I’m spending this weekend working on my book site https://el-kitap.org/ — adding some small improvements to the browsing experience and cleaning up a few outdated sections.
I’m also taking time to think about the bigger picture: what strategy makes the most sense for growing the project. Should I focus more on curating high-quality content for a niche audience, or go broader and try to reach casual readers with trending topics? Still figuring out the right balance between steady organic growth and faster promotional pushes.
A CLI tool to make reasoning about Postgres locks a bit easier.
Given a database[1] and a set of DDL statements/migrations you want to check, pglockanalyze will open a transaction, execute the statements, read the pg_locks view to analyze the locks they acquire and rollback (or commit, depending on the flags you passed) the transaction. Then, it will output the results for each statement.
I think there's merit in this idea, that said it's very much an experiment so there could be flaws and/or corner cases that this strategy won't work well for.
It's meant to act as a complement, not a replacement, to things like static analysis and the official Postgres docs.
I'm working on an article/tutorial about how to develop an intuition behind the WCAG contrast rules so you can spot bad contrast problems yourself instead of getting surprised (and irritated) by contrast checker tools giving you failures later. And how to solve color contrast problems like what design options you have when two colors don't contrast.
After speaking to other designers, I find most want to follow the WCAG rules but get stuck going in circles not knowing how to fix color contrast issues e.g. when changing one color throws another one off, and how to work with brand/primary colors that lack contrast (like orange/yellow on white).
Since working on my palette designer, I developed a lot of intuition in this area so wanted to share what I learned. There's lots of articles on WCAG rules, but many will say "pick accessible colors" without really helping you understand how, especially when you have to stick to certain brand colors.
I have one more chapter left to read in the book The Analogical Mind: Perspectives from Cognitive Science[1] so finishing that up is on the plate for this weekend. That largely wraps up a several month long deep dive into analogical reasoning that I've been engaged in. After that, I plan to shift back to studying associative networks / semantic networks / concept graphs and suchlike, along with revisiting some neural network stuff (not LLM's mind you).
Aside from that, I'll be doing one thing I don't normally do much of over the weekend, and that's some $DAYJOB work. I'm working on some experiments with the OBO (On Behalf Of)[2] login flow with Azure Entra, and since I'll be on PTO most of next week, I'm trying to finish things up, at least to a point where I can cleanly hand it over to a co-worker to finish this coming week.
This weekend, I have just integrated new model releases into VT: GPT-5, gpt-oss-20b, gpt-oss-120b, and Claude Opus 4.1.
Also optimized workflow utilities and chat pipeline handling, updated the UI/UX for improved performance, and added an icon provider for model selection.
I'm spending my weekend planning a new hobby project to for the first time use both what I know about electronics and the programming knowledge and make my own WiFi enabled touch screen computer based on a couple ESP32 chips.
My band is playing a small bluegrass festival this weekend. We’ve been recording an album since last year and are starting the work of promoting it by playing more gigs!
The idea is to give it a task, let it do whatever it decides and expect to produce a result *inside container*. I'm still exploring this idea whether it makes sense, and is cost effective.
Learning n8n a little more for some automated tasks on a hobby project. Maybe try re-learning sketchup to diagram some home improvement/woodworking projects.
Probably overthink buying a mini PC to run Ubuntu on to do homelab like things, like running n8n or other docker images on like plex. Then probably abandon the idea and watch something off a streaming service, lol.
I'm working on an addon for Godot called Godot Sandbox, and we're trying to also build it as a module. A module is embeddable into Godot, so that you can have only one executable be the whole game. Addons are dynamically loaded at run-time. It's been a lot of work, but integration is just as important as making the addon itself.
I’m spending this weekend working on my book site https://el-kitap.org/ — adding some small improvements to the browsing experience and cleaning up a few outdated sections.
I’m also taking time to think about the bigger picture: what strategy makes the most sense for growing the project. Should I focus more on curating high-quality content for a niche audience, or go broader and try to reach casual readers with trending topics? Still figuring out the right balance between steady organic growth and faster promotional pushes.
A CLI tool to make reasoning about Postgres locks a bit easier.
Given a database[1] and a set of DDL statements/migrations you want to check, pglockanalyze will open a transaction, execute the statements, read the pg_locks view to analyze the locks they acquire and rollback (or commit, depending on the flags you passed) the transaction. Then, it will output the results for each statement.
I think there's merit in this idea, that said it's very much an experiment so there could be flaws and/or corner cases that this strategy won't work well for.
It's meant to act as a complement, not a replacement, to things like static analysis and the official Postgres docs.
https://github.com/agis/pglockanalyze
[1] typically an ephemeral database spawned by your CI pipeline
I'm iterating on my accessible color palette designer:
https://www.inclusivecolors.com/
I'm working on an article/tutorial about how to develop an intuition behind the WCAG contrast rules so you can spot bad contrast problems yourself instead of getting surprised (and irritated) by contrast checker tools giving you failures later. And how to solve color contrast problems like what design options you have when two colors don't contrast.
After speaking to other designers, I find most want to follow the WCAG rules but get stuck going in circles not knowing how to fix color contrast issues e.g. when changing one color throws another one off, and how to work with brand/primary colors that lack contrast (like orange/yellow on white).
Since working on my palette designer, I developed a lot of intuition in this area so wanted to share what I learned. There's lots of articles on WCAG rules, but many will say "pick accessible colors" without really helping you understand how, especially when you have to stick to certain brand colors.
I have one more chapter left to read in the book The Analogical Mind: Perspectives from Cognitive Science[1] so finishing that up is on the plate for this weekend. That largely wraps up a several month long deep dive into analogical reasoning that I've been engaged in. After that, I plan to shift back to studying associative networks / semantic networks / concept graphs and suchlike, along with revisiting some neural network stuff (not LLM's mind you).
Aside from that, I'll be doing one thing I don't normally do much of over the weekend, and that's some $DAYJOB work. I'm working on some experiments with the OBO (On Behalf Of)[2] login flow with Azure Entra, and since I'll be on PTO most of next week, I'm trying to finish things up, at least to a point where I can cleanly hand it over to a co-worker to finish this coming week.
[1]: https://direct.mit.edu/books/edited-volume/2224/The-Analogic...
[2]: https://learn.microsoft.com/en-us/entra/identity-platform/v2...
This weekend, I have just integrated new model releases into VT: GPT-5, gpt-oss-20b, gpt-oss-120b, and Claude Opus 4.1.
Also optimized workflow utilities and chat pipeline handling, updated the UI/UX for improved performance, and added an icon provider for model selection.
https://vtchat.io.vn
I'm spending my weekend planning a new hobby project to for the first time use both what I know about electronics and the programming knowledge and make my own WiFi enabled touch screen computer based on a couple ESP32 chips.
I work on various github projects: 1) Header-only ASCII font renderer for embedded/SDL/terminal https://github.com/Ferki-git-creator/ascii-render
2) TurboStitchGIF: A fast, header-only C GIF decoder without dynamic allocations, ideal for embedded systems and cross-platform projects. https://github.com/Ferki-git-creator/TurboStitchGIF-HeaderOn...
My band is playing a small bluegrass festival this weekend. We’ve been recording an album since last year and are starting the work of promoting it by playing more gigs!
Using GH Copilot Agent to vibe code a LLM+Docker omnitool for my coding assistant: https://github.com/gtanczyk/genaicode/pull/25
The idea is to give it a task, let it do whatever it decides and expect to produce a result *inside container*. I'm still exploring this idea whether it makes sense, and is cost effective.
Learning n8n a little more for some automated tasks on a hobby project. Maybe try re-learning sketchup to diagram some home improvement/woodworking projects.
Probably overthink buying a mini PC to run Ubuntu on to do homelab like things, like running n8n or other docker images on like plex. Then probably abandon the idea and watch something off a streaming service, lol.
I'm working on an addon for Godot called Godot Sandbox, and we're trying to also build it as a module. A module is embeddable into Godot, so that you can have only one executable be the whole game. Addons are dynamically loaded at run-time. It's been a lot of work, but integration is just as important as making the addon itself.
My budget - the reality of living off my savings hasn't totally reflected my spending yet (mostly though). I'm trying out switching to cash.
My family - they're coming to visit for a few days
My health - evaluating what's been working for weight loss, charting my progress.
Trying to understand Variational Inference. Struggling like hell though so if anybody has some great resources, I would be thankful.
Adding tests to the test automation of a personal project
Not my mental health
Learning Spanish to at least get to a B1 level by next February when my wife and I will be in Costa Rica for a month and half.
Spending time in the gym and spending time with my wife.
Working on my tan today. Travelling across Ukraine by train tomorrow.