This is the discussion repo for the first online QuPath hackathon, 26-29 April 2022.
The goal of the hackathon is to bring together people who are interested in QuPath from a programmer's perspective.
Be warned: it'll be technical!
Most QuPath users who don't like coding aren't likely to find the discussions very interesting, useful or comprehensible. Sorry.
This is to try to rebalance the training a bit. QuPath's 'normal' use is already described in the official docs, YouTube channel and discussion forum - as well as fabulous user-created resources such as Mike Nelson's website, the La Jolla Institute workshop and the recent SciLifeLab webinar.
However, programmers who want to get the most out of QuPath don't have a lot. They can browse the source code or the javadocs, but otherwise an excessive proportion of QuPath's technical documentation exists only in my leaky brain.
That's neither healthy for me nor for the project. It's hard to keep on top of answering all the developer questions that come my way while always being on the lookout for buses.
This hackathon is intended to help more people understand QuPath at a technical level - so they can use it to do and develop new things.
The exciting thing (for me) about QuPath being open-source isn't just that it's free to use, but that it can be reused and recombined to help make even more interesting and useful open tools.
Some examples already include:
- Warpy & ImageCombinerWarpy for whole slide image registration
- paquo for combining QuPath + Python
- clupath as a bridge between QuPath and clij for GPU-powered processing
- QuPath Edu for using QuPath in teaching
- QuPath Extension OpenVINO for using QuPath with Intel OpenVINO for Java
- QuPath Extension CellPose to bring CellPose segmentation to QuPath
- NoCodeSeg for using QuPath with DeepMIB and FastPathology
The goal of the hackathon is to make more work like this possible - and to have a larger group of people understanding QuPath's code, so we can share ideas about its future development.
It depends who joins!
It should be informal.
At a most basic level, the hackathon will be me hanging around image.sc island for 4 days from ~9am-5pm BST, hoping I can make myself useful.
But if others join then there can also be networking and discussion. I'll try to answer questions about QuPath's design, and we can talk about potential improvements - however the main focus will be on helping one another create or improve new QuPath extensions and scripts.
With that in mind, if you want to participate, you should already have an idea what you want to do.
For example, this could be:
- Integrate QuPath with some other tool
- Add a new feature as an extension
- Address some specific annoyance in the software
- Streamline analysis in a complex script
- Import/export data in a new way
- Create scripts of general interest, e.g. to validate classifiers
If you have strong coding skills but no specific project of your own, you're also very welcome to join and support the work of others.
Not necessarily - but you should be comfortable programming in some language.
QuPath is written in Java and scripted with Groovy, but the hackathon isn't really concerned with developing QuPath itself.
If you're a Python, R or C++ aficionado (for example) and want to link up your work with QuPath, that all fits.
Thanks to @jni, we'll be using image.sc Island Gather.Town virtual space, which is kindly sponsored by Gather.
It's the same space as has been used for other events by projects affiliated with the Scientific Community Image Forum at image.sc.
To join, you should first sign up for one of the webinars on 25 April 2022:
- QuPath: Past, Present & Future - 08:00 UTC (09:00 BST / 10:00 CET)
- QuPath: Past, Present & Future - 17:00 UTC (18:00 BST / 19:00 CET)
There you can select the option that you want to participate in the hackathon as well. If you do, you'll be emailed more details later.
You can then head over to Discussions to write a bit about what you want to achieve.
Important! Numbers on image.sc island are limited, so please do write a few sentences describing why you want to participate.
I hope everyone who wants to will be able to join - but in case it's badly oversubscribed, preference will be given to applicants who plan to develop new scripts/extensions that will be open-source and useful to the wider community. Everyone should at least demonstrate at least some knowledge of programming, since the hackathon won't be concerned with general use of the software.
That's fine!
Four days is a lot. Timezones and other commitments will make things tricky.
However, because space on the island is limited, please don't sign up unless you really plan to join for a substantial amount of time (e.g. a day or two, or equivalent number of hours).
Depending upon how it goes, there might be more hackathons and image.sc island events in the future.
You can find a bit more information and ask questions on Discussions.
If you want to contact me directly, you can find me on image.sc at https://forum.image.sc/u/petebankhead/
Any content on this repo is really just for information and discussion. Unless otherwise stated (e.g. in the header of a specific script), it's available under a Creative Commons Zero (Public Domain) license.
Note that this doesn't apply to QuPath itself, which is under The GNU General Public License v3.0. See https://github.com/qupath/qupath for information about QuPath license.