GSoC WESTPA Dashboard — prototype progress monitor + question about ZMQ vs HDF5 tracking

[ShyamAlancode]

Hi,

I’m preparing to apply for the WESTPA dashboard project for GSoC.

I set up a WSL-based environment and ran tutorial 7.1 locally to understand the WESTPA workflow and the west.h5 structure. While exploring, I implemented a small prototype CLI tool that monitors simulation progress by reading iteration and segment status from west.h5. The tool uses HDF5 SWMR mode for safe concurrent reads and NumPy-based counting for efficiency.

Repository:
https://github.com/ShyamAlancode/westpa-progress-monitor

While reading the project description, I noticed the mention of ZMQ work managers. My current prototype is disk-based (polling the HDF5 file), so I wanted to ask:

For the intended dashboard, is the expectation to subscribe directly to ZMQ messages for lower-latency updates, or primarily to monitor the HDF5 output?

Thanks!

[jeremyleung521]

Hello. Welcome.

For the intended dashboard, is the expectation to subscribe directly to ZMQ messages for lower-latency updates, or primarily to monitor the HDF5 output?

It's up to you, really, to weigh the benefits and downsides and decide. That's what we want to see in the pre-proposal/proposals.

Do note that Google doesn't like people start coding early for GSoC (at least not until after acceptance/official start date: https://developers.google.com/open-source/gsoc/timeline#may_25) so I would tread carefully about prototypes and what not.

EDIT: (also this: https://github.com/MDAnalysis/mdanalysis/wiki/GSoC-FAQ#should-i-start-a-pr-for-a-project-that-i-am-writing-my-proposal-for)

[ShyamAlancode]

Thanks for the clarification — that makes sense.

I’ll focus on working through the design trade-offs between file-based monitoring and ZMQ messaging and reflect that in the proposal. While experimenting locally helped me understand the workflow, I’ll treat it only as exploration and concentrate on the architecture and planning side moving forward.

Appreciate the guidance!