HNHacker News
TopNewBestAskShowJobs

oliwary

834 karma · joined October 14, 2013

For inquiries regarding squareword, you can reach me at squareword.org@gmail.com
submissionscomments
oliwary··on Show HN: The HN Arcade
Fun idea! I am glad there seems to be a community of people enjoying word games on HN, it tends to be the first place I post my games. https://squareword.org (wordle meets sudoku) and https://clickword.org (scrabble meets tetris... ish) are the largest so far.

I think https://videopuzzle.org has a lot of potential also, but it targets an audience that's a bit different as it is more visual, focused on reassembling videos like a puzzle.

Will post these to the arcade! :)

oliwary··on New York Times games are hard: A computational perspective
If someone wants a challenging Wordle variant, I've made this: https://squareword.org

It's a combination of 2D Wordle, crosswords and Sudoku. Been running for over 4 years :)

oliwary··on AI is a horse (2024)
IMO it really came into its own after the first season. S1 felt like mad men but with computers, whereas in the latter seasons it focused more on the characters - quite beautiful and sad at times.
oliwary··on AI is a horse (2024)
"Computers aren't the thing. They're the thing that gets you to the thing."

My favorite quote from the excellent show halt and catch fire. Maybe applicable to AI too?

oliwary··on Your app subscription is now my weekend project
This is a fun article and approach.

Subscription apps often have to target a wide userbase. However, most users only need a small subset of the entire feature set, and would be better served by a tailored version. This means that vibecoded apps can get away with being much less complex (specific featureset, no login etc), while still being more useful.

I have also created tools with LLMs that are exactly tailored to what I need, and still much more polished than what I could do without LLMs. Will have to think about if there is anything else I can do this with.

oliwary··on Uncrossy
Nicely done! This is an excellent experience, both visually and gameplay wise.

The only thing that felt a bit weird was being able to change the word to the same word, if the slid word had the same letter twice.

oliwary··on ChatGPT Health
Had a look and it does not show up anywhere else: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
oliwary··on A 30B Qwen model walks into a Raspberry Pi and runs in real time
Does anyone have any use-cases for long-running, interesting tasks that do not require perfect accuracy? Seems like this would be the sweet spot for running local models on low-powered hardware.
oliwary··on Show HN: Make audio loops online
Fun! Well done.

I created a loop in the drum mode, and then wanted to add a melodic track to it, so I clicked the melodic button. This erased my drum pattern instead - would be good if they could play over each other, and would not be erased by switching.

oliwary··on enclose.horse
Thank you! :) If you click the three dots on the left top side of the letters area, you can shift all tiles in a direction.
oliwary··on AWS raises GPU prices 15% on a Saturday, hopes you weren't paying attention
Ideally. Companies may also be hesitant to invest in expanding production capacity if they assume this is a bubble though.
oliwary··on enclose.horse
Another thing to try could be to rank people in realtime instead of the one-off submission approach. I do this in https://spaceword.org (create tight crosswords using 21 letters), and I think it's quite motivating to see how you compare to others as you improve your solution. On the other hand, its a bit more taxing on the server, and then you also could not show the optimal solution.
oliwary··on Ask HN: Any example of successful vibe-coded product?
https://spaceword.org - a daily word game inspired by banana grams, where you need to arrange 21 letters in a tight square. Has around 400 daily active players.

I'm pretty familiar with the underlying stack, which helped a lot since I knew the pitfalls. But pretty much all of the code is written by an LLM.

oliwary··on Ask HN: By what percentage has AI changed your output as a software engineer?
I think about 4x for greenfield development. For example, I implemented this motion controlled game from scratch in 5 days: https://motionparty.net

For this kind of no expectations, for fun development I find AI makes it much easier to develop and test hypotheses. For other styles it's different, especially if the stakes are higher.

oliwary··on Pocketbase – open-source realtime back end in 1 file
A couple of daily word games. https://squareword.org is the biggest one.
oliwary··on Pocketbase – open-source realtime back end in 1 file
I can mirror everyone singing praise to pocketbase here. Once you grasp the concepts (which map pretty closely to SQL concepts, with rules for row-based security), it is by far the easiest way IMO to create a maintainable, robust backend with direct auth integrations and a pleasant interface.

I have around 5 instances of pocketbase running on a 10 USD/month Hetzner server, serving thousands of users a day without breaking a sweat.

oliwary··on When UPS charged me a $684 tariff on $355 of vintage computer parts
Interesting that the Matrix picked the 90s as it's world to lock humanity into as well... Prescient?
oliwary··on Ask HN: What Are You Working On? (Nov 2025)
Been continuing to work on my daily puzzle game: https://videopuzzle.org

Yesterday was the one year anniversary, meaning there are 365 videos to unscramble. :)

oliwary··on Alterego: Thought to Text
The key to technology like this is how quickly it can get to say 99.5%. Not convinced it will be an easy path!
oliwary··on Survey: a third of senior developers say over half their code is AI-generated
Hey! I would encourage you to try our Claude code instead, which is also part of your subscription. It's a CLI that takes care of many of the issues you encountered, as it works directly on the code files in a directory. No more copy pasting or unscrambling results. Likewise, it can run commands itself to e.g. compile or even test code.
oliwary··on Nvidia DGX Spark
True... It would be very interesting to make a comparison of various open models based on token generation speed on these platforms. Presumably starting st some size the larger accessible RAM wins out over raw speed but low VRAM? Although I suppose things like MoE and FP would also matter.
oliwary··on The New York Times Mini Crossword Is No Longer Free to Play
In case anyone is looking for alternatives, my 2-dimensional daily word game (5x5 letters, five words down, five across) is free: https://squareword.org . Not exactly the same as you reveal words with wordle style guesses instead of clues, but might scratch the same itch.
oliwary··on Gemini 2.5 Flash Image
Also very confused at this... It told me "I'm unable to create images of specific individuals in different settings." I wish it would at least say somewhere which model we are using at the moment.
oliwary··on Reading for pleasure plummets by 40% in the US
I find that when I start reading a book that I enjoy, it is easily the most gripping form of entertainment, and I can barely put the book down. But, it also takes more effort to get started, so I rarely get into that zone and end up scrolling my phone instead.
oliwary··on Playing every game of Wordle simultaneously
Fun analysis! I run https://squareword.org, which has 5x5 letters featuring words both down and across. Guesses are applied to the horizontal words, but the vertical words can be used as support, a bit like in Sudoku.

I wonder how the analysis would work here - since the verticals form words as well I think you get a lot more information from each guess. Many players can now solve them consistently in 6 or 7 guesses.

oliwary··on Pfeilstorch
This is hilarious in hindsight... I wonder what kind of beliefs we hold that will seem equally funny to the people in 200 years. ;)
oliwary··on Show HN: Explainle – explain words, AI tries to guess
Hi!

I just launched Explainle (https://explainle.org), my latest word game. Every day, you have to explain 5 words without using forbidden words. An LLM tries to guess which word you are referring to. The fewer attempts you need, the more points you score!

This is the first game I've built with Claude Code, which I have found to be very powerful indeed. Also my first game to use LLMs as part of the game - quite intrigued to see how it will scale haha. The frontend is built using Vue.js, the backend is pocketbase.

Here is my score for the day:

Explainle #24 Total Points: 5000 (Top Score)

Word 1: (1000pt) Word 2: (1000pt) Word 3: (1000pt) Word 4: (1000pt) Word 5: (1000pt)

Would love any feedback!

oliwary··on Whatever happened to cheap eReaders?
I have a kindle 4 from 2011, still runs like a champ and holds multiple weeks of charge. I have had zero issues. I wish all gadgets were like this.
oliwary··on Square Theory
Glad you like it! :) And thank you for your comments, super interesting! Excellent point about the rarity of the perfect arrangement. Perhaps I should throw in a few lettersets that do have a solution, I am intrigued to see if people would discover it.

My other game, https://squareword.org focuses exclusively on perfect 5x5 squares, but here the goal is to uncover it wordle-style rather than arranging it from scratch. There are surprisingly few combinations that have ten unique, common words in a 5x5 letter square!

oliwary··on Square Theory
Fair enough! :) I'll look into adding a practice mode.
← PreviousPage 2 of 9Next →