A new device for filming (yes, there are devices, but no actual filming yet) – macro lenses for a phone.
Two pieces, each with x20 magnification, complete with a clip and lighting.
Unfortunately, my phone doesn’t have a native macro lens at all – only an ultra-wide and a telephoto lens (the latter up to x5).
So I had to upgrade the main camera with some extra gear.
Example videos show footage from the original camera, with one lens (x20), and with two lenses (which should theoretically be like x400).
I’m not sure yet where to use two lenses – the macro is just too strong. But one works quite well; perhaps I should also consider intermediate options like x5, x10, or x15 to cover the full range of possibilities.
Two pieces, each with x20 magnification, complete with a clip and lighting.
Unfortunately, my phone doesn’t have a native macro lens at all – only an ultra-wide and a telephoto lens (the latter up to x5).
So I had to upgrade the main camera with some extra gear.
Example videos show footage from the original camera, with one lens (x20), and with two lenses (which should theoretically be like x400).
I’m not sure yet where to use two lenses – the macro is just too strong. But one works quite well; perhaps I should also consider intermediate options like x5, x10, or x15 to cover the full range of possibilities.
About GPT-5 and Cursor
A week ago, a new model was released that was offered "for free" in Cursor.
I decided to use this opportunity to build a pet project for tracking income and expenses.
I’ve been tracking all my income and expenses in Google Sheets for the past six months. The main goal is to understand my average monthly expenses.
I even managed to "vibe-code" a Telegram bot to help with these spreadsheets, but now I felt like trying something different. And the models have improved a lot in the last six months.
So — GPT-5 is like night and day (though not drastically better than Claude 4). I deliberately chose a stack I wasn't familiar with — Go, React, and gRPC. And GPT-5 helped me build a working MVP in just a couple of days, which I’m now refining to fit my needs. I’ll announce the launch separately, but for now you can check out the repository.
P.S. The free GPT-5 in Cursor turned out not to be free — it stopped working after $200 worth of API usage (all under the $20 Cursor subscription).
A week ago, a new model was released that was offered "for free" in Cursor.
I decided to use this opportunity to build a pet project for tracking income and expenses.
I’ve been tracking all my income and expenses in Google Sheets for the past six months. The main goal is to understand my average monthly expenses.
I even managed to "vibe-code" a Telegram bot to help with these spreadsheets, but now I felt like trying something different. And the models have improved a lot in the last six months.
So — GPT-5 is like night and day (though not drastically better than Claude 4). I deliberately chose a stack I wasn't familiar with — Go, React, and gRPC. And GPT-5 helped me build a working MVP in just a couple of days, which I’m now refining to fit my needs. I’ll announce the launch separately, but for now you can check out the repository.
P.S. The free GPT-5 in Cursor turned out not to be free — it stopped working after $200 worth of API usage (all under the $20 Cursor subscription).
Everyone knows what an RNG is—an algorithm that generates a random number.
However, all algorithms are susceptible to attacks and vulnerabilities, so services with high security requirements tend to avoid relying on them—cryptography, encrypted communication, digital signatures, and so on. Some banks or systems ask users to move their mouse or press keys a certain number of times to generate random events.
Physical RNGs can be hardware-based or rely on physical phenomena, but in any case, they use a real source of entropy.
Some unusual sources of entropy used or studied:
1. Lava lamps – the chaotic movement of wax in dozens of lava lamps.
2. Geiger counter – intervals between events of radioactive decay.
3. Goldfish in an aquarium – random movement of fish in water.
4. A laundromat with a microphone – noise from washing machines.
5. A dance floor with pressure sensors – chaotic movements of people.
So your security might depend not only on software and algorithms, but also on fish in an aquarium.
However, all algorithms are susceptible to attacks and vulnerabilities, so services with high security requirements tend to avoid relying on them—cryptography, encrypted communication, digital signatures, and so on. Some banks or systems ask users to move their mouse or press keys a certain number of times to generate random events.
Physical RNGs can be hardware-based or rely on physical phenomena, but in any case, they use a real source of entropy.
Some unusual sources of entropy used or studied:
1. Lava lamps – the chaotic movement of wax in dozens of lava lamps.
2. Geiger counter – intervals between events of radioactive decay.
3. Goldfish in an aquarium – random movement of fish in water.
4. A laundromat with a microphone – noise from washing machines.
5. A dance floor with pressure sensors – chaotic movements of people.
So your security might depend not only on software and algorithms, but also on fish in an aquarium.
Today, YouTube surprised me with something wonderful – a CAD competition.
For a while, my information bubble was filled with Tetris competitions, then Japanese game shows and GeoGuessr (that’s the one where you're given a spot in Google Street View and have to guess the exact coordinates on the planet – totally addictive, highly recommend).
And now, CAD competitions:
1. Participants are given a technical drawing of a model.
2. They must recreate the model in CAD.
3. The winner is the one who sends the correct model weight to the chat the fastest.
Unlike other types of competitions, this one is more relevant to my interests – it’s fascinating to see the approaches participants take in modeling, and even if it’s not exactly about best practices, you can still pick up some useful techniques.
For a while, my information bubble was filled with Tetris competitions, then Japanese game shows and GeoGuessr (that’s the one where you're given a spot in Google Street View and have to guess the exact coordinates on the planet – totally addictive, highly recommend).
And now, CAD competitions:
1. Participants are given a technical drawing of a model.
2. They must recreate the model in CAD.
3. The winner is the one who sends the correct model weight to the chat the fastest.
Unlike other types of competitions, this one is more relevant to my interests – it’s fascinating to see the approaches participants take in modeling, and even if it’s not exactly about best practices, you can still pick up some useful techniques.
I came across a study on the topic of progress in various fields with and without AI.
From personal experience, I agree with the conclusions. Using AI provides a previously impossible boost in the early stages, which helps reduce the overall time needed for learning, improvement, etc.
At the same time, if you rely solely on AI, your maximum efficiency will be limited by the capabilities of the model you're using. The quality of models is improving, but I’m convinced that combining AI with your own thinking will yield better results in the foreseeable future.
The article contains plenty of additional information and charts - worth reading if you're interested.
From personal experience, I agree with the conclusions. Using AI provides a previously impossible boost in the early stages, which helps reduce the overall time needed for learning, improvement, etc.
At the same time, if you rely solely on AI, your maximum efficiency will be limited by the capabilities of the model you're using. The quality of models is improving, but I’m convinced that combining AI with your own thinking will yield better results in the foreseeable future.
The article contains plenty of additional information and charts - worth reading if you're interested.
All publication schedules were thrown off due to the vacation, so I’ll be writing about whatever is currently relevant, without sticking to specific dates.
Today’s post is about my recent development — a Telegram emulator. Why build such a "reinvention of the wheel," what it can do, and how long it took to develop — read about it on the blog: https://positroid.tech/en/post/local-telegram-emulator.
Today’s post is about my recent development — a Telegram emulator. Why build such a "reinvention of the wheel," what it can do, and how long it took to develop — read about it on the blog: https://positroid.tech/en/post/local-telegram-emulator.
This media is not supported in your browser
VIEW IN TELEGRAM
You can watch fire, water, and Cursor writing tests for three projects at once — forever.
I suddenly remembered touch typing—I mastered it about five years ago and haven’t looked at the keyboard since.
Before that, I could already type quickly and had developed some muscle memory, but I still had to glance at the keyboard from time to time, especially for special characters and function keys.
There were several attempts to learn it, starting back in childhood with Solo on the Keyboard, and the most successful one was five years ago using typingclub.com.
The service is as simple as it gets, fairly gamified—I recommend it.
I dug up some old stats: Russian—39 words per minute, English—33 words per minute.
I retested just now: Russian—61 words per minute, English—42 words per minute.
I won’t say whether that’s a lot or a little—it’s enough for me. Still, I’m surprised by people who still type with just two fingers, even though most of their job is writing. And during coding sections of interviews or design interviews (about architecture), it’s a secondary but still noticeable skill.
Before that, I could already type quickly and had developed some muscle memory, but I still had to glance at the keyboard from time to time, especially for special characters and function keys.
There were several attempts to learn it, starting back in childhood with Solo on the Keyboard, and the most successful one was five years ago using typingclub.com.
The service is as simple as it gets, fairly gamified—I recommend it.
I dug up some old stats: Russian—39 words per minute, English—33 words per minute.
I retested just now: Russian—61 words per minute, English—42 words per minute.
I won’t say whether that’s a lot or a little—it’s enough for me. Still, I’m surprised by people who still type with just two fingers, even though most of their job is writing. And during coding sections of interviews or design interviews (about architecture), it’s a secondary but still noticeable skill.
Today’s blog post is about image2model — generating a 3D model from an image: https://positroid.tech/en/post/image2model
I had done something similar with the Caretaker before, but in the article series, it was easy to overlook that part, even though the tools are very easy to use and quite helpful.
I had done something similar with the Caretaker before, but in the article series, it was easy to overlook that part, even though the tools are very easy to use and quite helpful.
I've been into 3D printing for over 7 years now.
And I've never dried filament. Occasionally, there was some stringing, but nothing out of the ordinary. Until last month.
I ordered two spools of filament from a no-name manufacturer for $8/1kg each. The packaging was poor, but that wasn’t the main issue.
The filament just wouldn’t print properly. I mean, it did print, but with a lot of stringing and defects. No matter the temperature setting, the results were awful. I even tried drying it in the oven at 50°C for 4 hours — no improvement.
Eventually, I gave up and bought a filament dryer. I got the cheapest one that has both heating and airflow — the Creality Dry Box 2.0.
Threw the spool in at 65°C for 8 hours and... it was a miracle — all the problems vanished. Flawless print quality, perfect temperature tower, and great Benchy.
It seems that all these years, many of my printing and retraction issues — especially with Bowden setups — could have been drastically reduced...
And I've never dried filament. Occasionally, there was some stringing, but nothing out of the ordinary. Until last month.
I ordered two spools of filament from a no-name manufacturer for $8/1kg each. The packaging was poor, but that wasn’t the main issue.
The filament just wouldn’t print properly. I mean, it did print, but with a lot of stringing and defects. No matter the temperature setting, the results were awful. I even tried drying it in the oven at 50°C for 4 hours — no improvement.
Eventually, I gave up and bought a filament dryer. I got the cheapest one that has both heating and airflow — the Creality Dry Box 2.0.
Threw the spool in at 65°C for 8 hours and... it was a miracle — all the problems vanished. Flawless print quality, perfect temperature tower, and great Benchy.
It seems that all these years, many of my printing and retraction issues — especially with Bowden setups — could have been drastically reduced...