axfelix.github.io

View My GitHub Profile

Recurse Center Blagg – Feb 4, Day 24

The final blogdown… everything’s ending here… these references are aging me… etc. I think this will be my last check-in, after a fairly heads-down week! In part I’ve been preoccupied with logistics – Canada announced additional airport quarantine measures which meant that we had to change our plans to head home a few weeks ahead of schedule. As of now I’m still in LA (which I don’t think I’d mentioned in my blogposts to date – I came to LA with my long-suffering wife and cats for a couple months in the middle of a pandemic to clear my head, it sounds cockamamie but it was a very good decision), but I’ll be on the road north by the end of next week, and my state of mind definitely reflects that. I’m defintely midcareer at this point if not middle aged, and getting a chance to bring all (or most) of my focus and energy to something new, especially in a year like this, is no small feat. And I feel like I actually managed that out of the gate at Recurse! I am proud of that, and relieved, and in fact I’m so relieved that I’m getting tired, and ready to go home.

I did manage to knock out one last big project, though 😎

Presenting: ffmpeg on Ice!

ffmpeg logo in an ice rink

It uses Tailwind (fullstack developers over 30 learn a CSS framework other than bootstrap challenge) with Rails on Heroku in order to provide some canned demos of the transformations provided in the very cool ffmpeg-artschool project. I’ve never actually shipped or contributed to a Ruby project before – as a longtime Python developer, my approach to Ruby has always been “it’s fine” – but it was the right choice for this app. As a big static site fan, it was also nicely illustrative to see how many templating frameworks seem to have come out of the Rails community, and likewise for webpack use cases; I’d picked a lot of this up elsewhere, but it makes a good deal more sense now.

Unfortunately, it also runs free Heroku instances out of memory almost immediately – it turns out it’s impossible to generate h264 from ffmpeg in less than ~1.5GB and I don’t really want to pay for more than the ~500M basic tier, so I probably won’t be keeping this up and working. I wish I’d realized this before starting the project, but them’s the breaks; I recommend cloning it locally and playing with it!

I also handled a bunch of other small stuff that came up over the past week and a half, including:

It’s funny – I’ve noticed that a lot of other folks in my cohort are interested in learning about CI, or test driven development, or managing large datasets, and I’ve found that I can give unexpectedly useful feedback in these conversations … because I’ve done some form of that for the past decade. Mid-career, like I said. But, more than that, I think I’m getting to a point where I like being able to give feedback and direction more than I like whipping through these various subfields. I like breadth. And I get to take some time off to decompress now, but I’m going to keep that in mind for the future.

Thanks Recurse!!!!!!