July 2007 - macports-dev - lists.macosforge.org

livecheck.regex problems due to r26041?
by Ryan Schmidt 17 Jul '07

17 Jul '07

Could our recent livecheck regex issues (where regexes that we thought worked before suddenly are requiring additional escaping) be due to r26041? http://trac.macosforge.org/projects/macports/changeset/26041 The log message reads "Fix livecheck to properly de-escape livecheck.url" and the revision does indeed do that. But it also changes the processing of livecheck.regex without explaining why.

1 0

Base Enhancements
by Chris Pickel 17 Jul '07

17 Jul '07

Hi all, I'd like to call your attention to four reports I just posted to Trac for developer review. None of them are too major, but I figure it's safer to ask before committing. To sum up: 12309 [1] - Currently we use sed to do some replacement in port, portindex, and portmirror. I think autoconf is better suited to this. 12310 [2] - Right now we detect some libraries (curl, readline, etc.) and then we link them all into every dylib we compile. This patch slims it down so that decisions can be made separately for each dylib. 12311 [3] - On the same note as my previous post to -dev about macports_fastload.tcl. This is a fix for the version comparison problem. 12312 [4] - Several manpages reference paths according to variables, rather than giving the actual path; we can do the latter and that's more helpful to users. [1] http://trac.macports.org/projects/macports/ticket/12309 [2] http://trac.macports.org/projects/macports/ticket/12310 [3] http://trac.macports.org/projects/macports/ticket/12311 [4] http://trac.macports.org/projects/macports/ticket/12312 The only one I have a strong attachment to is 12310, but I feel the others are improvements too. Chris

2 2

Re: [macports-mgr] Re: MacPorts Guide infrastructure requirements
by Juan Manuel Palacios 17 Jul '07

17 Jul '07

On Jul 11, 2007, at 3:38 AM, markd(a)macports.org wrote: > I started fairly clean because I think the logical structure and > organization of the old guide is fundamentally flawed. So I'm of the > opinion that we should not merge new stuff into the old guide, and I'm > taking the tack of merging the old stuff into a new guide. However, I > just went with the DocBook type "article" and a single file since > that was > familiar to me and fastest for me. So I don't favor merging new > into old, > but yet I don't know if I should convert my article to a "book" and/or > break it up into multiple files along major section. The latter I > suspect > yes; the former I don't know. Can anyone give advice here? > > BTW, I just reorganized the doc structure. Take a look at the TOC. I > like it better now; what do others think? Look past the ugly > colors. :) > > http://homepage.mac.com/duling/macports/guide.html > > Mark Already replied to most of these comments as a PS in my reply to your last mail to me (rsync replicating servers setup), but I'll do here briefly too to emphasize: -) totally love the TOC sidebar, definitely a keeper! -) I believe the source should be split in multiple xml files, easier to manage like that in my opinion; -) article, book... don't know much about dockbook myself, I have to admit, but definitely do enjoy having each chapter in a single html page; is that doable in article mode (while retaining the sidebar in each chapter)? -) don't merge new into old, lets populate new and move it in place once it's ready or close to ready, moving the source docs around as needed and (finally, removing the old!); -) lets set up automated regen in coordination with both Daniel, now, and Kevin on MacOSForge servers later on once he has the server side support finished. Regards,... -jmpp

4 5

Re: flagpoll Portfile for consideration in MacPorts
by Doug McCorkle 17 Jul '07

17 Jul '07

On Jul 10, 2007, at 5:43 PM, Doug McCorkle wrote: > > On Jul 10, 2007, at 8:19 AM, Doug McCorkle wrote: > >> >> On Jul 10, 2007, at 1:36 AM, Ryan Schmidt wrote: >> >>> >>> On Jul 9, 2007, at 15:57, Doug McCorkle wrote: >>> >>>> Attached is a Portfile for flagpoll https:// >>>> realityforge.vrsource.org/view/FlagPoll . Please let me know if >>>> there is anything else that needs to be done for its inclusion >>>> in MacPorts. Thanks. >>> >>> I think you should be using the python24 or python25 portgroup. >>> It would simplify the portfile. >>> >> I am still learning all the details of Portfiles so do not >> understand what a portgroup is. Could you provide more details on >> what I need to do to make flagpoll part of the python portgroup? >>> I found the man page that describes PortGroups but there is no description of what the Python PortGroups override. Is that documented anywhere? >>> Also, we have a python category. Though there are ports that use >>> a python portgroup that aren't in the python category. I don't >>> use python software, so I'm not sure what the rules are there, >>> just wanted to point out that there is such a category. Ports in >>> that category should have names that begin with "py-" (for 2.4) >>> or "py25-" (for 2.5). Perhaps a MacPorts python user can provide >>> guidance on what needs to go in that category and what can be >>> elsewhere. >>> >>> >> >> OK. I am happy to change the category to whatever you think the >> best home for the port is. > > I looked at pkgconfig today and it is stored in the devel category. Doug

3 4

Re: [27041] trunk/dports/lang/gcc40/Portfile
by Ryan Schmidt 16 Jul '07

16 Jul '07

On Jul 16, 2007, at 08:00, source_changes(a)macosforge.org wrote: > Revision: 27041 > http://trac.macosforge.org/projects/macports/changeset/27041 > Author: mww(a)macports.org > Date: 2007-07-16 06:00:03 -0700 (Mon, 16 Jul 2007) > > Log Message: > ----------- > change program suffix from '-dp-' to '-mp-', increment revision > > Modified Paths: > -------------- > trunk/dports/lang/gcc40/Portfile Fixed ports using gcc40: http://trac.macosforge.org/projects/macports/changeset/27055

1 0

Re: [27040] trunk/dports/lang/gcc41/Portfile
by Ryan Schmidt 16 Jul '07

16 Jul '07

On Jul 16, 2007, at 07:58, source_changes(a)macosforge.org wrote: > Revision: 27040 > http://trac.macosforge.org/projects/macports/changeset/27040 > Author: mww(a)macports.org > Date: 2007-07-16 05:58:31 -0700 (Mon, 16 Jul 2007) > > Log Message: > ----------- > change program suffix from '-dp-' to '-mp-', increment revision > > Modified Paths: > -------------- > trunk/dports/lang/gcc41/Portfile But why would you commit just that change? You've now broken all the ports that depend on gcc41. This is what I meant when I wrote "we'll have to update a lot of portfiles too." Fixing ports that depend on gcc41: http://trac.macosforge.org/projects/macports/changeset/27052

1 0

Fwd: [27045] trunk/dports/security/cyrus-sasl2/Portfile
by Juan Manuel Palacios 16 Jul '07

16 Jul '07

Begin forwarded message: > From: source_changes(a)macosforge.org > Date: July 16, 2007 2:42:35 PM GMT-04:00 > To: macports-changes(a)lists.macosforge.org > Subject: [27045] trunk/dports/security/cyrus-sasl2/Portfile > Reply-To: macports-dev(a)lists.macosforge.org > > Revision 27045 > > Author jmpp(a)macports.org > > Date 2007-07-16 11:42:34 -0700 (Mon, 16 Jul 2007) > > Log Message Commit patch contributed by Ryan in #12279 to > tempoarily fix archival issue of packages that list two variant > names in a single code block. Closes #12279. NOTE: Consider this > only a temporary, Portfile specific fix, the real bug in MacPorts' > base should be sought and fixed seperately. Tracked in #ticket > 12306. As for this particular Portfile, Ryan's fix is also a better > approach 'cause 'darwin' and 'freebsd' shouldn't be listed as > variants but platforms rather, whereas 'kerberos' is a valid > variant name. > > Modified Paths trunk/dports/security/cyrus-sasl2/Portfile Hey guys! I just committed Ryan's patch to fix the error installing cyrus- sasl2 all of you reported. Please do excuse me for taking so long to act on it, I've been busy ;-) My commit log contains some information of what's at fault here, but further inspection (reading through all the mails I got for this) reveals the error is not in the archivemode code but actually independent of it, both Tabitha and Blair proved the same error crops up when installing directly (either image mode or direct, I just meant having archivemode disabled). Interestingly enough, the first report I got for this error was on June 15th, the same day I merged my dp2mp-move branch back into trunk. Not sure if the two are related though, as I did not touch anything in the installation process other than consistently changing names across all our sources. I'll investigate however. Ticket tracking this error is #12306 per my commit log, but if you're only interested in getting back up and running simply wait about half an hour from now and sync your sources, Ryan's patch should flow into your ports tree at the time. Regards (and sorry again for the wait!),... -jmpp

1 0

Questions about hardcoded rsync_path and svn_path
by Simon Ruderich 16 Jul '07

16 Jul '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, today I had a problem that "port sync" wasn't working anymore. The problem was that it used a rsync version which was removed by me. I didn't knew that macports was using hardcoded paths and so had to search for the offending configuration data. After some time I thought I found it in this file: ${prefix}/share/macports/Tcl/port1.0/port_autoconf.tcl I changed the values but it wasn't working. So I looked further and found it here: /Library/Tcl/macports1.0/macports_autoconf.tcl So I have two questions about it: 1. Why is macports using hardcoded paths for this sort of values. I tried just using "rsync" and it works. 2. Why are there two (configuration) files similar data. I had some problems with macports hardcoded paths some time ago when I tried to install multiple port installations. I never really fixed this and have to rename my port installations to get it work. Thanks for your help, Simon - -- + privacy is necessary + using http://gnupg.org + public key id: 0x6115F804EFB33229 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (Darwin) iD8DBQFGlnlAYRX4BO+zMikRCpMyAKCuZtn/5shIB4KC+8sh+q1HFTsNJACePM5x bSUwK8IzT+JggRQ+rIIM6IA= =1X0O -----END PGP SIGNATURE-----

2 2

1.5.0 packaging bug
by Blair Zajac 16 Jul '07

16 Jul '07

Doing a complete rebuild of 1.5.0 today, I ran into this bug: man8/sasldblistusers2.8.gz: changing permissions from 00644 to 00444 man8/saslpasswd2.8: 51.8% -- replaced with man8/saslpasswd2.8.gz man8/saslpasswd2.8.gz: changing permissions from 00644 to 00444 ---> Packaging tbz2 archive for cyrus-sasl2 2.1.21_0+kerberos darwin ---> Creating cyrus-sasl2-2.1.21_0+kerberos darwin.i386.tbz2 bzip2: Can't open input file darwin.i386.tbz2: No such file or directory. /usr/bin/tar: -: Cannot open: Bad file descriptor /usr/bin/tar: Error is not recoverable: exiting now Error: Target org.macports.archive returned: shell command " cd "/opt/ local/var/macports/build/ _opt_local_var_macports_sources_rsync.macports.org_release_ports_securit y_cyrus-sasl2/work/destroot" && /usr/bin/tar -cvf - . | /opt/local/ bin/bzip2 -c9 > /opt/local/var/macports/packages/darwin/i386/cyrus- sasl2-2.1.21_0+kerberos darwin.i386.tbz2 " returned error 1 Command output: bzip2: Can't open input file darwin.i386.tbz2: No such file or directory. Regards, Blair -- Blair Zajac, Ph.D. <blair(a)orcaware.com> Subversion training, consulting and support http://www.orcaware.com/svn/

2 2

Re: [27033] trunk/dports/textproc/chmlib/Portfile
by markd＠macports.org 16 Jul '07

16 Jul '07

macports-dev(a)lists.macosforge.org on Sunday, July 15, 2007 at 9:53 PM -0800 wrote: >Log Message > >Add examples variant. > >This includes useful tools that use chmlib. For example, >extract_chmLib takes a chm file and extracts it to a directory. > >72 hour rule. > > >Modified Paths > > [ >fcp://@bubbs.biola.edu,%235000280/Mailbox/%23218807261#trunkdportstextprocchmlibPortfile >]trunk/dports/textproc/chmlib/Portfile > Is there is a reason not to enable examples by default? If it doesn't have undesirable requirements and it is useful, perhaps it should be enabled by default, adhering to our principle of minimizing variants. Mark

2 2