Hey all. I will do a short presentation tomorrow for the team - I'm going to show remote and local caching for tests, and that tests run in parallel and explain some of the core concepts of pants (as far as I understood them) ... maybe someone has a suggestion or so what I could add to the presentation to create the excitement I'm hoping for 🙂 I'm using this repo for it, so a django example app https://github.com/elmcrest/example-django-multiple-monoliths-pantsbuild

I think the dependency inference is a big thing, which allows pants to operate on a subset of all available targets/files for whatever goal using either changed since or just path wildcards etc.

Yeah I try to cover this, any suggestions for some maybe easy to get command(s)? I vaguely remember that one could reference branches etc. I will look things up from docs too of course but maybe you have something particular in mind 🤷🏼‍♂️

pants --changed-since=origin/main test

like that perhaps?

I don't know how much time you have but you can show how dependency inference can lead to docker images much smaller than without using it. I was using pants at my previous job and we were able to build python images which were like 200-300 MBs. Here in my new job we just put everything and no images are less than 1 GB. There are some big culprits like tensorflow which you can't avoid. Otherwise pants and PEX gives you so much smaller images in comparison to what usually people deal with.

Sounds great, thx guys!

Let us know how it goes!

I had to skip the presentation cause I got sick, so I gained some more time to prepare it 🙂