2025 Development Priorities

[MicahGale]

I realized that @tjlaboss and I have an informal plan for MontePy, and with new people coming to the community it would be helpful to put these ideas on paper.

Deadlines

• There is one impending deadline coming up: the M&C workshop at the end of April this year. This is tracked with this milestone. This requires:
  • Developing Workshop demo notebooks
  • Deploying version 1.0.0

Upcoming Milestones

• Release version 1.0.0

  • 1.0.0a1 is currently out for alpha-testing, and some bug reports are coming in.
  • Bugs are being found while making demo notebooks
  • Will likely be deployed after the workshop demos are ~80% complete

• PyOpenSci and JOSS reviews

  • We have had two reviews conducted as part of the PyOpenSci software peer-review process
  • We believe we have addressed all reviewer comments
  • Waiting on a reviewer who is very busy to confirm that we adequately addressed all comments.
  • We have been accepted by PyOpenSci
  • JOSS article is currently being finalized shortly after this

Long Term Priorities

This is where we need some feedback.

Feature Requests

Currently it seems like the priorities for new features are:

1. Tallies
2. source defintions
3. everything else

Code Improvements

This I am less clear on. My vague priorities are:

1. Performance
2. Functionalize and make type enforcement more DRY.
3. Implement Copy on Write?

Bugs

We probably need to go back through existing bug reports after 1.0.0 and double check our triage.

Note: This is all volunteer work so keep that in mind.

[MicahGale]

October Update

Since release 1.0 our efforts have focused more on code improvements over feature development.

Performance

1. Currently there is an open PR Implement concurrent parsing #743 implementing concurrent (multi-processed) parsing. This so far has significant bottle-necks that is given marginal improvement
2. Work is underway for Just-in-time parsing Add support for Just-In-Time parsing #529 on branch jit_parse. This requires a rewrite of all object's constructors, which is slow going. Once this JIT parsing is implemented this should remove a lot of the bottlenecks for concurrent parsing. Together these two features should create a great minor release.

Infrastructure.

In #691 a new decorator was added args_checked. For a decorated function the type (and sometimes its values) will be checked for the arguments every time it is called against the type hints provided. This will remove significant amounts of boiler plate (#674). This is being rolled out in alpha-test-checkvalue