echo installed' right?
[14:45:23] <Raim> exactly :-D
[14:45:24] * anddam nelsons Raim
[14:45:45] <anddam> http://goo.gl/I9y0
[14:45:49] <anddam> ;-)
[14:45:59] <Raim> heh
[14:47:11] <Raim> sixcorners: hm, I have a feeling that OS X no longer allows to use a script in a shebang line, only executable files
[14:47:19] <Raim> s/executable/binary/
[14:47:45] <sixcorners> but it works with wakeonlan
[14:48:03] <Raim> sixcorners: '#!/usr/bin/env port -F' should work as well
[14:48:20] <sixcorners> portf ... ? softlink to port
[14:48:34] <sixcorners> huh
[14:48:35] <Raim> sixcorners: portf is equivalent to port -F
[14:48:44] <sixcorners> ah
[14:48:46] <sixcorners> k
[14:49:04] <Raim> there are some OSes which do not allow additional arguments on a shebang line
[14:49:41] <sixcorners> .. I wonder why env is needed
[14:51:17] <neverpanic> maybe because Raim is right and OS X doesn't let you use scripts in a shebang? env is a binary
[14:52:21] <Raim> however, I would expect to error out in this case and not silently use $SHELL instead
[14:53:35] <Raim> well, it's not even using $SHELL, maybe it is the login shell
[14:54:55] <Raim> hah, "./foo: /opt/local/bin/doesnotexist: bad interpreter: Permission denied"
[14:55:24] <Raim> but using an existing file, that is not a binary simply falls back to bash
[14:58:13] <sixcorners> looks like I can't have comments in my cmdfile
[14:58:18] <anddam> Raim: http://www.faqs.org/faqs/unix-faq/faq/part3/section-16.html
[14:58:53] <anddam> check      "errno == ENOEXEC"
[14:59:08] <anddam> that's just an example but explains the fallback
[14:59:19] <anddam> we should look at XNU's source
[14:59:32] <anddam> btw is anyone able to update xulrunner?
[15:00:08] <anddam> I tried and failed, asked jeremyhu (hi, there) but it requires times
[15:00:17] <sixcorners> nvm comments work
[15:00:24] <anddam> mmm maybe s/times/time/
[15:01:37] <anddam> Raim: lol I just got the shell fallback thing, I was using "echo outdated" as test script so shell was obviously using builtin echo
[15:03:16] <Raim> anddam: uhm yeah, I totally wasn't aware of this fallback
[15:03:50] <anddam> tell me about it, I had to ask bash
[15:04:03] <neverpanic> maybe that's by design? If I create an executable file without a shebang it gets executed by my $SHELL
[15:04:40] <anddam> neverpanic: that seems exactly the example in the link
[15:04:54] <anddam> if (errno == ENOEXEC) execl ("/bin/sh", "sh", "-c", program, (char *)0);
[15:05:24] <anddam> that is file is executable but magic bits don't match
[15:05:49] <anddam> again we should be looking at XNU
[15:05:55] <anddam> does any of you build kernel btw?
[15:06:20] <anddam> I'd love to have a darwin system in my VM, useless but adorable
[15:06:29] <anddam> like a little Hexley-pet
[15:06:35] <sixcorners> Thanks for the help everyone. It's working like a charm now
[15:06:44] <anddam> sixcorners: I'm curious, how did you fix it?
[15:06:48] <sixcorners> env
[15:07:01] <anddam> we may add that to wiki
[15:07:27] <sixcorners> feel free to
[15:07:34] <anddam> oh, thanks :-)
[15:08:22] <sixcorners> Not sure how I could help with that
[15:11:47] <sixcorners> I guess I should have said 'neat' ... I don't want to sound uncaring.. I said "not sure hwo I could help with that" because I didn't mean to imply that you needed my permission...
[15:14:21] <anddam> I didn't think you were uncaring, just wordplaying
[15:15:19] <sixcorners> Might want to say to not add comments on to the ends of lines.. as they will be interpreted as arguments... or not.. not sure how obvious that is
[15:15:43] <sixcorners> It's probably mentioned elsewhere.. nevermind
[15:17:41] <sixcorners> As far as the contents of the file I posted.. Are those good commands to run every time I want to update my installed ports? http://pastebin.com/hG1dtQUY
[15:17:48] *** wwii has quit IRC
[15:20:08] <Raim> sixcorners: 'upgrade outdated requested' seems redundant
[15:20:26] <sixcorners> the idea is to only update the ports I requested so I don't update ports that might not be depended upon anymore
[15:21:15] <Raim> sixcorners: okay, then you want 'upgrade outdated and requested' (the implied logical operation is an or)
[15:21:37] <Raim> sixcorners: which means, upgrade all ports which are both in the list of outdated and requested ports
[15:22:43] <sixcorners> heh.. "upgrade outdated and requested" sounds like upgrade both groups..
[15:24:06] <sixcorners> If my requested ports depend on outdated ports or if they depend on any version of a different port.. Is it possible that unsetrequested ports might never be updated?
[15:24:43] <sixcorners> err.. nvm.. let me look at a portfile..
[15:31:06] <Raim> sixcorners: upgrade is a recursive operation, it upgrades dependencies first
[15:31:30] <sixcorners> ah, ok. thanks :)
[15:34:27] <neverpanic> upgrade outdated and requested is a good idea; maybe I should be doing this, too.
[15:35:03] <neverpanic> or, no, it's not. If there's a security update in some library that was pulled in as a dependency I want that to be upgraded, too
[15:36:43] <sixcorners> That's what I was thinking but if upgrade is recursive ports that were depended upon will be upgraded
[15:36:58] <sixcorners> leaves I uninstall though.. so there is that
[15:38:41] <Raim> sixcorners: I think you want 'leaves and not requested'
[15:38:44] <neverpanic> oh, right, so the only problem are libs linked against although not specified as dependencies, but those should be reported and fixed anyway
[15:39:23] <neverpanic> Raim: aren't requested ports never contained in leaves?
[15:40:07] *** phw has joined #macports
[15:40:09] <mpbot> phw is maintainer of py26-metar, libcaca, py26-tweepy, and toilet (of 106 total) and is in Nürnberg; local time is Sun 14:40 UTC
[15:41:23] <Raim> neverpanic: hm, right. I was mislead by all the build dependencies returned.
[15:41:59] *** drw has quit IRC
[15:42:52] <neverpanic> Raim: I recently ran port_cutleaves and rev-upgrade, turns out quite the number of leaves were actually being linked against
[15:43:29] <neverpanic> (that was the reason I asked about trace mode)
[15:44:35] <sixcorners> There are ports that have unlisted dependencies?
[15:45:03] <neverpanic> yes, unfortunately
[15:45:44] <neverpanic> autotools' configure scripts often enable features if the needed libraries are installed
[15:45:54] <neverpanic> in my case, some kerberos stuff when building subversion
[15:46:09] *** fab23 has quit IRC
[15:47:41] *** fab23 has joined #macports
[15:48:38] <sixcorners> So.. I upgrade all of my ports.. then I uninstall all my leaves.. Would kerberos stay? Would subversion become nonfunctional?
[15:51:21] <neverpanic> kerberos doesn't get pulled in by subversion, so chances are you'll never even have it
[15:51:58] <neverpanic> if it gets pulled in by some other port you have installed and you uninstall that it would become a leave, get removed during the next run of your script and break subversion
[15:52:13] <neverpanic> if you're running trunk you can port rev-upgrade to detect and fix the problem
[15:53:20] <sixcorners> So.. It's never a good idea to uninstall leaves then?
[15:53:49] <neverpanic> yes it is. if this happens it's a bug and should be reported and fixed
[15:54:54] <sixcorners> Seems like there is a way to solve this pragmatically.. Like DKMS and building kernel modules.
[15:55:49] <sixcorners> If the leaf is removed then ports that had a "transient" dependency on it need to be rebuilt
[15:57:15] <neverpanic> yeah, the ports dependency system needs some work
[15:57:28] <neverpanic> but we already have a working solution for this problem in trunk
[15:57:48] <neverpanic> rev-upgrade will detect missing links and rebuild, without the need for a transient dependency
[15:59:52] <sixcorners> Is there ever a time when you would want to have subversion installed then install kerberos and want those things to hook up automatically?
[16:00:47] <sixcorners> by rebuilding subversion?
[16:00:57] *** justdave has quit IRC
[16:01:08] <sixcorners> I don't know.. nevermind
[16:01:16] *** justdave has joined #macports
[16:01:48] <neverpanic> not really -- the correct MacPorts way would be to have a kerberos variant on subversion, and if you want kerberos support in your svn, enable it, and if you don't leave it disabled (and that should pass --disable-kerberos or something to ./configure, so it won't link against it, even if it's there)
[16:02:34] <sixcorners> How would a package manager handle it?
[16:03:11] <sixcorners> You only get one subversion binary right? Does it depend on kerberos or go without it?
[16:04:14] <neverpanic> what package manager do you mean?
[16:04:36] <sixcorners> apt
[16:06:11] <neverpanic> they either have two conflicting packages or go with shipping everything, i.e. with kerberos
[16:06:52] <neverpanic> but there's enough whining already about macports compiling the-world-and-even-more dependencies, so I'm always hesitant to add more dependencies that aren't strictly required without using a variant
[16:06:53] <sixcorners> huh.. k
[16:07:25] <neverpanic> think gentoo's use flags, they would probably require you to USE=+kerberos emerge subversion to get a svn with kerberos
[16:25:46] <neverpanic> ugh, trace mode is painfully slow, isn't it?
[16:33:05] <sixcorners> neverpanic: So are you going to use "upgrade outdated and requested" ?
[16:34:45] <neverpanic> it probably doesn't make a difference, so I'm not going to bother myself with typing it ;)
[16:34:56] <neverpanic> but I don't see a problem with it
[16:35:23] <sixcorners> k
[16:39:26] *** Bemmu has joined #macports
[16:41:44] *** TheInfinity has joined #macports
[16:50:34] *** weichkaes has joined #macports
[16:57:18] *** wwii has joined #macports
[16:59:37] <anddam> woa, went running - definitely cold for me
[17:00:29] *** TheInfinity has quit IRC
[17:05:14] *** TheInfinity has joined #macports
[17:05:40] *** Ionic has quit IRC
[17:07:09] *** TheInfinity has quit IRC
[17:08:14] *** Guest2561 has joined #macports
[17:11:01] <Raim> sixcorners: packages for apt (or rather dpkg) are usually built using fakeroot, which only allows access to files listed in the dependencies. this is the same what trace mode is supposed to achieve.
[17:12:19] <Raim> (it also removes the need to be root for building a package)
[17:15:35] <neverpanic> Raim: Btw, why aren't we using fakeroot, too?
[17:15:41] <neverpanic> It seems to compile fine on OS X.
[17:18:40] <sixcorners> hmm.. I have heard of chroot..
[17:18:55] <Raim> neverpanic: most probably, nobody implemented it yet? :-)
[17:20:40] <sixcorners> I have some reading to do later..
[17:21:17] <Raim> neverpanic: I think it would require some modifications as the load does not respect LD_PRELOAD, but DYLD_INSERT_LIBRARIES
[17:21:22] <Raim> *loader
[17:22:27] <Raim> neverpanic: and there was also something about the two-level namespaces of shared libraries, but I never cared to read more about how this works
[17:23:47] <neverpanic> Raim: I think it handles DYLD_INSERT_LIBRARIES fine, I tried it and it worked on OS X
[17:24:05] <neverpanic> not sure about the "denying access to files not in the whitelist" thing though, I haven't found that in the docs
[17:24:35] <Raim> neverpanic: maybe I was mistaken and that is achieved by building in a chroot...
[17:25:17] <neverpanic> chroot is always such a PITA to set up... :(
[17:25:43] <sixcorners> fakeroot seems like it doesn't have anything to do with being able to see files
[17:26:06] <Raim> neverpanic: ah, actually I meant fakechroot, not only fakeroot
[17:30:23] <sixcorners> interesting
[17:41:29] *** drw has joined #macports
[18:14:51] <neverpanic> I can't even get my bash working in darwintrace, it always terminates with Abort Trap: 6
[18:16:29] *** Achtwerk has joined #macports
[18:18:41] * anddam googles darwintrace
[18:19:30] <anddam> neverpanic: is it specific to darwinbuild?
[18:21:26] <neverpanic> yeah, somewhat
[18:21:46] <neverpanic> but since it can be loaded with DYLD_INSERT_LIBRARIES you can use it whatever you're trying to run
[18:22:02] <neverpanic> (as long as it's dynamically linked against libSystem)
[18:26:05] <neverpanic> ah, it works much better with the version from trunk
[18:35:36] * neverpanic wonders why a simple port search opens PortIndex 4(!) times
[18:36:10] * neverpanic should stop playing around with darwintrace before he starts to feel an urge to fix all these things
[18:38:45] *** wwii has quit IRC
[18:45:52] * anddam needs to install darwintrace
[18:46:15] <anddam> neverpanic: should this kind of stuff be done with dtrace?
[18:46:28] <neverpanic> no, not really
[18:46:33] <neverpanic> but it's interesting to see
[18:47:13] <neverpanic> oh, you mean dtrace(1) from solaris?
[18:47:17] <neverpanic> I have no idea what that does
[18:48:16] <anddam> oh lol
[18:48:42] <anddam> I have no idea what darwintrace does, I guessed it's something like linux's strace
[18:49:38] <neverpanic> it is to some extent, but strace has kernel support wheres darwintrace is only a preloaded library overwriting libc syscall wrappers
[18:49:41] <anddam> dtrace is the native framework to inspect the system and apps
[18:50:02] <neverpanic> and it allows to emulate a partial chroot on the fly
[18:50:18] <anddam> neverpanic: dtrace uses D scripts to define its behaviour, OSX comes with several prebuilt (useful) scripts
[18:50:30] <anddam> neverpanic: apropos dtrace
[18:52:18] <neverpanic> oh, so it's the backend for the GUI tools in Xcode doing that kind of stuff
[18:52:22] <neverpanic> forgot the name, though
[18:52:29] <Raim> Instruments.app
[18:53:27] <neverpanic> but that's a watch-only tool, you can't use it to make open(something) open somepath/something, right?
[18:53:42] <neverpanic> because if it could do that I wonder why darwinbuild doesn't use it
[18:53:48] <anddam> idk
[18:54:18] <anddam> I see the point, I've always seen it used in read mode tho'
[19:23:17] *** TheInfinity has joined #macports
[19:42:03] *** anddam has quit IRC
[19:58:51] *** lipsinV2 has quit IRC
[19:59:28] *** wwii has joined #macports
[20:38:05] *** wwii has quit IRC
[20:40:14] *** wwii has joined #macports
[20:48:11] *** cbmuser has quit IRC
[20:48:22] *** Lichtlos has quit IRC
[20:49:21] *** Lichtlos has joined #macports
[20:49:46] *** Achtwerk has left #macports
[20:52:06] *** cbmuser has joined #macports
[21:01:28] *** mattg has quit IRC
[21:05:17] *** jkyle has joined #macports
[21:06:02] *** wwii has quit IRC
[21:06:33] *** mattg has joined #macports
[21:10:19] *** jkyle_ has joined #macports
[21:10:34] *** jkyle has quit IRC

top