Paul Moore <[email protected]> wrote: > [1] I'm assuming that we don't have any cases where authors of > maintained packages hosted outside of PyPI refuse to set up an index > page. There's no technical reason why they should do so, but there > remains the possibility of non-technical issues that need to be > thrashed out, and consensus reached with the conclusions documented in > the PEP.
I think there are quite some cases, for example: https://pypi.python.org/pypi/bzr/2.6.0 I *suspect*, but I may be entirely wrong, that they are using PyPI as a catalog and don't particularly care about automatic downloads. Reasons could include that on Linux distros one would use the package manager to install bzr, and for Windows and OS X they have installers. Also, some companies do not like deep links to download material on their pages and get litigious about it. I think these pages should be left untouched. Under no circumstances should direct links be created automatically for ethical and maybe even legal reasons. My view is that the respective maintainers should just be informed that download tools will abandon crawling support. They should be able to keep the pages, but pip should just comment that the package needs to be installed manually and users should contact the package maintainer and not the pip team in order to change the situation. Stefan Krah _______________________________________________ Distutils-SIG maillist - [email protected] https://mail.python.org/mailman/listinfo/distutils-sig
