On Thu, May 5, 2016 at 6:37 PM, Robert Collins <robe...@robertcollins.net> wrote:
> > Thats good. It occurs to me that scientific builds may be univerally > broken because folk want to avoid easy-install and the high cost of > double builds of things. E.g. adding bootstrap_requires will let folk > fix things? > scientific builds are universally "broken" because their dependencies are often non-python (and oftne complex and hard to build) external libs. that is simply a problem that setuptools and pip were never designed to address. > But the main question is : why are these packages staying inaccurate? > Even in the absence of a systematic solution I'd expect bug reports > and pull requests to converge on good dependencies. > the problem is not that people haven't declared the proper dependencies in their packages -- it's because it's impossible to declare the proper dependencies in their packages... The solution is to do one of three things: 1) punt -- and let folks use Canopy, or conda, or third party collections of wheels like the Gohlke repository, or let folks build it themselves (a lot of packages have all kinds of custom code to go look in macports, or homebrew, or fink locations for libs on OS-X for instance. 2) build binary wheels that statically link in everything needed (or include the dlls in the package) 3) make binary wheels out of C libs, so that other packages can depend on them -- this is kind of a kludgy hack (abuse?) of pip/wheel but should be do-able. Some folks are making a valiant effort to do a combination of (2) and (3) for Windows, OS-X, and now many linux -- but it is too bad that the tools don't make it easy. And so far, all we really have are the core scipy stack (and pyQT5???) In short -- building fully supported binary packages is possible, but requires a fair bit of expertise and a valiant effort on the part of someone... -CHB -- Christopher Barker, Ph.D. Oceanographer Emergency Response Division NOAA/NOS/OR&R (206) 526-6959 voice 7600 Sand Point Way NE (206) 526-6329 fax Seattle, WA 98115 (206) 526-6317 main reception chris.bar...@noaa.gov
_______________________________________________ Distutils-SIG maillist - Distutils-SIG@python.org https://mail.python.org/mailman/listinfo/distutils-sig