A review of Q1 2021
I realize I haven’t done one of these in a while, a few … status-update-worthy things have piled up. Bangalore is currently in a lockdown after ranking in the top 3 worst COVID-hit cities, in India. As such, I’m sitting at home, in my home-office thing which now has an extra monitor, writing this. We’ve got a lot to talk about — let’s get to it!
working at DeepSource
Starting January of this year, I’ve been working as a Site Reliability Engineer at DeepSource. As a Kubernetes shop, nearly all of my day to day work involves working with Kubernetes. I didn’t think I’d say this, but I actually quite like using K8s now. The more I understand it, the more I find myself appreciating the need for it. Of course, I wasn’t around for the initial setup of everything — I’m merely enjoying using an already stable cluster environment.
So far, I’ve set up an event-driven observability pipeline (pictured below), some infosec work and the usual SRE stuff. I’ll probably write about the observability infrastructure in detail sometime — perhaps on the company blog.
I even managed to break prod within the first month! But in all seriousness, it’s pretty riveting work, with some very fun people.
There haven’t been too many, I must be honest. I did briefly consider the idea of building a small SaaS — a simple email-based bookmarking service. Send a bunch of links you want to bookmark to an email address, and get back the entire webpage as a thread of emails.
I got as far as writing the backend for it, at
forlater/donkey — a simple
Flask app that pulls web pages and sends an email, on a webhook. And a
helper tool forlater/mdawh in
Go that takes mail in STDIN and sends a webhook to an endpoint. I used
OpenSMTPD to call
mdawh when mail arrived. Overall, a pretty simple
system. Except, HTML email sucks. And the modern web sucks. I quickly
got bored of it — dealing with websites not getting rendered correctly,
email delivery being pretty shit overall and the fact that nobody would
actually use something like this, let alone pay for it. Of course, I
could be wrong and someone looking for a service like this could be out
there — and if they’re reading this, please email
That said, I’m open to revisiting this project sometime. Perhaps with a different use-case, even.
In 2020, I began getting into the Cosmere. I’ve read nearly all the books in it, save for the Mistborn series, which I’m reading at present. Still in Era 1, having finished The Final Empire and The Well of Ascension. I’m taking a break before I dive into The Hero of Ages.
Aside from high fantasy reading, I’ve begun spending some time reading the essays at Slate Star Codex — more specifically, the selected few at https://www.slatestarcodexabridged.com.
I’d begun learning Russian sometime last year, but stopped studying it for about 6 months or so. Recently, I decided to pick it up again after coming across this video by Johnny Harris on how he learnt Italian. In essence, he talks about why the textbook method of learning a language, i.e., the grammar: conjugations, rules, exceptions, etc. is ineffective. Instead, he suggests starting with learning the 1000 most frequently used words in that language which helps build an intuition for the language.
I’ve found two decks from the publicly shared decks to be really good: 1000 frequently used words, 7000 sentences in order of difficulty. This lets me 1) learn the basics (vocabulary, sentence construction) and start applying them in making my own sentences and learning the grammar intuitively, instead of memorizing rules and conjugations.
For a good portion of last year, I was unable to get any physical exercise done — gyms were closed, running downstairs just wasn’t fun. Early this year, before wave 2 of the virus, the gyms here opened for a brief bit and I managed to get some swole on. They’re closed again, now but I’ve taken to running about 3 – 4 km everyday. It’s still quite boring compared to lifting, but it needs to be done.
I’ve hit a writer’s block, so to speak, with this blog. I have a few ideas for future posts, but they require actual research (read: I can’t just pull up Vim and type away), so they will remain as ideas for now. If you have some ideas for things I can write about, please shoot them my way.