I updated to Mojave and the error vanished. :)
_
-. .´ |∞∞∞∞
', ; |∞∞∞∞∞∞
˜˜ |∞∞∞∞∞∞∞∞∞ RdB
,., |∞∞∞∞∞∞
.' '. |∞∞∞∞
-' `’https://rdb.is On 22 novembre 2018 a 11:37:11, Ryan Schmidt (ryandes...@macports.org) scritto: On Nov 17, 2018, at 16:25, Ruben Di Battista wrote: > On 17 novembre 2018 a 03:24:40, Ryan Schmidt scritto: > >> On Nov 16, 2018, at 18:00, Ruben Di Battista wrote: >> >> > I tested again the build (with also rev-upgrade). No luck, same error. >> > Does anyone here can try to install: vtk +hdf5 +python27 +mpich to see if >> > it’s just a misconfiguration for me or it’s a general problem? >> >> Some file on your system that is being used by the vtk build is linked with >> /opt/local/lib/libhdf5.101.dylib. We need to identify that file and rebuild >> it so that it is linked with /opt/local/lib/libhdf5.103.dylib instead. >> >> That's what running "sudo port rev-upgrade" does, so I'm surprised that >> didn't fix the problem for you. >> >> Maybe rev-upgrade is failing, but isn't printing an error message. You could >> run it with debug output: "sudo port -d rev-upgrade". >> >> If the output doesn't look like an error is happening, then maybe the >> problem is that the file that is linked with >> /opt/local/lib/libhdf5.101.dylib is not in MacPorts, but the vtk build is >> finding it for some reason. (For example, maybe it is in /usr/local.) If so, >> you could use trace mode, which tells MacPorts to ignore any files not >> explicitly listed as dependencies: "sudo port clean vtk && sudo port -t >> install vtk +hdf5 +python27 +mpich". > > Thank you. I looked in all the computer and there is not any > libhdf5.101.dylib. Exactly. But some library or program on your computer is linked with that nonexistent file, which is what's causing the error message we're trying to solve. We need to identify what that file is. Did you try running "sudo port -d rev-upgrade"? Did any of its output mention libhdf5.101.dylib? > I also tried with the trace mode, but I got an error like “bin/ps operation > not permitted” before it reached the actual point where the previous error > was given. It's difficult to offer any recommendations for this without seeing more lines of the output. > Can I somehow overcome the sandboxing? I'm not sure. What makes you think the sandboxing is implicated in the problem?
signature.asc
Description: Message signed with OpenPGP using AMPGpg
