October 2006 - macports-dev - lists.macosforge.org

BTW, 1.3.2 .dmg
by Jyrki Wahlstedt 20 Oct '06

20 Oct '06

Hi, should someone package the current release as .dmg? Should the .dmg (of the current, or next release) be named MacPorts? I saw that Ryan has mentioned this on Sep 11th, but with no answer or action. Should a ticket be opened? ! ! Jyrki Wahlstedt ! Kolmas linja 12 A 18 mob. +358-400-347 015 skype:jyrkiwahlstedt ! FI-00530 Helsinki www.wahlstedt.fi ! ! Our life is no dream; but it ought to become one and perhaps will. ! PGP key ID: 0x139CC386 fingerprint: F355 B46F 026C B8C1 89C0 A780 6366 EFD9 139C C386

3 3

Problem Hotlist entry for gettext trouble
by Mark Duling 18 Oct '06

18 Oct '06

Frank McPherson has contributed a shell script to help people upgrade their gettext dependent ports. The previous command I thought might work (port -fR upgrade gettext) has proved to be highly unsatisfactory. http://trac.macosforge.org/projects/macports/wiki/ProblemHotlist Mark

1 0

New Wiki page - Problems hotlist
by Mark Duling 16 Oct '06

16 Oct '06

I added a new "MacPorts Support Requests" top-level section to the Wiki. It also has a page called "Problems Hotlist" to track frequently encountered problems. It is different from a FAQ entry and is intended to be a fairly short and fluid list that is intended to be checked before filing bugs. The first and only entry is on the current gettext iproblems. I don't know if the description I gave is adequate or if the command I listed really works. If not, please adjust it so it makes sense so we can just refer people to the list from now on. Mark http://trac.macosforge.org/projects/macports/wiki/ProblemHotlist http://trac.macosforge.org/projects/macports/wiki/TracTicketing

2 1

macports without root?
by Paul M. Lambert 16 Oct '06

16 Oct '06

I'm trying to build a macports installation that runs entirely out of a user's home directory. I've compiled macports without launchd support, and set the install user and group to the appropriate user. However, when I try to run "port install mysql5" it fails, apparently because it's trying to chown a file to root. Is this a problem with that port, or does macports require root in general? If it's the latter, any suggestions on how I could go about patching it to allow fully non-root user installations? Paul

5 8

Trac/Python24
by Dominik Madon 15 Oct '06

15 Oct '06

Hello, I'm having troubles to run tracd on my MBP. It looks like the _locale module of python has not been compiled and, in fact, it's not when I do a port install python24: Traceback (most recent call last): File "/opt/local/lib/python2.4/site-packages/trac/web/main.py", line 356, in dispatch_request dispatcher.dispatch(req) File "/opt/local/lib/python2.4/site-packages/trac/web/main.py", line 206, in dispatch populate_hdf(req.hdf, self.env, req) File "/opt/local/lib/python2.4/site-packages/trac/web/main.py", line 74, in populate_hdf hdf['trac'] = { File "/opt/local/lib/python2.4/site-packages/trac/util/ datefmt.py", line 61, in format_datetime encoding = locale.getpreferredencoding() File "/opt/local/Library/Frameworks/Python.framework/Versions/2.4/ lib/python2.4/locale.py", line 399, in getpreferredencoding import _locale ImportError: No module named _locale Question: Is there any easy way to include this module when installing python24 or is there something special to do get trac/ postgres to run? Dom

3 8

Panics?
by Jyrki Wahlstedt 12 Oct '06

12 Oct '06

Hi, this must be an odd question to make, but I've recently had difficulties with my system and I'd like to confirm that my system (i.e. computer, currently at 10.4.8 and Xcode 2.4) is a more probable culprit then the ports system. The symptom is that installing ports or getting info about them fails. I've even had panics like the following: ======= ********* Thu Oct 12 12:17:34 2006 panic(cpu 0 caller 0x00215C84): bad dir Latest stack backtrace for cpu 0: Backtrace: 0x00095138 0x00095650 0x00026898 0x00215C84 0x000FB124 0x000E1EE8 0x000E1A7C 0x000EE74C 0x000EEB04 0x000EEBA4 0x002AAE28 0x000AB930 0xFF4F0000 Proceeding back via exception chain: Exception state (sv=0x20F27C80) PC=0x90021A0C; MSR=0x0000F030; DAR=0x00306000; DSISR=0x42000000; LR=0x9003CB68; R1=0xBFFF9F70; XCP=0x00000030 (0xC00 - System call) Kernel version: Darwin Kernel Version 8.8.0: Fri Sep 8 17:18:57 PDT 2006; root:xnu-792.12.6.obj~1/RELEASE_PPC ********* ======= This one was obtained by writing "ll $(port contents mysql5)" in the Terminal (should produce some garbage due to "port x contains", but then produce a long list of information on the files, this is related to the symptom that symbolic links often are empty in my installs). The previous panic port was ruby, but it did not install at all. I don't really see a chance myself for the port command to cause a panic in such a predictable way (as tcl is using normal userland system services!?). It is more probably so(?) that there are some bad blocks or something that these installs have hit. If I am the only one who has experienced things like this, I'll start to complain in one direction… ! ! Jyrki Wahlstedt ! Kolmas linja 12 A 18 mob. +358-400-347 015 skype:jyrkiwahlstedt ! FI-00530 Helsinki www.wahlstedt.fi ! ! Our life is no dream; but it ought to become one and perhaps will. ! PGP key ID: 0x139CC386 fingerprint: F355 B46F 026C B8C1 89C0 A780 6366 EFD9 139C C386

1 0

Bug filing criterion / application errors
by Mark Duling 11 Oct '06

11 Oct '06

Is it fair to day that in general that bugs should not be filed for application issues? I'm working on documentation and it sems to me we need some rough criteria for filing bugs. Esoteric application errors are something we can do little about, and I susoect when people do this they are not reporting them upstream where they may be fixed. Of course if it is a bug that afects our users and would generate questions and/or cause us to delay an upgrade is always an exception just to name one, but esoteric application bugs that don't affect other apps at all don't seem to be a good idea, especially if the user neglects to report it upstream. Do others agree that for bug filing guidelines that bugs for MacPorts should generally be confined to installation and configuration issues, with exceptions roughly spelled out? I'm not saying it is a big problem but still I think some rough guidelines like this would be a good thing. It is better than stating nothing and then marking bugs "invalid" immediately because people get sorta miffed. Mark

2 1

X11 ports
by Mark Duling 11 Oct '06

11 Oct '06

So I've never dealt with porting X11 programs before. Since we all use a non-MacPorts X11 now, does that make it acceptable to write outside ${prefix] for X11 ports? Specifically, this submission writes to /usr/X11R6/bin/ and /usr/X11R6/lib (via destroot). Please advise. http://trac.macosforge.org/projects/macports/ticket/7488 Mark

3 6

Default revision of 1
by Blair Zajac 10 Oct '06

10 Oct '06

When I'm developing a new port or want to override the default port in MacPort's svn with a local port of a newer upstream release, it would be nice to have a range of revision numbers that is smaller than any revision number from MacPort's svn repository. This would enable me to have a local port, say subversion 1.4.1-0, and then when the real subversion 1.4.1-1 comes out, I'll get an out of date warning. Also, is the revision an integer, a floating point value, or does it compare values like the version number, it 0.10 is greater than 0.9? Would people mind defaulting the revision to 1? I don't know how this would be rolled out. It would cause a massive upgrade. Regards, Blair -- Blair Zajac, Ph.D. CTO, OrcaWare Technologies <blair(a)orcaware.com> Subversion training, consulting and support http://www.orcaware.com/svn/

5 12

Consistent --program-prefix for GNU programs
by Blair Zajac 10 Oct '06

10 Oct '06

Some of our GNU ports use a --program-prefix of gnu: findutils gnutar gnumake-devel gsed indent while others use g: binutils coreutils diffutils gmake gpatch gtime gunits gwhich libtool libtool-devel m4 It would be nice to have a MacPorts' standard here and it looks like changing the gnu to the g ones would be less work I'd also like all of these ports to have a variants to disable the program-prefix. Currently, only diffutils has a variant named 'normal', which is not a good name in my opinion. I'd like to see a variant named no-program-prefix. What do people think? Regards, Blair

2 2