February 2008 - macports-dev - lists.macosforge.org

tuxmath Portfile
by David Bruce 17 Feb '08

17 Feb '08

Hi, After some study of the guide (as well as several existing Portfiles), I've taken my first stab at writing a Portfile for tuxmath and submitted it as trac ticket #14279. How can I start testing it to see what needs to be fixed? Tuxmath uses automake, so I am fairly sure it supports DESTDIR already. Thanks, and pardon my noob-hood with regard to MacPorts. -- David Bruce

2 1

Splitting the guide
by Juan Manuel Palacios 16 Feb '08

16 Feb '08

Hey guys! The guide is advancing so nicely that I'm afraid we've hit the point when it's maybe becoming a tad too large for a single page, wouldn't y'all think? Every time I have to load a section of the guide a have a moment of hesitation in fear of loading such a large document! At this point in time, wouldn't it be better to have a single html page per chapter? Just my thoughts,... -jmpp

1 0

RE: linking for dummies
by markd＠macports.org 16 Feb '08

16 Feb '08

>> Thanks for the reply. It appears that the problem was not the addition >of >> "-lgcc_s.10.5 -lgcc -lSystem" (only seen in verbose gcc mode), but the >> inclusion of the "-lrrd". I've never tried verbose mode before so it >was >> unfamiliar to me. For some reason the "-lrrd" was unncessary. I'm >> following up with the developer to see if a source fix is in order. > >What is the error that you are getting? An extraneous library on the >link line is not usually a big issue. > >> >> BTW, is there a way to turn on gcc verbose mode with a MacPorts flag? > >You shouldn't need to. Actually, the error was not a compile error but a runtime error in the app after install. I really don't know why the extra library was causing the problem. Maybe the developer will be able to explain. Mark

1 0

Re: linking for dummies
by markd＠macports.org 16 Feb '08

16 Feb '08

>> I have an app where >> >> "-lgcc_s.10.5 -lgcc -lSystem" >> >> is getting appended at the end of the stuff getting linked, and I can't >> figure out where this is coming from. It appears to me that gcc is >doing >> it on its own. Instead of the linker item "-lrrd" normally appearing >at >> the end, instead I see it in the middle like this: >> >> "-lrrd -lgcc_s.10.5 -lgcc -lSystem" >> >> and I'm told that is a problem. But I can't for the life of me see the >> Makefile doing anything that would cause this. Does anyone know how >this >> extra stuff gets added on, or how to make "-lrrd" go on the end of the >> linker line? > >The compiler appends these, you do not normally want it not to. Are you >seeing errors because of this? Thanks for the reply. It appears that the problem was not the addition of "-lgcc_s.10.5 -lgcc -lSystem" (only seen in verbose gcc mode), but the inclusion of the "-lrrd". I've never tried verbose mode before so it was unfamiliar to me. For some reason the "-lrrd" was unncessary. I'm following up with the developer to see if a source fix is in order. BTW, is there a way to turn on gcc verbose mode with a MacPorts flag? Mark

2 1

Re: Trac ticket that need your attention: py25-mysql
by js 16 Feb '08

16 Feb '08

Sorry to disturb, but another reminder. > Maintainers, > > This is the reminder for the ticket which hasn't received no response > from maintainers more than 72 hours. > http://trac.macosforge.org/projects/macports/ticket/13410 > > Please take a look and commit the patch if everything's fine >

2 1

linking for dummies
by markd＠macports.org 15 Feb '08

15 Feb '08

I have an app where "-lgcc_s.10.5 -lgcc -lSystem" is getting appended at the end of the stuff getting linked, and I can't figure out where this is coming from. It appears to me that gcc is doing it on its own. Instead of the linker item "-lrrd" normally appearing at the end, instead I see it in the middle like this: "-lrrd -lgcc_s.10.5 -lgcc -lSystem" and I'm told that is a problem. But I can't for the life of me see the Makefile doing anything that would cause this. Does anyone know how this extra stuff gets added on, or how to make "-lrrd" go on the end of the linker line? Mark

2 1

The ticket that need your attention: py25-feedparser
by js 15 Feb '08

15 Feb '08

The maintainer didn't answer the ticket within 72 hours. (According to MacPorts Guide, a maintainer's supposed to apply patches within 72 hours) Here's the ticket: http://trac.macosforge.org/projects/macports/ticket/14265 Please take a look at this and commit the patch. Thanks.

1 0

Improvements for automated port lint
by Ryan Schmidt 14 Feb '08

14 Feb '08

Thanks to William Siegrist for setting up the automated port lint and email following each commit! Now here's some thoughts about some problems and how they could be dealt with. * I'm worried what will happen when someone does a batch cleanup operation affecting dozens or hundreds of portfiles in a single commit. (We've had this situation in r33441, r30218, r28561, r22478, r19376, etc.) I don't want this to fire off dozens or hundreds of lint emails. A thought here is that if a single commit affects, say, more than 5 portfiles, no lint report is run and no emails are sent. Or, just one email could be sent to the committer, letting them know why lint was not run. * The subject line of the lint emails reads "[MacPorts Lint] Portfile Lint Errors for: <port>". Not all information returned by lint is an error though; some of it is just warnings. There's also a lot of words up front there in the subject that I don't need. I would change the subject to "[<rev>] <port> lint report" * Possibly obsoleting both of the above observations, what would people think about appending the port lint report to the diff email that's already generated and sent to macports-changes? I'm not sure if this is the best idea, or even possible with the diff email script we use, but the idea occurred to me so I thought I'd see what others think.

2 4

Re: [34108] trunk/dports/math
by Ryan Schmidt 13 Feb '08

13 Feb '08

On Feb 13, 2008, at 08:11, raimue(a)macports.org wrote: > Revision: 34108 > http://trac.macosforge.org/projects/macports/changeset/34108 > Author: raimue(a)macports.org > Date: 2008-02-13 06:11:37 -0800 (Wed, 13 Feb 2008) > > Log Message: > ----------- > math/libsvm: > New Port submitted by Florian Ebeling, closes #13779 See my follow-up ticket for some concerns: http://trac.macosforge.org/projects/macports/ticket/14310

2 1

Re: [34110] trunk/dports/mail/archivemail/Portfile
by Markus Weissmann 13 Feb '08

13 Feb '08

I havent tried, but this looks totally like a case for the Python group code; also you should just rely on a concrete version of Python as the path to the "real" python executable gets hardcoded in the "archivemail" python program ($prefix/Library/Frameworks/Python.framework/Versions/ 2.4/Resources/Python.app/Contents/MacOS/Python for Python 2.4..) cheers, -Markus On 13 Feb 2008, at 18:16, simon(a)macports.org wrote: > Revision34110Authorsimon(a)macports.orgDate2008-02-13 09:16:14 -0800 > (Wed, 13 Feb 2008)Log Message > mail/archivemail: Updated to version 0.7.2. Also removed unnecessary > compression of man page. > Modified Paths > • trunk/dports/mail/archivemail/Portfile > Diff > Modified: trunk/dports/mail/archivemail/Portfile (34109 => 34110) > --- trunk/dports/mail/archivemail/Portfile 2008-02-13 16:57:22 UTC > (rev 34109) > +++ trunk/dports/mail/archivemail/Portfile 2008-02-13 17:16:14 UTC > (rev 34110) > @@ -3,7 +3,7 @@ > PortSystem 1.0 > > name archivemail > -version 0.7.0 > +version 0.7.2 > categories mail python > platforms darwin > maintainers nomaintainer > @@ -16,20 +16,30 @@ > > homepage http://archivemail.sourceforge.net > master_sites sourceforge > -checksums sha1 fcba04b7ae748d54b8ebcbecb0aad08c4f58e927 > +checksums md5 e444424688e6ec063e829176e4eb62e2 \ > + sha1 0ff9b8991b04f09cf9536c45b6f9e05d2427a459 \ > + rmd160 da85dff5114100d76f8f5ac19172d8a557801455 > > depends_lib bin:python:python24 > > use_configure no > build {} > > -destroot.cmd ${prefix}/bin/python setup.py install > +# I don't know a better way so it works even if python_select is not > +# installed. If there is please tell me (simon(a)macports.org) or > just change > +# it. > +if {[file exists ${prefix}/bin/python]} { > + set python python > +} else { > + set python python2.4 > +} > + > +destroot.cmd ${prefix}/bin/${python} setup.py install > destroot.destdir --prefix=${prefix} \ > --root=${destroot} \ > --install-data=${prefix}/share > post-destroot { > - system "gzip -9 ${destroot}${prefix}/share/man/man1/archivemail. > 1" > - xinstall -m 755 -d ${destroot}${prefix}/share/doc/${name} > + xinstall -d ${destroot}${prefix}/share/doc/${name} > xinstall -W ${worksrcpath} CHANGELOG FAQ README TODO \ > ${destroot}${prefix}/share/doc/${name} > } > _______________________________________________ > macports-changes mailing list > macports-changes(a)lists.macosforge.org > http://lists.macosforge.org/mailman/listinfo/macports-changes -- Dipl. Inf. (FH) Markus W. Weissmann http://www.mweissmann.de/

1 0