January 2016 - macports-tickets - lists.macosforge.org

[MacPorts] #50233: port upgrade --force shouldn't uninstall if no permission was given to install a missing dependency
by MacPorts 06 Jan '16

06 Jan '16

#50233: port upgrade --force shouldn't uninstall if no permission was given to install a missing dependency -------------------------+-------------------------------- Reporter: rjvbertin@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: base | Version: 2.3.4 Keywords: | Port: -------------------------+-------------------------------- This is a rare situation, that gave rise to a port being uninstalled in error, I consider: Given: a port "A" that has a build dependency on port "B1" which itself conflicts with port "B2". Once built, port "A" can co-exist with either "B1" or "B2"; in fact, "A" has a path-style *runtime* dependency on an executable provided by both B1 and B2 (path:bin/bb). Scenario (leaving out sudo, and ignoring the reasons why I proceeded this way): ``` > port deactivate B2 > port activate B1 > port destroot A > port deactivate B1 > port activate B2 > port -nv upgrade --force A ---> Computing dependencies for A.. The following dependencies will be installed: B1 Continue? [Y/n]: n ---> Deactivating A @version ---> Uninstalling A @version ---> Cleaning A ---> Removing work directory for A ---> Computing dependencies for A.. The following dependencies will be installed: B1 Continue? [Y/n]: n ---> Scanning binaries for linking errors ``` I consider that the process should have been aborted when I first refused to install B1. One could also consider that - the depends_build dependency should be ignored since the build process has been completed - the depends_run dependency is satisfied by the presence of ${prefix}/bin/bb (provided by B2 at the time). So in a sense there should be no missing dependency when doing an upgrade . -- Ticket URL: <https://trac.macports.org/ticket/50233> MacPorts <https://www.macports.org/> Ports system for OS X

1 1

Re: [MacPorts] #42220: py-netcdf4 @1.0.4_1: build failure when using ccache
by MacPorts 05 Jan '16

05 Jan '16

#42220: py-netcdf4 @1.0.4_1: build failure when using ccache -------------------------+-------------------- Reporter: egall@… | Owner: sean@… Type: defect | Status: closed Priority: Normal | Milestone: Component: ports | Version: 2.2.1 Resolution: worksforme | Keywords: Port: py-netcdf4 | -------------------------+-------------------- Changes (by sean@…): * status: new => closed * resolution: => worksforme Comment: I just tried to recreate this locally but even with ccache everything seemed to work. I'll still work on the mpi stuff, just to get it inline with netcdf4. -- Ticket URL: <https://trac.macports.org/ticket/42220#comment:19> MacPorts <https://www.macports.org/> Ports system for OS X

1 0

[MacPorts] #50221: Cannot install control package in Octave 3.8.2 on Mac OSX 10.11.2
by MacPorts 05 Jan '16

05 Jan '16

#50221: Cannot install control package in Octave 3.8.2 on Mac OSX 10.11.2 -------------------------+-------------------------------- Reporter: nico.mack@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: -------------------------+-------------------------------- Attempt to (clean) install control package in octave 3.8.2 fails on OSX 10.11.2 (El Capitan) due to a link error saying: {{{ :info:configure ld: warning: could not create compact unwind for _dhseqr_: stack subq instruction is too different from dwarf stack size :info:configure ld: warning: could not create compact unwind for _ATL_dJIK0x0x66NN0x0x0_aX_bX: stack subq instruction is too different from dwarf stack size :info:configure Undefined symbols for architecture x86_64: }}} I've attached the log file -- Ticket URL: <https://trac.macports.org/ticket/50221> MacPorts <https://www.macports.org/> Ports system for OS X

1 3

[MacPorts] #50231: emacs-app does not emulate three-button mouse
by MacPorts 05 Jan '16

05 Jan '16

#50231: emacs-app does not emulate three-button mouse --------------------------+-------------------------------- Reporter: duanhuaiyu@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: emacs-app --------------------------+-------------------------------- In a previous version of emacs-app I could use a two-finger click on trackpad as mouse-2, e.g. for auto correction in flyspell. It does not work anymore on emacs-app(a)24.5 on OS 10.11.2. -- Ticket URL: <https://trac.macports.org/ticket/50231> MacPorts <https://www.macports.org/> Ports system for OS X

1 1

[MacPorts] #49747: Update py-pymc to 2.3.6, correct GitHub PortGroup add gcc5 (as default) and py35 variants
by MacPorts 05 Jan '16

05 Jan '16

#49747: Update py-pymc to 2.3.6, correct GitHub PortGroup add gcc5 (as default) and py35 variants -----------------------------+-------------------------------- Reporter: russell.jones@… | Owner: macports-tickets@… Type: update | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: py-pymc -----------------------------+-------------------------------- Portfile and diff attached. All tests pass in python 3.4 and 3.5 (well, it "fails" with a set of `DeprecationWarning`s, but I think that's not a problem-- or at least one to be fixed upstream), but one fails in 2.7 as below. I'll check if the 2.7 failure is an upstream bug. I'm also rebuilding py27-numpy and scipy gcc5 to see if that's the cause. Perhaps this should be enforced or warned about by the port if it is the problem? {{{ FAIL: test_consistency (pymc.tests.test_distributions.test_skew_normal) ---------------------------------------------------------------------- Traceback (most recent call last): File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7 /site-packages/pymc/tests/test_distributions.py", line 1075, in test_consistency assert_array_almost_equal(hist, like, 1) File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7 /site-packages/numpy/testing/utils.py", line 886, in assert_array_almost_equal precision=decimal) File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7 /site-packages/numpy/testing/utils.py", line 708, in assert_array_compare raise AssertionError(msg) AssertionError: Arrays are not almost equal to 1 decimals (mismatch 10.0%) x: array([ 0. , 0. , 0.1, 0.3, 0.6, 1. , 1.7, 2.1, 1.9, 0.3]) y: array([ 0. , 0. , 0.1, 0.3, 0.6, 1.1, 1.7, 2.2, 1.7, 0.3]) }}} (The 3.5 deprecation warnings are all as below) {{{ DeprecationWarning (inspect.getargspec() is deprecated, use inspect.signature() instead) }}} -- Ticket URL: <https://trac.macports.org/ticket/49747> MacPorts <https://www.macports.org/> Ports system for OS X

1 4

[MacPorts] #50227: port select --set ruby ruby23 fails on rake symlink creation
by MacPorts 05 Jan '16

05 Jan '16

#50227: port select --set ruby ruby23 fails on rake symlink creation ---------------------+-------------------------------- Reporter: allen@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: ruby23 ---------------------+-------------------------------- allen@badapple ~ $ sudo port select --list ruby Available versions for ruby: none ruby21 ruby22 (active) ruby23 allen@badapple ~ $ sudo port select --set ruby ruby23 Selecting 'ruby23' for 'ruby' failed: could not create new link "/opt/local/share/man/man1/rake.1.gz": target "/opt/local/share/man/man1/rake2.3.1.gz" doesn't exist -- Ticket URL: <https://trac.macports.org/ticket/50227> MacPorts <https://www.macports.org/> Ports system for OS X

1 2

[MacPorts] #50216: Migration failed on El Capitan
by MacPorts 05 Jan '16

05 Jan '16

#50216: Migration failed on El Capitan ---------------------+-------------------------------- Reporter: daryl@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: ---------------------+-------------------------------- Following the procedure at https://trac.macports.org/wiki/Migration I, perhaps with too much trust, followed the instructions explicitly, arriving at the "put it all back step" : sudo ./restore_ports.tcl myports.txt. I got three lines of output: Warning: Skipping Can't (not in the ports tree) Warning: Skipping Please (not in the ports tree) Warning: Skipping To (not in the ports tree) Smelling a rat, I had a look at myports.txt (which I should have done immediately on creating it, I now realize) and here it is: Can't map the URL 'file://.' to a port description file ("Could not find Portfile in /Users/daryl"). Please verify that the directory and portfile syntax are correct. To use the current port, you must be in a port's directory. Now when I run $ port -qv installed I get nothing. Being of a philosophical sort, I suppose I'll take that as a sign that this was a good time to clean out the attic and install only what I want now. But before going down that path, do you suppose restoring from my TimeMachine backup would succeed, or would that be just too complicated? -- Ticket URL: <https://trac.macports.org/ticket/50216> MacPorts <https://www.macports.org/> Ports system for OS X

1 1

[MacPorts] #49777: julia @0.4.1 should use libgit2 from MacPorts
by MacPorts 05 Jan '16

05 Jan '16

#49777: julia @0.4.1 should use libgit2 from MacPorts -------------------------+-------------------------------- Reporter: braumann@… | Owner: macports-tickets@… Type: enhancement | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: libgit2 | Port: julia -------------------------+-------------------------------- As {{{julia @0.4.1}}} uses its own {{{libgit2}}} without openssl usage but instead with SecureTransport, building {{{libgit2}}} during julia build fails on systems less than 10.8, as e.g. some protocols like TLS 1.1 and 1.2 were not yet implemented e.g. in 10.6 (as far as I can see) inside SecureTransport. However, using {{{libgit2}}} with openssl support (just from MacPorts) should help enable {{{julia}}} builds on older MacOSX versions. BTW, I also wonder why {{{julia}}} apparently requires own {{{openblas}}}, {{{libuv}}} and {{{libunwind}}} libraries. -- Ticket URL: <https://trac.macports.org/ticket/49777> MacPorts <https://www.macports.org/> Ports system for OS X

1 3

[MacPorts] #50076: pdfgrep @1.4.1 Build error
by MacPorts 05 Jan '16

05 Jan '16

#50076: pdfgrep @1.4.1 Build error -------------------------------------------------+------------------------- Reporter: franky.cyber.ethics@… | Owner: macports- Type: defect | tickets@… Priority: Normal | Status: new Component: ports | Milestone: Keywords: unique_ptr mountain lion snow | Version: 2.3.4 leopard | Port: pdfgrep -------------------------------------------------+------------------------- On my system with OS X 10.8.4 (and presumably any other pre-10.9 system), building the latest version of pdfgrep gives compile errors, due to C++11 library incompatibilies. Please see the attached log for more information. Had to revert to older Portfile to install version 1.3.2. -- Ticket URL: <https://trac.macports.org/ticket/50076> MacPorts <https://www.macports.org/> Ports system for OS X

1 4

[MacPorts] #50184: MacVim @7.4 Update to snapshot 88
by MacPorts 05 Jan '16

05 Jan '16

#50184: MacVim @7.4 Update to snapshot 88 ------------------------+-------------------------------- Reporter: joelfred@… | Owner: macports-tickets@… Type: update | Status: new Priority: Normal | Milestone: Component: ports | Version: 2.3.4 Keywords: | Port: MacVim ------------------------+-------------------------------- Updates snapshot to 88. Recommend removing patch-src-MacVim-icons- python.diff as the port no longer needs this patch. -- Ticket URL: <https://trac.macports.org/ticket/50184> MacPorts <https://www.macports.org/> Ports system for OS X

1 2