On 08/19/2011 04:32 PM, Jakub Wilk wrote: >> to resolve "libassimp.so.2" dynamically. > > But why? The Python code has hardcoded assumptions about the library > ABI, so you should open libassimp.so.2, not any other libassimp.so.N > (which ABI might be different). thanks, sounds like a good idea. > >> since i am no python programmer, i wonder whether this is portable >> (w32, osx) > > AFAICT ctypes.util.find_library is implemented for all platform, though > using it is almost always bad idea. > >>>>> Also, why did it print stuff on stdout? :/ >>>> ask upstream :-) >>> ... >> no, seriously. > > Well, I have no interest in your package, I don't even know what it's > all about. I was just (implicitly) asked by Michael to review the Python > part of it, so I did. > i appreciate your review and think it was helpful. even more so, if you are not interested in the package per se. i still don't know how to deal with problems like the "print to stdout" in general. for this package it seems quite easy, as it is pretty clear that those printouts are there mainly for debugging reasons and have been left over. but is there a general guideline for fixing not-yet-bugs? e.g. if i have an commandline app that uses flags like "-help" (rather than "-h" or "--help"), this could be considered unexpected behaviour by many people. should those be fixed? fixing might break a lot of peoples scripts... mgfasd IOhannes
Attachment:
signature.asc
Description: OpenPGP digital signature