Mailman 3 May 2020 - pytest-dev

A Short Personal Post Morten for the pytest exodus
by Ronny Pfannschmidt May 20, 2020

May 20, 2020

Hello everyone, ** This was originally intended to be send out mid April but got intentionally delayed to ensure Holger could do the communication for the resolution, i finally picked it up again to send it out ** After failing a number of times to sit down and actually throw together a reasonable post mortem, I decided to at least try to write up a basic summary. Before continuing please take a look at Adrin Jalali: Open Source - CoC - Conflicts <https://adrin.info/open-source-coc-conflicts.html> His writing gives an outside perspective on the issues at hand. Overall I don't want to attribute malice to Daniel, however we have built up a sad story about failed communications and a failed process. Early in the process just a bit after the first CoC Report on Daniel, I had a phone call with him and exited it under the impression that things are better now, the next morning I woke up to a complete “annihilation” of all his issues/prs (rage closed?). At that point in time I reached the frustrating conclusion that I couldn't get through to him, and as far as I know we never got a satisfying explanation for that. That event also changed my perception of him as my trust was gone. In the weeks and months that followed the behaviour stayed the same and Bruno and I tried to get the CoC to act, as small little details that border on CoC violations and just drain our will to work on things kept accumulating. At the same time I also didn't want to push for intense action, as I consider myself a “affected party” and would like to see someone with a more clear head on things cut out some clear decisions. What I perceived then was the helpless situation that 1. I couldn’t give Daniel the direly needed feedback in a sensible manner 2. The CoC Team also didn't. So the Situation kept “giving” - to the point where I would sometimes fail to communicate to Daniel the way I would like to as the frustration kept getting to me.I then quit the CoC team in frustration, but only with a relatively short unhappy note. Some more weeks passed in which Bruno and I worked with each other to have a little outlet. However when Bruno finally called quits, that setup was gone, I hesitated a while and then decided to join the exodus. As far as i can tell in retrospect, there are many points where i see potential “what if’s”, However when things happened I didn't see those. My own inexperience at CoC enforcement plays a role, given the current situation I sure wish I had acted earlier and more. Also I tried to be nice even though the situation begged for a more direct and honest approach, after all bad behaviour needs early and effective feedback to correct and to be clear for all sides. With my departure from the CoC team I felt that I shed myself of that part of the responsibility (as I didnt perceive an effect) , but it certainly is my responsibility to protect myself from bad behavior. While I would prefer that this was done in an effective and positive way, I recognize that shortcomings in the communication between Daniel and me, The CoC team and me as well as my inexperience with situations of that kind, have contributed towards the current outcome. After I sent out the short message about quitting as well, I underwent a number of emotions, but in the end I do feel a great relief, I have shed myself of a role that became negative to me. Now I have quite some more emotions to unpack and reorient myself. I want to be part of a CoC-Team again - but next time I want to actually be trained in the process,I do believe that CoC Teams are an important tool to keep communities working,but as Adrin wrote, one has to hone a particular set of skills to do it decisively and quickly. I believe part of why the CoC process with pytest went so frustrating in the case at hand is that all 4 of the initial members are trying to be nice people, and I'm very sure the other 3 are way nicer than I am. However that particular trait hinders a certain positive harshness and in a way paves a road to hell with good intentions. It’s my impression that we all missed our opportunities to sort this out more favourable for everyone involved. I also want to continue to contribute to the python testing ecosystem in one way or another,After all I had myself lined up for a multi-year process of refactorings and minor breaking changes of pytest internals to get node structures, fixtures and test execution in general ready for the things to come. What I see myself doing in the near future is something like minimal modular composable testing components. I also can see myself returning to pytest, but that's not yet something for which I have a concrete idea. For now I want to reinvent myself a bit and see where that leads to. -- Ronny

1 0

Manually running test steps/getting user input
by Brianna Laugher May 16, 2020

May 16, 2020

I work in an environment where there is a lot of manual text execution (blame hardware). Moving people to ATDD/BDD style tests (such as Gherkin) is a first step. The good thing about tests written in this style is it works for both automated and manual execution. We are looking at adopting such a framework to start with some system level testing for embedded software. Robot Framework is well known and solid in this space. One of my colleagues pointed out that RF even has a capability to prompt the user for input or to record a step as pass/fail. https://robotframework.org/robotframework/latest/libraries/Dialogs.html (I've just tried it out and it is an extremely basic Tk dialog that pops up from the command line on Windows, but it is kinda nifty.) I am loath to discover that RF has a feature pytest does not... This is kind of a neat tool to have on a path towards automation. Of course the steps run all the time need to be automated, but there are plenty of others that could stay manual, being able to do this is an interesting hybrid. (Obviously these are not the kind of tests you run in CI!) So has anyone heard of a plugin like that for pytest? The RF code looks like it might not take that much porting, tbh... https://github.com/robotframework/robotframework/blob/master/src/robot/libr… Any thoughts about this as an idea, somewhere between awesome and cursed? :) cheers Brianna -- They've just been waiting in a mountain for the right moment: http://modernthings.org/

2 1

pytest 5.4.2
by Bruno Oliveira May 8, 2020

May 8, 2020

pytest 5.4.2 has just been released to PyPI. This is a bug-fix release, being a drop-in replacement. To upgrade:: pip install --upgrade pytest The full changelog is available at https://docs.pytest.org/en/latest/changelog.html. Thanks to all who contributed to this release, among them: * Anthony Sottile * Bruno Oliveira * Daniel Hahler * Ran Benita * Ronny Pfannschmidt Happy testing, The pytest Development Team

1 0

Resolution of recent conflict
by Brianna Laugher May 7, 2020

May 7, 2020

Many of you know that there has recently been some conflict within the pytest core maintainer group. Thanks all for your patience while discussions took place in the background. In the end, we as active maintainers agreed to exclude Daniel from pytest project membership for at least six months, because of interactions which we deemed innappropriate and harmful to the project. After six months we will be ready to talk again. Daniel has been a prolific and significant contributor to pytest, and we recognise and deeply appreciate the time he has devoted to the project, and thank him for his efforts. We thank everyone for allowing some time and space to the resolve the situation, in particular to Holger who discussed with Daniel and everybody else in the background. We ask that you resist the temptation to create a hero/villain narrative, because it is both not correct and not helpful. This is also the reason we don't cite his full name as we don't want to put public blame or burn bridges that we'd like to remain walkable in the future. Thanks, Anthony, Brianna, Bruno, Florian, Floris, Ran, Ronny.

2 1

Pytest-stress Maintainers
by Imran M May 1, 2020

May 1, 2020

Hope everyone is enjoying quarantine 😷. Would any of the pytest-stress maintainers check out the issues page when they get a chance please? https://github.com/pytest-dev/pytest-metadata/issues I'd like to contribute by adding an enhancement but haven't gotten a response for a few weeks. I'd also be happy to become a maintainer if that's possible? Looks like a new release in needed. Take care, and keep washing those hands 🧼. -Imran

3 3