2024-07-11 undate check-in

[rlskoeser]

Agenda

Goals

• Check in on open pull requests
• Decide about numpy integration
• Determine any priorities for next steps

Open PRs

• integrate numpy datetime64 for greater range of years #77 / Use numpy datetime64 instead of datetime.date #77 #84
• setup ci with treon for jupyter notebooks to ensure they continue to run as the codebase changes #76 / Check notebooks in CI check workflow #76 #83
• add overview and example usage to the readme so people can more easily understand the project #75 / Add example usage to the readme #75  #82
• Reorganize and improve documentation #81

Discussion

• Decision on numpy approach: 👍 / 👎 ? optional vs. required dependency?
• Any feedback or changes on documentation / example structure
• How important is django/database integration?
  • @ColeDCrawford and @rlskoeser discussed implementation specifics in undate development meeting 2024-06-04 #72 - what about higher level need/importance of this feature?
• check in on 1.0 milestone
  • next priority / sequencing?
• Scheduling / process
  • Propose we skip meeting in August (busy prepping for and attending DH2024).
  • September? "Regular" schedule?
  • Any feedback or adjustments we want to make on process?

[rlskoeser]

Maybe numpy is a smaller dependency than it used to be; 2.0 release in June. Given that, then it seems ok. Probably worth documenting but not require - test against both versions for compatibility. Don't pin it. if working with the older version is hard, carve into a separate issue to tackle that. Check how heavy it is and if 1.2 / 2.0 change breaks anything.

Value of django / db integration? Useful for many people who would want to use it in an app, especially in the DHy community - widest adoption route. Additional django integrations from S&co for template tags would be valuable.

python packaging: we can use pyproject.toml for everything now; let's standardize on that and cleanup the other files - #80

• tox? simpler to drop that and do a matrix build in github actions (both Cole and RSK do this on other projects)

we'll need to allow different input/output formatters eventually (may not be the case)

schedule: we can do something ad hoc at dh2024 if we want to put our heads together and work on undate; otherwise we'll try to get back to our regular first thursday schedule starting in september. This meeting is useful to get reviews done if nothing else !

(stickers/logo for undate?)

next steps:

• numpy date integration
• update packaging
• integrate edtf parsing as an undate parser/formatter; once it's integrated we can test and check off functionality (refer to tests in upcoming release of python-edtf Cole has been working on)

decisions:

• use numpy; clean up code in draft pr and document versions; test against new release and old
• keep plan to implement django db field & related utilities (queryset filter, date string methods, etc)
• schedule: first thursday of the month starting up again in September (September 5)