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

Re: [27705] trunk/dports/x11/gtk2/Portfile
by Randall Wood 14 Aug '07

14 Aug '07

What would be the best way to check for darwin platforms with a version less than 8 in a Portfile? On 12 Aug 2007, at 09:50, source_changes(a)macosforge.org wrote: > Revision 27705 Author rhwood(a)macports.org Date 2007-08-12 06:50:02 > -0700 (Sun, 12 Aug 2007) Log Message configure.env -> > configure.*flags-append statements redo dependencies based on trace > output replace ui_msg blah ; exit 1 statements with return 1 blah > statements Modified Paths > trunk/dports/x11/gtk2/Portfile > Diff > Modified: trunk/dports/x11/gtk2/Portfile (27704 => 27705) --- trunk/ > dports/x11/gtk2/Portfile 2007-08-12 12:03:22 UTC (rev 27704) +++ > trunk/dports/x11/gtk2/Portfile 2007-08-12 13:50:02 UTC (rev 27705) > @@ -26,26 +26,44 @@ use_bzip2 yes -depends_lib port:glib2 port:tiff > port:libpng port:atk port:pango \ - port:gettext > lib:libX11.6:XFree86 +depends_build \ + port:perl5.8 \ + > port:pkgconfig \ + port:gzip +depends_lib \ + port:cairo \ + > port:fontconfig \ + port:freetype \ + port:glib2 \ + port:jpeg \ + > port:tiff \ + port:libiconv \ + port:libpng \ + port:atk \ + > port:pango \ + port:gettext \ + port:render \ + port:xrender \ + > port:zlib \ + lib:libX11.6:xorg pre-configure { if {[variant_isset > quartz]} { if { ![file exists ${prefix}/include/cairo/cairo- > quartz.h] } { - ui_msg "\nYou must first build cairo with the > quartz variant enabled. Please\nuninstall (or deactivate) the cairo > port and reinstall by running:\n\n\"port install cairo +quartz\"\n" > - exit 1 + return 1 "\nYou must first build cairo with the quartz > variant enabled. Please\nuninstall (or deactivate) the cairo port > and reinstall by running:\n\n\"port install cairo +quartz\"\n" } } > if {[file exists ${prefix}/bin/cups-config]} { - ui_msg "\nThe > cups-headers port may prevent building this port. Please uninstall\n > (or deactivate) cups-headers and restart the build.\n" - exit 1 + > return 1 "\nThe cups-headers port may prevent building this port. > Please uninstall\n(or deactivate) cups-headers and restart the > build.\n" } } +# older CFLAGS declaration - if the default > optimization still causes problems +# we will revert to this +# > CFLAGS="-O3 -funroll-loops -fstrict-aliasing" +configure.ldflags- > append -lpango-1.0 +configure.cppflags-append -no-cpp-precomp - > DX_LOCALE +configure.cflags-append -funroll-loops -fstrict-aliasing > -configure.env LDFLAGS="-L${prefix}/lib -lpango-1.0" \ - > CPPFLAGS="-I${prefix}/include -no-cpp-precomp -DX_LOCALE" \ - > CFLAGS="-O3 -funroll-loops -fstrict-aliasing" - configure.args -- > disable-shm --disable-gtk-doc --with-included-loaders \ --mandir=$ > {prefix}/share/man > _______________________________________________ > macports-changes mailing list > macports-changes(a)lists.macosforge.org > http://lists.macosforge.org/mailman/listinfo/macports-changes Randall Wood rhwood(a)mac.com http://shyramblings.blogspot.com "The rules are simple: The ball is round. The game lasts 90 minutes. All the rest is just philosophy."

4 12

The mtree and man pages
by Randall Wood 14 Aug '07

14 Aug '07

I have noticed that almost 1840 ports have configure.args set to install man pages in ${prefix}/share/man instead of the default location ${prefix}/man and that the mtree-violation code does not like ${prefix}/man. Can we allow ${prefix}/man to be used and perhaps even set it to be the preferred location for man pages since it seems to fit more with the UNIX default location? On 13 Aug 2007, at 19:23, susanmackay(a)optusnet.com.au wrote: > I'm trying to install gnome-doc-utils and I get the following (just > the last output lines - all lines above appeared to be 'normal' > operation): > > /usr/bin/install -c -m 644 gnome-doc-xslt-C.omf /opt/local/var/ > macports/build/_opt_local_var_macports_sources_rsync.macports > .org_release_ports_gnome_gnome-doc-utils/work/destroot/opt/local/ > share/omf/g > nome-doc-xslt/gnome-doc-xslt-C.omf > make[3]: Nothing to be done for `install-exec-am'. > make[3]: Nothing to be done for `install-data-am'. > make[2]: Nothing to be done for `install-exec-am'. > test -z "/opt/local/share/gnome-doc-utils" || /bin/sh ./ > mkinstalldirs "/opt/local/var/macports/build/ > _opt_local_var_macports_sources_rsync.macport > s.org_release_ports_gnome_gnome-doc-utils/work/destroot/opt/local/ > share/gnom > e-doc-utils" > /usr/bin/install -c -m 644 'template.omf.in' > '/opt/local/var/macports/build/ > _opt_local_var_macports_sources_rsync.macport > s.org_release_ports_gnome_gnome-doc-utils/work/destroot/opt/local/ > share/gnom > e-doc-utils/template.omf.in' > /usr/bin/install -c -m 644 'template.make' > '/opt/local/var/macports/build/ > _opt_local_var_macports_sources_rsync.macport > s.org_release_ports_gnome_gnome-doc-utils/work/destroot/opt/local/ > share/gnom > e-doc-utils/template.make' > /usr/bin/install -c -m 644 'template-document.xml' > '/opt/local/var/macports/build/ > _opt_local_var_macports_sources_rsync.macport > s.org_release_ports_gnome_gnome-doc-utils/work/destroot/opt/local/ > share/gnom > e-doc-utils/template-document.xml' > DEBUG: Executing destroot_finish > DEBUG: checking for mtree violations > Error: violation by /opt/local/man > Error: Target org.macports.destroot returned: mtree violation! > Warning: the following items did not execute (for gnome-doc-utils): > org.macports.activate org.macports.destroot org.macports.install > Error: Status 1 encountered during processing. > Titania:~ susan$ sudo port -dv install gnome-doc-utils > > (The last line is there to show the command I used). > > I had recently ( as in about 1200UTC 13/8/2007) done a selfupdate > and a sync which included several new versions of some packages. > These were 'upgraded' beforehand. > > Is there something I have to do to get this to work? > > Thanks > > Susan > _______________________________________________ > macports-users mailing list > macports-users(a)lists.macosforge.org > http://lists.macosforge.org/mailman/listinfo/macports-users Randall Wood rhwood(a)mac.com http://shyramblings.blogspot.com "The rules are simple: The ball is round. The game lasts 90 minutes. All the rest is just philosophy."

3 2

mtree violations should be debug info, should not be fatal errors
by Ryan Schmidt 13 Aug '07

13 Aug '07

I'm not sure I like that mtree violations are fatal errors. The ports that are now failing to install because of mtree violations installed just fine in MacPorts 1.5.0. Why should they now fail to install? Their content has not changed. Sure, they may be installing things in places they shouldn't, but why should we make the user suffer? We have a -t switch which informs us about forgotten dependencies in the port -- but this does not issue a fatal error. It's just a message which portfile authors can use to improve their portfiles. Maybe mtree violations could be handled similarly. I'm also concerned about needing to specify in the portfile that the port intends to violate the mtree. For example, I'm going to have to add that to the php5 port, because it wants to install an apache2 module and the apache2 layout is considered nonstandard. So just because I want to install one item in a weird place, I have to turn off the mtree violation checks in the entire php5 portfile. It would be nicer if port would just issue nonfatal debug messages letting us know exactly which files were violating the mtree. This way I could assure myself that my port is installing the apache2 module in a weird place, yes, but that everything else is being installed in sane locations.

6 7

i386 vs i686
by Anders F Björklund 13 Aug '07

13 Aug '07

Certain platforms report their "machine" (or `uname -m`) as "i686" rather than "i386" like Darwin/FreeBSD does... This arch should probably be converted to "i386" as well, or else the variants and binaries won't match the other ? set arch $tcl_platform(machine) if {$arch == "Power Macintosh"} { set arch "powerpc" } if {$arch == "i686" || $arch == "x86"} { set arch "i386" } Occasionally someone tries to use the variant "intel" (which doesn't exist), so maybe that should somehow be made into yet another synonym for "i386" as well ? (interestingly, the darwin8 cross-compilers use i686) --anders

1 0

*.env not working
by Blair Zajac 12 Aug '07

12 Aug '07

The qt3-mac used to work with this: configure.env QMAKESPEC='' QTDIR='' DYLD_LIBRARY_PATH='' build.env QMAKESPEC='' QTDIR='' DYLD_LIBRARY_PATH="${worksrcpath}/lib" but now using port -d -v install qt3-mac, you can see that the environments are not passed to the configure and build steps. Can somebody who knows this part of the code take a look? Also, having a standard set up Portfile's that ensure that things work correctly would be a good thing. We seem to break things and only later find out that something has changed. In this case, the Portfile could set the env and the Makefile could ensure that the environment is indeed set. Another thing is to have a beefy build box that builds a large number of Portfile's before a new release and looks for any regressions in ports that used to build. Regards, Blair -- Blair Zajac, Ph.D. CTO, OrcaWare Technologies <blair(a)orcaware.com> Subversion training, consulting and support http://www.orcaware.com/svn/

2 2

mtree_violation?
by markd＠macports.org 12 Aug '07

12 Aug '07

I had to add an mtree_violation statement for my logrotate port I I don't understand why. Here is the destroot phase. destroot { xinstall -m 755 ${worksrcpath}/logrotate ${destroot}${prefix}/sbin xinstall -m 755 ${worksrcpath}/examples/logrotate-default ${destroot}${prefix}/etc/ xinstall -d ${destroot}${prefix}/man/man8 xinstall -m 755 ${worksrcpath}/logrotate.8 ${destroot}${prefix}/man/man8 } What is wrong with this? Mark

2 2

Re: [27673] trunk/dports/editors/abiword/Portfile
by Juan Manuel Palacios 12 Aug '07

12 Aug '07

On Aug 11, 2007, at 9:33 PM, source_changes(a)macosforge.org wrote: > Revision > 27673 > Author > nox(a)macports.org > Date > 2007-08-11 18:33:19 -0700 (Sat, 11 Aug 2007) > Log Message > > abiword: > * Removed superfluous configure.env variables. > * Added check to be sure that either use_binary or use_source is > enabled. > Modified Paths > > trunk/dports/editors/abiword/Portfile > > configure.env \ > - CPPFLAGS="-L${prefix}/lib -I${prefix}/include" \ > - LDFLAGS="-L${prefix}/lib" \ > CFLAGS="-no-cpp-precomp" Shouldn't this remaining old style CFLAG declaration be taken care of too? "configure.cflags-append no-cpp-precomp", 'cause otherwise (as it is now) you're overriding the default CFLAGS we provide out of the box. Or is that intentional? (we have seen cases of default CFLAGS braking some builds due to, for instance, optimization). Regards,... -jmpp

2 1

mtree violation check: no files matched glob pattern "*"
by Ryan Schmidt 12 Aug '07

12 Aug '07

I'd say installing into /Applications/MacPorts should not be considered an mtree violation. Check this out: $ sudo port install minivmac ---> Fetching minivmac ---> Verifying checksum(s) for minivmac ---> Extracting minivmac ---> Configuring minivmac ---> Building minivmac ---> Staging minivmac into destroot Error: Target org.macports.destroot returned: no files matched glob pattern "*" Error: Status 1 encountered during processing. $ That's a weird error message. Let's try it again. $ sudo port install minivmac ---> Staging minivmac into destroot Error: Target org.macports.destroot returned: no files matched glob pattern "*" Error: Status 1 encountered during processing. $ Ok. One more time. $ sudo port install minivmac ---> Staging minivmac into destroot Error: Target org.macports.destroot returned: error copying "/opt/ local/var/macports/build/ _Users_rschmidt_macports_dports_emulators_minivmac/work/projects/ bgcmachs/minivmac.app" to "/opt/local/var/macports/build/ _Users_rschmidt_macports_dports_emulators_minivmac/work/destroot/ Applications/MacPorts/Mini vMac.app/minivmac.app": file already exists Error: Status 1 encountered during processing. $ Whoa. Why do I now get a different error message? Let's start over and see if debug info helps. $ sudo port clean --work minivmac ---> Cleaning minivmac $ sudo port -d install minivmac [snip] DEBUG: Executing org.macports.destroot (minivmac) xinstall: mkdir /opt/local/var/macports/build/ _Users_rschmidt_macports_dports_emulators_minivmac/work/destroot/ Applications/MacPorts DEBUG: Executing destroot_finish DEBUG: checking for mtree violations Error: Target org.macports.destroot returned: no files matched glob pattern "*" Warning: the following items did not execute (for minivmac): org.macports.activate org.macports.destroot org.macports.install Error: Status 1 encountered during processing. $ Well, not exactly. Presumably the glob that's being tripped over is this one at line 233 in src/port1.0/portdestroot.tcl: # test files in ${prefix} foreach f [glob -directory "${destroot}${prefix}" *] { This port installs no files in ${prefix}; only one application into /Applications/MacPorts. There are many other such ports (see the aqua category for example); they must all be experiencing this error now.I think we will need to fix this quickly and release a 1.511 soon.

2 1

Re: gtk2 +quartz bug
by Randall Wood 12 Aug '07

12 Aug '07

On 12 Aug 2007, at 03:38, Davide Bertola wrote: > this bug : > http://paste.lisp.org/display/46006 > > has been solved here : > http://lists.macosforge.org/pipermail/macports-users/2007-March/ > 002213.html > > i tried with gtk+-2.10.14 but has not been fixed yet.. > it's just a line > maybe you can embed the patch in the portfile somehow Since GTK+ 2.12.0 is expected to be released later this month, I'm not going to fix this unless it is not fixed in that version. GTK+ 2.12 will include significant upgrades to the OS X support in GTK+. Randall Wood rhwood(a)mac.com http://shyramblings.blogspot.com "The rules are simple: The ball is round. The game lasts 90 minutes. All the rest is just philosophy."

1 0

Re: MacPorts 1.5.1 released (Needs a Trac ticket????)
by markd＠macports.org 11 Aug '07

11 Aug '07

Juan Manuel Palacios <jmpp(a)macports.org> writes: > This is a new feature in MacPorts 1.5.1 to detect violations of the >pre-established hierarchy that governs the ${prefix} tree, so as to >keep a sense of order when installing software. In order to minimize >conflicts with the bundled Apache, Apache 2 installs its entire world >onto its own prefix within MacPorts' ${prefix}, ${prefix}/apache2/ >{bin,lib,include,<etc>}, so that its files are not readily found by a >default configured shell; this setup violates the default hierarchy >and thus produces the error you saw when upgrading to MacPorts 1.5.1. >I just added and committed workaround to the Portfile to indicate the >violation is intended in this case, "destroot.violate_mtree yes". >Please re-sync your tree in about half an hour from this posting and >try installing again, report back to me if you still experiencing the >same problem. I may have missed a previous discussion, but can someone describe exactly what conditions violate the mtree guidelines? Mark

1 0