[MacPorts] #52212: hatari update to 1.9

MacPorts noreply at macports.org
Mon Sep 12 03:39:23 CEST 2016


#52212: hatari update to 1.9
-------------------------------------+--------------------------------
  Reporter:  kenneth.f.cunningham@…  |      Owner:  macports-tickets@…
      Type:  update                  |     Status:  new
  Priority:  Normal                  |  Milestone:
 Component:  ports                   |    Version:  2.3.4
Resolution:                          |   Keywords:
      Port:  hatari                  |
-------------------------------------+--------------------------------

Comment (by kenneth.f.cunningham@…):

 Thanks, Lawrence,

 I finally concluded (as per the mailing list discussions and with your
 help and that of others) that the cmake build script (in all versions of
 hatari, including the existing 1.7 version on macports now) has skipped
 the part about prefacing python calls with
 ${PYTHON_EXECUTABLE)/path/to/script.py. They just call /path/to/script.py,
 and let 'shebang' figure it out with the internal shell setup cmake uses.

 And because cmake doesn't inhale environment variables, and there seems to
 be no way to set the shell call environment variables that the cmake shell
 will honour that I can discover after several days of looking, there seems
 to be no way to modify the shebang python target cmake calls. I could have
 monkeyed with the path and forced it to python27 as we discussed, but this
 seemed less desirable as per our previous discussion. The proper way is to
 rewrite the cmake build script with
 ${PYTHON_EXECUTABLE)/path/to/script.py

 Upstream will someday, maybe, fix the python calls in the build script. Or
 perhaps I might fix it for them at some point, for their next release
 (whenever that might be). I have a couple of cmake reference tutorials on
 the go just now, but like Ryan pointed out, it's a bit of challenge
 getting the hang of it.

 In the meantime, the fix here will makes the existing python script
 compatible with all known python versions back to Tiger, and probably
 robust for at least all python 3.x versions that might show up, and it's
 certainly better than the existing port's python script that fails if the
 user has python 3.x selected.

 So I guess it's kind of a question of "better" being acceptable, vs
 "perfect".

 Hatari 1.9 is the last released version, and I'm not aware at present of
 an imminent new release that might fix this.

-- 
Ticket URL: <https://trac.macports.org/ticket/52212#comment:2>
MacPorts <https://www.macports.org/>
Ports system for macOS


More information about the macports-tickets mailing list