[Distutils] who is BDFL for the boostrap/requires declaration? (was: moving things forward)

[Nick Coghlan]

On 8 May 2016 at 09:23, Brett Cannon <brett at python.org> wrote:
> On Sat, May 7, 2016, 12:16 Chris Barker <chris.barker at noaa.gov> wrote:
>> how is this about "bootstrapping" -- to me, bootstrapping is when you need
>> X to build X. Isn't this just regular old configuration: you need x,y to
>> build z?
>
> Sure, if you don't like the term "bootstrap" then you can call it "build
> requirements". We have not been calling it " configuration" in a general
> sense as this doesn't cover how to invoke the build step (that will probably
> be the next PEP), just what needs to be installed to even potentially do a
> build.

The reason I think "bootstrap" is a better name at this point than
"build" is that there are actually three commands we're installing the
relevant dependencies for:

* egg_info/dist_info (i.e. metadata generation)
* sdist (i.e. archive generation)
* bdist_wheel (i.e. building)

The bootstrapping at the moment is taken care of by "assume everything
uses setuptools, install setuptools by default, if you want to use
something else, use setuptools to define and retrieve it".

The new metadata aims to take the place of setuptools in that
bootstrapping process: if the software publisher so chooses, they'll
be able to both create an sdist from a source tree and a wheel archive
from an sdist without ever installing setuptools. (Of course, one or
more of their dependencies are likely to bring in setuptools anyway
for the foreseeable future, but at the level of their own project
they'll be able to ignore it)

Cheers,
Nick.

-- 
Nick Coghlan   |   ncoghlan at gmail.com   |   Brisbane, Australia