• PackagingWG
• 2020-02-19-pip

Bernard / Pradyun: PyCon UX Research planning

19th Feb 2020

Packaging Summit:

• Pradyun has a survey for people to suggest topics to discuss at packaging summit. These topics will then be discussed in seperate sessions. They will be attended mainly by package maintainers, but with some users also.

The purpose of these discussions will be to build consensus amongst people (tool maintainers, distro maintainers, potential design discussions etc) will be involved in design, implementation and possibly end users and progress towards a solution on the topic.

The objective of having these packaging summit discussions:

• Come up with solutions to the problems identified for discussion
  • These solutions will contain a number of assumptions on how the problem should be addressed (technical design, and implementation, also user workflows, expectations).
  • Some of these assumptions (mainly user related) will need to be validated with users i.e. the people who will end up using this software

Depending on how the summit goes, there are a number of ways this could happen.

16th April (Packaging Summit)

During the discussions about these topics:

1. We will take notes of what's discussed

• Action items:
  • identify assumptions that come up
  • identify questions the people in the room have and need to be answered

2. After the discussion ends

• Bernard will take these user-related research questions (what do you expect to happen, how/if do you expect to use this, Here's how I expect this to work for me and other questions that need to be identified)

• If the question isn't clear enough, the people who've attended the discussion can talk with Bernard. This can be used to define what research questions need to be asked at PyCon (over lunch, a coffee, unstructured time at the summit ("this is where Bernard sits!") etc)

17th-19th April (PyCon Conference Days)

3. We need to advertise the research to PyCon attendees.

Questions:

• How to do this?
  • Ask Sumana (Sumana responded on Zulip)
  • Options: do an open space/"UX table", twitter, put up posters/fliers, etc

4. Bernard interviews participants (users of packages and package maintainers) at PyCon using the quesions identified in step 2 above. The output will be user requirements/expectations that can be used to inform the work on the problem during the PyCon sprints (or after).

5. The interview then need to be analysed and some output needs the be produced.

Questions:

• What format does that output need to take?
• The research output needs to be in a format that the packaging maintainers can readily use in the sprints and later on
  • some information about the users interviewed, their background

  • GitHub issues

  • User stories
  • more?

20th-23rd April (PyCon Sprints) (partcipant recruitment: in teh early days there will be more "end-users" attending. Towards the end, the people left will be more invested - e.g. maintainers, long time contributors)

6. The package maintainers will take the output from step 5 and use it as input to their sprint work.

7. (optional) If there is a possilbilty to do some usability testing of early "prototypes" of the work from the sprints?

• The objective would be to find users at the sprints to test the prototypes with

8. Conduct interviews with package maintainers and end users about other topics too

(The sprints will be relaxed)

• python packaging on the whole
  • pip output messages
  • commonly used pip commands
  • if all pip commands went away what would be the 5 (X) you would want to keep?
• identify and talk to new 1) python and 2) packaging, contributors (roving reporter style)

9. Output

Question: Amongst whom does the consensus need to be built?

Getting consensus from people who are not in attendance at the summit, is also important - we might not be able to "finish" things at the sprints

Other questions:

• what can Bernard bring to PyCon that would be useful to talk about, present, etc

  • This would be work carried out before PyCon

  • survey
  • co-relating use cases with problems people have
    • e.g. people doing python packaging for education purposes - what are the problems you have
    • what are the packaging problems you have
    • what they use Python for
    • how Python packaging has gaps for them
    possible source of use cases
    • Existing surveys!

      • https://www.jetbrains.com/research/python-developers-survey-2018/

      • https://pyfound.blogspot.com/2019/10/the-2019-python-developer-survey-is.html

      • https://www.jetbrains.com/lp/devecosystem-2019/python/

    • what people are trying to do (with python + packaging tools) and what are their problems
    • what are the users we have
    • what their problems are