What are you doing this week? Feel free to share!
Keep in mind it’s OK to do nothing at all, too.
It appears that this week I’m dealing with angry emails from people who think I’m part of some backstabbing conspiracy to ruin the unimpeachable and absolutely innocent saint of all saints Dear Leader Richard Stallman. (I made the mistake of responding to one such email – never again.)
Other than that, I’m studying for my citizenship exam. This Friday!
Finishing up a personal blog (again) and writing up the first legit post (again). Maybe it’ll teach me to think before I open my mouth.
I am trying to understand how a web framework works by building one in Nim.
By the way, learning Nim is really nice: the language is well thought and the standard library is surprisingly really rich already :)
Do let me know how that ends up. I’m currently writing a site using jester, and I plan on writing more docs and examples for it, since it seems to be the most comprehensive web framework I’ve seen so far.
If you like how jester does things, I think it’d be nice to bring some more people to help maintain it. I’m not a maintainer for it right now, but I’m very interested is helping it be fleshed out, at least with more docs and examples.
If you’re not interested in that right now, I understand, just trying to coordinate with other people interested in Nim
Hi Yumaikas !
This project is mainly to get used to code with Nim as I never built something meaningful with it. When I will be more confident with the language and its ecosystem, I will be really happy to contribute to Jester too (be it documentation or code).
I will contact you when I am ready !
Chip8 Emulator is about to be wrapped up. Gonna make a debugger on top of it. Writing a debugger would be a good exercise.
Improving the onboarding process for another team. I have removed 4 steps from the process and will remove another tomorrow.
Working with Microsoft support on an obscure issue with the Windows Photos application.
Attending an HOA board meeting.
Baking an apple pie for a contest at work.
Making plans for this weekend.
More tickets this week, fiddling with how I keep tactical track of what I’m working on, switching from an electronic journal to a pen and paper. I’ll probably need to take some time to clean up my work journal entries a bit.
I currently am still working on porting my mini-wiki from Erlang to Nim. Once I get it ported, I’ll be allocating one more instance, that I plan on putting at feed.junglecoder.com. I also plan on writing some scripts to try to migrate the wiki state from ets files to SQLite databases. That migration will probably be the last bit of Erlang I write for a while. I might also just punt, and scrape my site’s markdown files down into the SQLite databases in question. It’ll be a good way to see how sophisitcated Nim’s web-scraping tools are, or a good excuse to look into Python’s beautiful soup if Nim is lacking what I need to parse that information out.
Right now I find myself talking more in small bites on Lobsters, the orange site and mastodon, as opposed to on my blog. The idea behind feed.junglecoder.com is to build a place where I can aggregate my activity across the various tech social sites for my own use (as well as anyone that wants to see that sort of thing). I also plan on using it to curate an ongoing “Best Of” series of the best comments I find on the sites I frequent, since I’ve noticed some rather quality comments here and on the orange site in the past few weeks that I think are worth highlighting.
Once I get this port finished and usable, I plan on writing some more on my blog. I have some post ideas I want to get back to, and I have have been forming Opinions about Nim as I’ve been working with it, and those would probably be best in a blog post.
Exploring youtube playlists of old DOS and Amiga game soundtracks, which has been a nice change of pace.
My team has (starting today) a fortnightly dedicated Research Day where we are free to study whatever programming thing we want and not do chores or work on the product.
I’m booking all the travel for our first company off-site. The word “off-site” is a bit weird given that we’re distributed and we don’t actually have a site to be on-site on.
I’ve been sent a whole bunch of super impressive résumés, and it’s going to be difficult to choose our next hire.
I’m working on a digital time capsule app that currently just a single Python file, and I’m also trying to start doing consulting work as a software dev. No luck yet!
When I have spare time, I’m rewriting the web frontend code of my endless arcade multiplayer game, https://sneakysnake.io. I finally fixed all the major bugs on mobile Chrome/Safari, and I’m ready to start improving the frontend after this cleanup.
Currently, the frontend is a big mess of Promise-heavy typescript with a complicated state machine used to track and update the visibility of DOM divs each frame. I’m rewriting it using svelte to update the DOM, and cleaning up the typescript as I go, making use of asyn/await to make the async portions easier to follow. Also greatly improving the error handling.
Once that’s done, it will be much easier to maintain and extend. Next new feature will be player-specified names. I just need to decide how to handle black-listing and/or reporting (I want to keep the game family-friendly).
I’m learning a ton about explainability in machine learning (XAI). In short, it’s about helping the user understand why a prediction was made. The problem is most CS papers in XAI neglect to specify who the user is and what “why” means. I found this gem that overviews the research in social sciences on explanations. It’s really helped me build a structured mental model for approaching XAI. Still, I’ve found some fascinating CS papers in XAI. This one explores having humans annotate which features were important in a training example. They claim that by annotating ~2% of the data, they can achieve the same model performance as a data set with 10X as many examples. This seems like a fascinating possibility for deep learning with less data, as well as simplifying data science problems via lay people.