RESTX Rename

[SteadBytes]

Addresses #1

TODO:

• Rename references to Flask-RESTPlus, flask-restplus e.t.c to Flask-RESTX, flask-restx e.t.c
• Remove Flask-RESTPlus logo
• Add new Flask-RESTX logo
• Remove AUTHORS.rst
• Remove CHANGELOG.rst
• Update LICENCE
  • Remove Akamai
  • Add another Modifed work Copyright line (see questions)
• Update setup.py author and author_email (see questions)

Questions:

• Which name should go into setup.py as the author and into the licence copyright? Are the other maintainers happy for my name to be used here (as I'm making these changes)?

[apryor6]

Woo hoo for moving forward!

[andreykurilin]

Which name should go into setup.py as the author and into the licence copyright? Are the other maintainers happy for my name to be used here (as I'm making these changes)?

python-restx team?

[ziirish]

Agreed, "python-restx team" is good as author as from now on, this is a community driven project.

[SteadBytes]

Sounds good to me! I just want to check if "python-restx team" is valid as the copyright line in the licence i.e. Modified work Copyright (c) 2020 python-restx team i.e. do we need to be some sort of legal entity? Is anyone (more knowledgeable of licencing minutia than myself) able to comment on this?

Edit: I've found an example from the chromium project https://github.com/chromium/dom-distiller/blob/d16a68c1b885f679ae1266a9eeee66e69c12e67c/javatests/org/chromium/distiller/OpenGraphProtocolParserAccessorTest.java#L1

Should we follow suit with python-restx Authors ?

[SteadBytes]

We also need an author_email for setup.py - any suggestions?

[andreykurilin]

@SteadBytes actually, only 3 fields are required: name, version, and packages. We can just drop author_email.

Also, based on first clause of BSD license I'm not sure that we can remove Akamai copyright: Redistributions of source code must retain the above copyright notice. I guess the only available "workaround" is to revert an original commit that adds copyright

[SteadBytes]

@andreykurilin author_email is required if author is specified, so we either have both or neither.

I defer to @j5awry on the Akamai licence bit, but he stated that he never made any copyrightable commits to the code whilst working at Akamai and won't be moving forward either.

[andreykurilin]

@SteadBytes

author_email is required if author is specified, so we either have both or neither.

oops, I did not know this. But after googling, it looks like not true anymore - python/cpython#17388 (setuptools->distutils->cpython). I guess we can specify invalid email for now (if there are no other options) and remove it as soon as new python packages will be available.

I defer to @j5awry on the Akamai licence bit, but he stated that he never made any copyrightable commits to the code whilst working at Akamai and won't be moving forward either.

yeah, but technically current change looks like 'forking with loss of one of the original authors'. At least, it is better to do it separately.

[SteadBytes]

@andreykurilin
Ah yeah, that's change is quite recent - on my Python 3.7 install setup.py check throws a warning:

python3 setup.py check
/usr/lib/python3.7/distutils/dist.py:274: UserWarning: Unknown distribution option: 'dev_require'
  warnings.warn(msg)
/usr/lib/python3/dist-packages/setuptools/dist.py:474: UserWarning: Normalizing '0.13.1.dev' to '0.13.1.dev0'
  normalized_version,
running check
warning: check: missing meta-data: if 'author' supplied, 'author_email' must be supplied too

As long as it doesn't affect building the project (i.e. it's only affect is showing that warning) then I'm happy to remove it - good spot thankyou!

I see your point, I will wait for @j5awry to confirm how to proceed with this it was originally introduced by/for him. The commit message has a clear explanation of why the change is taking place so as long as the explanation is actually valid then I think we're ok. That said, I'll make sure to have confirmation from @j5awry before proceeding further 👍

[andreykurilin]

LGTM :) Thanks for pushing this forward!

[j5awry]

Legally, it's fine. There were 0 commits done with Akamai attached to it. Akamai normally only cared about commit level (if i made a commit, add in "akamai copyright"). They normally don't do a top level.

In fact, doing a giant top level with a bunch of entries is highly irregular.

[j5awry]

Essentially, Akamai can't claim copyright when no copyrightable work by an Akamai employee is done. If you'd all rather, i can push the specific commit and have it in the commit message.

[andreykurilin]

I do not pretend to be a legal expert and I guess Akamai is not offended here, but it would be nice to save all this context into the history of the project. That is why I raised this question.

Please accept my apologies, @SteadBytes, I used to review PRs on GitHub where people forgot to add proper commit messages to all commits. Thanks to add a good commit message for a change to LICENCE!

[ziirish]

The tests are failing due to the lack of author_email:

warning: Check: missing meta-data: if 'author' supplied, 'author_email' must be supplied too

I tried locally with a dummy author_email and the tests succeed.

[andreykurilin]

@ziirish I guess we can modify https://github.com/python-restx/flask-restx/blob/master/tasks.py#L160 with warn=False to avoid CI failures. Another thing, as you mentioned, is setting author_email, but there is no team specific email yet...

[SteadBytes]

Thanks for clarification @j5awry - I'll leave the licence changes in 👍 No need to apologise @andreykurilin, one cannot be too careful when it comes to licencing!

Since the warning for author_email has been officially removed as @andreykurilin helpfully pointed out, I'm happy to adjust https://github.com/python-restx/flask-restx/blob/master/tasks.py#L160 to either: remove it entirely or make it output the warning but not fail the build. It's also somewhat confusingly named readme_check when it's testing setup.py 🤔

Of course the other option is to include the email, I'm not in favour of using a placeholder email. I see several options:

1. Use one of our own email addresses
2. Create a team email account (i.e. a Gmail account python-restx@gmail.com) and share access amongst the core team
3. Similar to point 2 but done 'properly' using a domain (i.e. G Suite)
   • Major downside for this is cost, hence I'm not in favour

[ziirish]

I'm on mobility atm, but I think I tested the underlying command `python setup.py test` and it doesn't return 0So I'm not sure patching the task as pointed earlier will resolve anything.There is another warning though. It suggests that `python setup.py test` has been deprecated in favor of `twine test dist/*`I tried that command and it doesn't complain about the lack of email. But this requires us to build at least the sdist first.Now I still think the name/description is valid because the purpose of this test is to check whether the documentation is renderable on Pypi.

[SteadBytes]

Ok awesome thanks @ziirish , looks like no author_email will be fine 👍

[ziirish]

LGTM

[SteadBytes]

Brilliant, thanks for adding the logo @ziirish 😄 I'll give it one more check over, fixup the commits and then merge 👍