Framework team meeting minutes 2017-06-27

Attendees

Eric Steele
Eric Brehault
Ramon Navarro Bosch
Gil Forcada
Johannes Raggam

Topics

Python 3

Discuss about driving an effort similar to the Zope 4 sprints lead by gocept to both clean our dependencies and port them to python 3
The gist of it would be that the framework team lists all the tasks that need to be done in order to get us to Plone X with python 3 support. If done in a way that can be parallelized and there’s clear instructions/mentors overlooking it, it can be a theme for sprints and also train newbies.

Test management

Gil will contact Alexander about his PLIP regarding moving to TOX to manage Plone tests https://github.com/plone/Products.CMFPlone/issues/2072

#Next meeting

Next FWT meeting is in two weeks (2017-07-11).

Software engineering in my humble opinion is aproximately 90 % a matter of communication (ideally starting with some pragmatical requirements engineering). So sometimes it would be quite helpful to spend some more words and in order to allow some luxury even some charts, graphics etc.

I hope i am right about TOX: tox ???

Please do not feel offended about that, but I think, it would make live easier for people interested in Plone to start using it or contributing ...

that sound pretty interesting; I would love to have a similar approach for our add-ons; I will start talking about this on the local community email list.

can you elaborate? I read the whole introductory documentation and found almost no reason to use tox, besides running different versions of Python which, BTW, can be done in CI.

can someone show me how life could be easier on something like this?

https://github.com/collective/collective.cover/blob/master/.travis.yml#L16-L30

tox is a complement to Travis-CI, not a replacement. I don't have Travis-CI on my local machine, but I do have tox and pyenv to work with multiple versions of Python. This allows me to run whatever environment, tests, docs, PEP8, or anything else before committing and pushing to a remote repo.

See also my comment in the Products.CMFPlone issue tracker.

yes tox seems to be a good way to enable more possibilities in successful contribution and quality. As @stevepiercy said, it is a complement to Continuous Integration Servers, like Travis-CI, Jenkins or Gitlab-CI, not a replacement.

The only tool that could become unnecessary might be plone.recipe.code-analysis as this functionallity could be easily included into a tox environment.

I would suggest that I will prepare some documentation on the travel to the midsummer sprint in Finland next week and will discuss that there with the attending people of the Framework / documentation and training team in a small session.
@stevepiercy would you mind if i copy some of the Pyramid documentation about that, and like your comment on @gforcada documentation for removing bootstrap.py / virtualenv those parts to make it more streamlined?

@gforcada if you have any future question on that we could have a short hangout or chat.

Yes that's the TOX I were talking about. Sorry if I did not elaborate a lot about it in the meeting minutes but I was not familiar with this PLIP at the time of the meeting and I am not familiar with TOX either :slight_smile: .
As we are starting discussing with Alexander about it, I guess we will have more details after our next meeting.

@loechel please, be my guest!

Also feel free to stop into pyramid on IRC. The core contributors there (raydeo and x58) often know a lot more detail about this stuff than I.