Ask HN: What are you working on (August 2024)?
What are you working on? Any new ideas that you're thinking about?
So lately I've been working on a "v2" that exposes a full superset of GLSL, so you can write arbitrary shaders -- even foregoing SDFs altogether -- in a high-level lisp language. The core "default" raymarcher is still there, but you can choose to ignore it and implement, say, volumetric rendering, while still using the provided SDF combinators if you want.
The new implementation is much more general and flexible, and it now supports things like 2D extrusions, mesh export for 3D printing, user-defined procedural noise functions... anything you can do in Shadertoy, you can now do in Bauble. One upcoming feature that I'm very excited about is custom uniforms and embedding in other webpages -- so you can write a blog post with interactive 3D visualizations, for example.
(Also as a fun coincidence: my first cast bronze Bauble arrived today! https://x.com/ianthehenry/status/1827461714524434883)
About 50% of my days are spend doing the coin dealer stuff - hunting for inventory, buying collections/doing appraisals, going to coin shows and buying and selling in person, etc.
The other 50% I’m writing code and building out the tech stack for this business. I’ve written the whole backend for the retail site myself, which includes my own inventory management system, sync with eBay and other marketplaces, etc.
I’ve also built out a research tool which includes an ML price prediction engine engine (which sounds fancy but is really just a tabular regression model).
Backend is written in Crystal because I love the language and there’s nobody stopping me from using it :) Frontend is all Svelte and they’re glued together using a mini framework I wrote:
https://github.com/noahlh/celestite
I probably have 5 years worth of ideas I still want to build and I wish I could spend even more time building it all, but it’s super fun actually using it in the real live marketplace so I’d never give that up.
Happy to chat about this stuff with anyone who’s interested or vaguely interested in numismatics.
It’s something I’ve been wanting for a while, for example to read a book with a group of friends or with a work team, but there’s lots of other possibilities including author reading parties, proofing and education. Got the basics of it working now, need to polish the UI and add the commenting and highlighting features.
I’m using Next.js and Supabase, neither of which I’ve used before so it’s been a fun but often frustrating process. Claude has been an amazing assistance, fixing my mistakes and countless type errors.
This is a hierarchical representation of any given piece of knowledge.
It starts with a tree root node that you specify (let's say Kung fu), then it branches out into multiples subcategories (techniques, styles, philosophy, weaponry, ect...) and then you can click on these subcategories to branch out even more into the graphical tree.
This is all generated on-the-fly with Claude 3.5. There is no limit to what knowledge you might explore.
The killer feature is that it is totally free and does not require to login. Just click the link and have fun. I'll keep it free like that, as long as I can.
I hope you like it guys because it is the best project that I have up my sleeve.
Enjoy!,
Pierre
As such, I’ve been working on the software to bring this vision to life, which is called Your Commonbase (a portmanteau of Commonplace Book and Vector Database).
In short, the purpose of the work is to create a data structure that works the way humans store, retrieve, and share information. By making these three elements as close to zero stress as possible, you catalyze creativity through remixing and augmentation of memories. My hypothesis is a lifetime building a Commonbase creates an idiosyncratic system, filled with the interpretations of an individual or a group. This individualized structure then creates demand that others want. I.e, a curation of all of the books you have read, organized by the marginalia you have added to them. This is a system people would pay for, and also a system that becomes more valuable over time.
I’ve been “working in public” by posting updates on my site [2], and am just beginning a small waitlist alpha testing phase (email me if you want in!)
It's a ton of work, especially with the number of systems - you've got combat, resource management, settlement development, food managment, espionage, diplomacy etc - these all need to play together well. And then you've got to add in the storyline, graphics, marketing -but I think I'm making pretty good progress.
I'm still using Unreal Engine 4 because I started work on that version and I haven't needed to upgrade to 5 since it's been released. I've got a free prologue that I'm releasing on October 1, so now until then is a lot of polishing to make it work well.
And I took my new quadcopter drone out and did some flying last night for the first time. As in, my first time flying a drone, ever. The results were... predictable. Let's just say, I bought a cheap (< $100) drone for a reason. This thing will wind up destroyed. In less than an hour I managed to crash it into fences, walls, bushes, cars, dumpsters, the ground, an armadillo, Elvis Presley, a 1974 AMC Gremlin, and Nickelback. Well, more or less.
It brought to mind this famous scene[3] from the movie Days of Thunder:
Harry: I want you to go back out on that track and hit the pace car.
Cole: Hit the pace car?
Harry: Hit the pace car!
Cole: What for?
Harry: Because you hit every other god-damned thing out there and I want you to be perfect.
[1]: https://jacamo-lang.github.io/
It's a ton of fun, especially writing Fusion 360 scripts to do all the parametric modeling of decks and molds. Then 3D printing molds with different parameters, pressing decks with veneer, making art for the decks, packaging, etc. It's an incredibly niche hobby but I've always found fingerboarding and making fingerboards to be infinitely creative.
It's the first solo project I've done in many years and I'm having a great time learning. I've also been able to get some great recommendations for myself and it's fun to use!
It's going to have online text courses with interactive examples and coding exercises, but I'm also in the process of adding video tutorials. These videos will be of 2 types: ones where I teach you the theory, and ones where we actually build a project from scratch.
I feel like CSS has always been something that was made to look harder than it actually is. In its essence, the syntax is very simple, and the vocabulary is quite basic. There are only a few things you need to know to be able to code an attractive and flexible responsive web page. For comparison, I find programming backends much more difficult.
Even though I've been working on this project for almost a year, I decided that next month will be the day I finally launch it.
I love Chicago and the incredible wealth of free festivals, museums, public art, music, parks, workshops, and so many more activities available to appreciate. I've found there isn't a good way to discover them though so I'm building the tool that I've always wanted. Right now I have a landing page up, have done some work on sourcing data, and I'm working towards defining a discovery interface.
Essentially combining the softness of fine count cotton with the weight and durability of a heavier garment.
Also a lot of ERP, warehouse, UI work and fun with Cursor.
I actually built it for my girlfriend who was writing a systematic review paper. She had to compare 7.500 papers against inclusion and exclusion criterias. She obviously did this manually because she cares about scientific integrity, but it sparked the idea to make an AI tool to automate repetitive tasks for people like her who would rather avoid programming. Now I just find it useful myself for a lot of ad-hoc analysis tasks like prompt engineering, rag tuning, and comparing model outputs from anthropic, openai, and google.
No microtransactions, no advertisements. No networking! The game is meant to function completely off line. The game will start with a mobile release in probably October and should be available on Steam shortly after.
I’ve been using Flame, Flutter’s game engine to develop my app. My artists are amazing and have been providing fully animated sprites using Spine. I’ve used Flutter before but this is my first time getting into Flame.
If you’re curious how it’s going, hop on the mailing list at https://danger.world to learn more. I’ll be looking for my initial playtesters here in the next few weeks.
Sometimes I feel I'm loosing the time, because nowadays there is a lot of AI generated content and even more competence in self-published books.
After a long walk against myself, of about 10 months, it's nearly finished (in my native language, Spanish). It still needs a few more reviews and retouching.
I got recently unemployed after +20 years as Linux sysadmin, and my wife is now unemployed too (after +20 years in HHRR), fortunately we have still a few savings.
I dream that it will (economically) work, but most of the time I intuit there will be less than a few sales from family and friends.
Depending on how it goes, I've already the script for the second and third parts.
In parallel I'm researching different ways to generate cash flow without working for another person. I would like to avoid going to search for a job in the current market of cloud, docker and kubernetes, as I'm more a hardware/colocation guy, and 99,9% of job offers request for docker/kubernetes.
It analyzes the social situation of a person, or group of people that live together, and forecasts those benefits that they could access. It is an old base code that uses OpenFisca for the back, and React for the front. However, it's quite fun though. Currently I'm working on make it more stable as there are still some parts that could break the process.
Previously used Reclaim but found $10 a month to keep 2 calendars in sync was excessive and the software was increasingly becoming more team oriented, no longer for individuals. Felt like I was paying for a product and also the product. I just needed to keep 2 calendars in sync, not smart meetings, analytics, integrations, etc. ideally I set it up and forget about it.
I really needed a way to sync my calendars privately, without all the extras. Now that Dropbox has purchased Reclaim, it's even more important I feel like my calendars are not spied on. I knew I could provide a similar service.
Here’s what makes me excited about it as a user:
Affordable: Just $20/year with a 7-day free trial. Cancel at any time for a pro-rated refund.
Privacy First: No need to store events on a servers enabling unlimited calendar syncs.
Working Hours: Adaptive feature that adjusts if an event falls outside your specified working hours.
Round Event Times: Opt to round event start and end times to the nearest 15 or 30 minutes for cleaner scheduling.
Invitations: Block time between calendars when you receive a meeting invitation that you haven't responded to yet. Decline it? The time block on your other calendar is removed.
My plan is to keep it small and focused, while also listening to user feedback for how you'd like to manage your personal and work calendars more efficiently.
Thanks for checking out my new project.
I took some time off from work to teach myself Rust and to build a WASM colony simulation game. You've got a colony of ants, they're in a cold, foggy crater, and you help them grow and survive. The simulation runs 24/7, like a Tamagotchi, but a bit more complex, like a simplified RimWorld or Dwarf Fortress. I am hoping to design game play systems which focus on mental health, self-care, addiction, motivation, and personal growth and to use the gameplay as a means of encouraging awareness in the player.
https://github.com/MeoMix/symbiants
I haven't added any features in a while, unfortunately, but it's on my mind. For a while I was just adding whatever popped into my head, as a means of learning Rust, and I naively thought the full idea would crystallize with time, but it hasn't. So, I'm trying to take a step back and figure out how to actually make a coherent game that does justice to the mechanics I want to see in my simulation. I've spent a lot of time thinking rather than coding, but I'm optimistic that I'll get through this phase sooner or later. I will admit, though, that trying to take a novel approach with game design is overwhelming at times. That's okay, though! I'm enjoying the process of tinkering with the project and will likely continue tinkering on it for many years to come as a creative outlet for self-actualization.
Recently I’ve been taking more of being able to flexibly run sql against this data, and this past week I’ve been working with d3 to make fancy sankey graphs to show income/expense flows. Quick preview here: https://demo.tender.run/reports/sankey
I thought it would be absolutely terrifying to be in front of a bunch of strangers and try to make them laugh, but it turns out if you're prepared, it's not that hard. Open mic crowds are benevolent and don't expect you to be the next Richard Pryor or George Carlin anyway, and I don't "engage" the public at all; I just tell my jokes.
I try to come up with new material each time so there's some work to do, but it's fun.
My first job was in a FinTech and the way production access was managed scared me. This is my approach at streamlining the process. Basically a PR review flow for SQL queries, enforcing the 4-eyes principle so you never accidentally can do a Delete * form users, forgetting the where clause.
The way position used to be stored in the search engine was approximate, using something like a bloom filter. With this change, they'll be stored in an exact fashion using a gamma coded positions list instead.
In the phase of dotting t:s and dashing i:s at this point. Hope to have it in production in September.
I'll do it anyway, as I need to learn to own where I'm at and have a backbone. :)
I'm learning the Zola static site generator and using to to build my blog at https://jeff-mitchell.dev. The focus is my mis-adventures learning the Rust language.
I've found it a challenge to get off my feet with Zola, but I'm slowly figuring things out. Little victory this evening, finally figured out how to get images linked in posts to render properly.
I recently left a startup I had cofounded after a few years because we were a remote team and I came to the conclusion that we weren't going to be successful without working together in person. I wanted to make this site to help people like me use location as an early filter to find good companies and teams to work with. Let me know if you have any feedback!
Link: https://camprompter.app
The insight is that if you read text from a tiny enough box right below the front camera on the MacBook, it appears as if you're talking to the camera.
Boom! Easy Eye Contact With Camera.
I scrunched up the Notion app and placed it as well as I could in the safe area to test it out for a couple of videos, but then I just wrote a web app because it seemed like the next logical step.
Once you paste the script and hit play... you can only see the text in the safe area so your eyes don't wander.
Fire up PhotoBooth and try it out!
PS: Press F for fullscreen.
It works with 90% of the bikes/motor brands on the market, so I assumed that some people here might be interested, if they got a non-functional batteries but they still want to use their e-bike?
We believe that everybody should have control about stuff they own, and we should fight against planned obsolescence!
Here are a few videos about our founder on the battery itself, why we built it, and how to assemble it:
- What is the Gouach Battery: https://www.youtube.com/watch?v=NsuW1NPkvNk
- Presentation of the pack: https://www.youtube.com/watch?v=mLoCihE0eIA
- Presentation of the fireproof and waterproof casing: https://www.youtube.com/watch?v=EDJpt7RDbRM
Here are the juicy bits: https://docs.gouach.com
We'd love some feedback from the e-bike DIY builder community
Oh, and it's launching as a Kickstarter in September and there is an offer for early-backers here https://get.gouach.com/1 for a 25% discount on the battery!
You can follow us on Instagram https://www.instagram.com/gouach.batteries to get the latest news!
The open source project: https://github.com/memfreeme/memfree
An Open Source Hybrid AI Search Engine: Instantly get accurate answers from the internet, bookmarks, notes, and documents. Obtain the most precise answers in the shortest time. With one click, AI indexes your personal knowledge base, eliminating the need to remember or manage it.
Worked as a software dev/manager for a decade, went through workaholism, burnout, then alcoholism, depression, all that. Doing a ton better now, and taking some time off to write about what I went through and hopefully help out others going through the same thing some: https://depthsofrepair.com/
It started out as just a prototype asking what happens if you task an LLM with generating crafting recipes for every combination of items in a game (which was already super fun), but it's exploded into all kinds of crafting, harvesting, and item manipulation systems that literally weren't possible in games just 5-10 years ago.
Now we're working on NPC simulations based on last year's Generative Agents: Interactive Simulacra of Human Behavior [1] paper. Dialogue and persistent memory are obvious, but we're also using the ReAct pattern [2] to give NPCs an influenceable decision loop that dictates what actions they take throughout each game day. And there's some other fun stuff like quest generation and using LLMs as a decision engine to determine if certain player actions complete these dynamic quests.
There's still a lot of work to do to make it feel more like a polished game, but we've been focused on the underlying systems and getting them feeling great and I'm really excited to see the game come to life.
[0] https://store.steampowered.com/app/3039840/Elixir_Emporium/
I get to lead a team of 175 doctors and students across premed, medical, and dental education. I am the first doctor + full stack technologist in the country. It's super rewarding. No funding, just off our immensely low price point that things are still growing quickly. All software written in house.
I've been working on and off on building an app for learning chest x-ray interpretation, but that's shelved until I finish applications.
TestFlight for now - try it out here: https://testflight.apple.com/join/wKY4eYV8
First time writing C and the task has been both daunting and rewarding, I am mostly working hls enc/dec and learning about av has been a ton of fun!
Most Inuit in Canada speak Inuktitut, which is a language with long words and two different writing systems: the latin alphabet, and syllabics. Syllabics are specially adapted for Inuktitut and well-loved by the Inuit, but unfortunately can be a pain to input on a computer, so often times the more cumbersome latin alphabet is used in casual writing.
Tom Scott has an awesome video about how syllabics work[1], but breifly, the shape of the character determined its inital consonant spund, and the rotation determines its vowel sound. So ᐱ = pi, ᐳ = pu, ᐸ = pa, ᑎ = ti, ᑐ = tu, ᑕ = ta, etc. The word "Inuktitut" becomes ᐃᓄᒃᑎᑐᑦ in syllabics.
Transliterating between the two is fairly simple, but there are edge-cases around dialects and whatnot. The more interesting problem from a technical perspective is having a web extension that can detect Inuktitut on a web page (in wither writing system), and transliterate that into whatever writing system the user desires, whilst never accidentally transliterating the other text on the page ("inhabitants" could be picked up and transliterated as "ᐃᓐᕼᐊᖯᐃᑕᓐᑦᔅ", for example, even though that makes no sense).
The project is mostly using Rust via WebAssembly, which has been a lot of fun to work with and has let me do some awesome things, like avoiding heap allocation and using compile-time hashmaps to do conversions on the text. The build system has to do a lot and I eventually settled on python. Right now I'm trying to wrangle JS and the DOM (there's a lot of edge cases to deal with), and that's been difficult as it's not my wheelhouse.
The idea is that phones have very nice cameras and lot of media processing power and such API could enable hackers to easily incorporate capable camera stack in to their projects without android development experience in any language they like.
There are millions of working phones in the drawers battery backed, with plenty of computing power, nice cameras, hardware acceleration, good connectivity (WiFi, ethernet over usb), that could get new purpose.
Previously I was working on automated photogrammetry rig and this is perfect use case for that.
If you see other use case for such API share your ideas.
If you think that's cool idea to develop and would like to help write me at hn@zelo.pl
It lets you view and edit plasmid sequences, features, and primers, and has some tools like automatic primer creation for cloning, primer QC, protein sequence viewing, and interop with common file formats.
Eventual goal before I return to work next year is to have a robot I can take on walks with me that will pick up trash.
Looking back, there wasn't any material to actually build a useful chat agent that resolved real-world problems. So I'm writing the guide that I wish I had when getting started.
I'll publish it on github and post here as more progress is made. I've created three parts. Part One is a non dev focused which explains how each of the moving parts work, what is Ai and what isn't. Part two gets technical and explains the tech stack. Part three is a bonus section that looks at how Ai assistants like google and Siri work and can be improved.
The working title is Automated Agents.
Edit: adding a link for those who want to follow along.
Hashtags can be used to split the text into sheets and columns, if so desired. Besides jotting down quick thoughts, this is very handy for short-form journaling such as tracking expenses, workouts, mood, period, weight, diet, etc., with the added bonus of easy charting and summarization from within the spreadsheet. It also supports pictures and other attachments that are uploaded automatically to Google Drive and linked into the spreadsheet.
Feel free to check it out, all feedback is appreciated: https://t.me/gsheet_notes_bot
Scams usually start with messages. Predominantly SMS/email sent with links.
ML models exist to determine this, but we are prototyping using Gemma2 2b to utilise its natural language understanding to see if a better firewall could be built, one which users can talk to.
It has been surprisingly easy to get started on this project: https://ai.google.dev/edge/mediapipe/solutions/genai/llm_inf...
We want to subsequently build a firewall for Calls, too; but that's a more challenging problem given its dynamic and online / realtime nature.
Hope to integrate these two features in our existing network firewall open source app.
A million ways to do this, of course, but I'm focused on using wireguard so that eg only my wireguard peers can get access to my local service, and for internal traffic (ie vpn).
At the moment I'm settling on having a simple script that I can run on a host alongside wireguard. The script will function like `wg-quick`, parsing a wireguard config file and handilng the routing stuff behind the scenes, and returning a cleaned up config to be passed to `wg`.
Ideally, the wireguard configs could be generated by some other tool or service, like https://wirehub.org, and automatically fetched and applied to the running wg interface.
So, a one liner on a server with a public IP and the services exposed by your wireguard peers can be accessed via a custom domain name while still respecting internal wireguard routing rules (based on AllowedIPs).
If anyone finds this interesting and wants to chat about it, I would love to! My contact info is in my profile.
I'm currently thinking about how to robustly Integrate different edit sources like iCloud, obsidian sync, git, and yjs updates. I think it could be cool to create a crdt persistence format that can live alongside markdown files (like note.md + note.md.crdt) to support edit history tracking from multiple users and their devices.
https://www.punycoder.com/ (a tool for Punycode to Text/Unicode and vice-versa conversion),
https://www.htmlwasher.com/ (a tool if you have some dirty HTML and need to clean it up),
https://www.htmlcorrector.com/ (a tool if you have some not so well-formed HTML and need to fix it),
https://www.htmlenc.com/ (a simple tool to HTML encode (escape) a text),
https://www.urlenc.com/ (a simple tool to URL encode (escape) a text),
https://www.gguid.com/ (a customizable GUID generator),
https://www.64baser.com/ (a Base64 and vice-versa encoder),
https://www.hexator.com/ (a Hex and vice-versa encoder),
https://www.cescaper.com/ (a C and C++ string escaper and unescaper) https://www.htmlinstant.com/ (a simple HTML editor).
https://www.notationer.com/ (a tool to switch across (code) notations, like camelCase, PascalCase, snake-case).
Frontend is mostly in React/Next.js (Some older projects have frontend still in the old ASP.NET MVC and Ajax :-)). Where the backend is needed, it is written in .Net/C# and running on Azure.
Just today, I finished first working version of the new compiler (https://github.com/ubavic/mint). It is written in Go, and there are lot of things on the TODO list, but it works :)
This is actually the second compiler for the atex. The first one was written in Haskell and compiled fixed document schema. I used it for writing a book on Haskell (https://github.com/ubavic/programming-in-haskell).
An example of it in action: https://streamable.com/2mxktc
Source code: https://github.com/beef331/potato
https://m.youtube.com/watch?v=S-D2dOTHpzo
Watch my blog https://tunn3l.pro for the full project release!
Cheers!
I released the first playtest of the alpha a few days ago, which you can try directly in your browser:
Some of the features:
* OpenID Connect based Identity Provider
* OpenLDAP synchronization - currently supporting users and groups synchronization
* MFA with TOTP, email, WebAuthn / FIDO2, crypto wallets
* wireguard client GUI integrated with OIDC, supporting multiple locations (https://defguard.net/client)
* secure enrollment & onboarding
* yubikey provisioning
Our github: https://github.com/DefGuard
(edit: formatting, github link)
The first video is here: https://youtu.be/W0_3rzvq9Ks (the second is coming out tomorrow)
And it's on GitHub here: https://github.com/masto/LED-Marquee
I also recently left the Big Tech world after 11 years at Google, so I'm trying to figure out what comes next. (I don't think I can make professional YouTuber pay the bills). If it's not inappropriate to mention here, my resume is at https://hire.masto.me/
Finished a feature to convert Markdown's fenced divs nested within blockquotes into the following XHTML:
<blockquote>
<div class="name"> ... </div>
</blockquote>
> ::: simul
> Simultaneous section 1.
> :::
>
> ::: simul
> Simultaneous section 2.
> :::
65 Words is an anonymous challenge to write 65+ words daily in the language you're learning.
The real focus isn't just on improving your writing but on having the opportunity to think about how you'd express something without the pressure of the moment.
There's no hidden science behind choosing 65 words. I found it’s achievable even on a very busy day.
The goal is to help learners practice daily and build confidence through consistent effort.
Recently I got a DIY strobe tuner I've been working on up and running, which pairs nicely with the Mosaichord.
https://www.youtube.com/watch?v=Lkz-fZN54OI
The Mosaichord works pretty well these days, and I've assembled a bunch of them. Just got a new batch of PCBs yesterday to make more. There's a basic website up, but I haven't yet set up a proper web store.
I left my home country for Portugal when I was 18, and was refurbishing iPods for money while living in a hostel. In my free time, I developed a Bluetooth kit, a crude first version which required hours of hand-soldering soldering, CNC milling, etc. to manufacture. While selling these I used the money to prototype and manufacture the second revision.
That's the version you can purchase today, which can now be assembled in about five minutes. This has allowed me to scale the business, and now we're a few employees big, shipping worldwide. A few grey hairs have appeared at 22 but finally I can relax a little.
I’m using this precious expanded free time to compose music, my primary form of artistic expression.
Software-wise, I’m on a platform team for a large company. I’m making some performance enhancements to our http client, and a plug-n-play library for app devs to easily integrate LLMs in their products.
It's not targeted at the 'cloud native' crowd, but rather more at traditional on-prem infra.
It is all manageable via an API and provides a Kafka API for streaming data (RedPanda under the hood).
Name: Morio Documentation: https://morio.it/ Code: https://github.com/certeu/morio
Note: here is no commercial angle here. This is an open source project of (the CERT of) the EU (license: EUPL).
http://github.com/leftmove/facebook.js
I was working with the Facebook API for another project, but I was surprised that a website as popular as Facebook had exclusively old API wrappers for it, that all required too much setup, and were cumbersome to write code with.
I wanted a better approach using the standards I have come to appreciate with modern API wrappers, and so I decided to start work on a more intuitive, faster approach for the Facebook API.
My initial launch goal is to allow for programmatic authentication, posting, and commenting, all with one setup command through a CLI.
I have decided to shelve my work on architectures that are biologically inspired for now. I was getting reasonable results with spiking neural networks and evolutionary training, but there are so many hyper parameters to think about and how they behave over time is really hard to predict. I was also struggling deeply with how to manage topological concerns like network growth over time.
With interpreted evolutionary programs, the memory access patterns are so much more ideal with the program counter stepping through (mostly) contiguous bytes vs totally insane recurrent spiking neural access patterns. You get so many more generations & candidates evaluated per unit time that it can make previously apparent "dead ends" viable, simply because you don't need to have extreme patience to find out anymore. I am discovering that iteration speed is the most important thing in this arena. The faster you find out how bad a certain parameter adjustment is, the sooner you can get to the good ones.
I am also working on an unrelated contract to integrate some back office banking systems. Not much worth discussing there.
There are some players in this space already, but we have repurposed some OEM grade hardware that we usually reserve for the big players and so we can offer differentiating features such as DC fast charging, bidirectional/V2X, and things like this to the retrofit market.
B2B to start with.
The app's website is vid-note.com, if you would like to check it out. The value of the app is that one no longer has to watch endless videos but can find exactly what one needs about a certain topic within the notes that we offer. What is good to know is that most of the notes are centered around business and self-improvement for the reason that we, the founders, are interested in these topics. We understand that many would advise us to let an LLM write the notes, but for this project, we have decided to sway away from the flawed results that LLMs bring with them.
This is the first venture my business partner and I have decided to pursue, and we look forward to finding out if we must pivot or preserve. For those wondering, the app has been approved for external testing on the App Store, but we have decided not to make the jump yet, simply because we still have to handcraft a few notes ;). We predict that a release on the Play Store will soon follow. Cheers.
Using PHP Slim Framework, MySQL, and vanilla Swift/Java. Prioritizing reliability and efficient sync between local storage and server. Currently finalizing API endpoints and feature matching the web app. Next up is adding a subscription model.
It's challenging as:
- Users want to use the app offline which means we need to sync
- We have to match our existing features on [2] in the first version and evolve the API and database to support reliable sync
- Users want to also track their cardio too e.g. Running Level [3] and Rowing Level [4] but that will have to come out in a future version
[1] https://strengthlevel.com/
My ultimate goal is to bring brain signals to the browser and develop neuro-apps, all in public. I share updates weekly on LinkedIn at https://www.linkedin.com/company/mind-tether
Next, I will try to build something using RL next but try not to use the Gym/Farama stuff to force myself to learn this from scratch.
- Here is a short video explainer of WHAT we do https://www.youtube.com/watch?v=PjXwUpBp_Kc
- And here's WHY we do that https://www.youtube.com/watch?v=-37CA6-m1Yk&list=PLVJfpLOJQj...
Our first campaign to fund Ethan Perlstein's research to find cures for rare diseases officially starts on Sep 9, but we already opened a pre-campaign. This one is a one-time commitment and it's heavily leaning on the charity side.
- Here's a landing page of campaign https://marabou.pro/perlara/
- And a video where we talk about it https://www.youtube.com/watch?v=YlO8IBQDMBY&t=9s
So my days are now filled with talks with rare disease foundations and influencers as well as usual people who may be willing to support us.
Happy to chat with anyone who is interested either in Marabou or the campaign itself!
Not sure if I caught the train but still. My name is Max, a software developer based in Central Europe, and I'm working on my course "Building Command-Line Interface App from Scratch in Go" (this is a working title).
In this course I want to teach how to write CLI application from scratch in Go without any external dependencies (exception: driver for DB).
This is not another one course about how to write CLI app using Cobra or any other command-line builder library. Instead, in my course you're gonna write your own Cobra-like command-line builder library (package) to build command-line interface application from scratch.
Thus, you will know how to build an API for building CLI like Cobra and you can use it later on for your future projects. I will also talk about Command-Line Interface Guidelines showing how to make powerful, useful and easy-to-use command-line application which you can use every day in your workflow.
The course is still work in progress and I don't even have a web page for it. But I do have my newsletter where I share best resources about Go. By subscribing to this newsletter you will be notified once course is ready to launch. Also, everyone who's subscribed to my newsletter and will be in first batch to buy a course gets a 50% discount.
So, if you are interested in this kind of stuff join the waitlist for course in my newsletter on https://kovalevsky.io
P.S. I also have daily base newsletter about Go - Daily Golang. If you subscribe to this newsletter you will also be notified once course is ready to launch you and will get 50% discount to buy the course - https://kovalevsky.io/daily-golang.
I am using the Godot engine: https://godotengine.org/
The playable demo is already available today on Github: https://github.com/Eccentric-Anomalies/Tungsten-Moon-Demo-Re...
We're releasing the demo on Steam for the first time in a few days, followed by an early access version probably in October: https://store.steampowered.com/app/3104900/Tungsten_Moon/
Keeping key bits of the idea to myself. If this admittedly vague idea excites you let's find some time to talk.
During university, I spent some time working on an AI agent to play Dominion, but a very large part of the work was building a way to simulate the game.
The goals are:
- Develop an engine that's efficient enough to use in simulations (for training AI agents or analyzing the game). - Still emit events that can be used to visualize the current state of the game when real people are playing. - Create primitives that are easy to distribute across a network for remote players/agents.
So far I just use it for my company billing, but it's quite delightful for me: My process is to save the services we provided in CSV files (exported from time tracking in org-mode, but you could use any tool that supports exporting), then I run this over it and it creates PDF invoices, and stores all the data (which I query occasionally, e.g. for quarterly VAT payments).
Technologically it's weird, the invoice templates are written in LaTEX, and the code is in Common Lisp.
I don't think about turning it into a business, but I think I'll open source it once it's a bit less messy. For now, I just focus on implementing the features I need (about 1-2 per quarter now).
I kinda doubt it, but if anyone is interested in command line billing tools, I'd be excited to talk about it. Contact info is in my profile.
I'm currently working on an app that provides a step-by-step guide to drawing from a reference photo. My goal is to make the sketching process more accessible and enjoyable for everyone.
javascript:void function(){ javascript:(function(){ var selection = window.getSelection().toString(); if (!selection) { alert("Please select some text on the page."); return; } var encodedSelection = document.createElement("div"); encodedSelection.textContent = selection; var processedContent = encodedSelection.innerHTML.replace(/\n/g, " <br></br> "); var words = processedContent.split(" "); var formattedText = ""; var speechContent = ""; for (var i = 0; i < words.length; i++) { var word = words[i]; var chunkSize = Math.floor(word.length / 3) + 1; var boldPart = "<span style='font-weight:bolder'>" + word.substring(0, chunkSize) + "</span>"; var lightPart = "<span style='font-weight:lighter'>" + word.substring(chunkSize, word.length) + "</span>"; var formattedWord = boldPart + lightPart; if (word.endsWith(".")) { formattedWord += "<span style='color:red'> *</span>"; } formattedText += formattedWord + " "; speechContent += word + " "; } var newWindow = window.open("", "_blank"); newWindow.document.write("<html><head><title>Spoken Content</title></head><body><input type='range' min='0.1' max='10' value='1' step='0.1' id='rate-slider'><p id='content' style='background-color:#EDD1B0;font-size:40;line-height:200%25;font-family:Arial'>"%20+%20formattedText%20+%20"</p></body></html>");%20var%20rateSlider%20=%20newWindow.document.getElementById("rate-slider");%20var%20utterance%20=%20new%20SpeechSynthesisUtterance(speechContent);%20rateSlider.addEventListener("input",%20function()%20{%20utterance.rate%20=%20rateSlider.value;%20window.speechSynthesis.cancel();%20window.speechSynthesis.speak(utterance);%20});%20window.speechSynthesis.speak(utterance);%20})();}();
https://github.com/WillAdams/gcodepreview
big things are reading:
_Abraham Lincoln and the Structure of Reason_ by David Hirsch and Dan Van Haften
and
_A Philosophy of Software Design_ by John Ousterhout
and getting the code to an actually useful state and then creating CNC joinery which isn't possible with other tools.
Very early stage right now but I hope to release alpha soon. I'm already using it privately right now but there's a lot to do to make it user friendly.
The idea is also to merge in a single app many needs of events, like group chats, photo albums, expenses splitting, and I'm planning to add more. I realized groups have similar needs to events, so I added them as well (think small private groups, not massive communities).
All of this is stil very early in development and it definitely needs some more work before a Show HN, but we've found it very useful for weekends and holidays already.
Consider it an open beta, so feel free to give it a try and give me your honest feedback :)
Goal is to automate or reduce the grunt work oncall engineers have to do.
Code is here: https://github.com/opslane/opslane
In short, if you want a person with Language A experience, but Language B is close enough, my software will provide the signal that the candidate is good enough for a closer look.
Example: PHP and Go devs are likely to have similar approaches to how software should be built. Ruby and Go, perhaps not so much.
There’s more nuance to it, but this is the idea.
Working as a consultant, but no extra energy or will to create anything outside of work.
Kids are getting bigger, and I have time.... but no will, or maybe it's just fear, to start anything.
My main barriers are 1) writing discipline and 2) some obsessive need to link/cite every claim I make. This results in me taking months to write an article, be it due to laziness or constant rabbitholes/google scholar searches about an idea.
Any clue on dedicated disciplined writing time or even considering "reducing" the rigor of writing?
And my long-time project. A Dungeons & Dragons utility to use Thermal Printer for handy Printouts: https://github.com/BigJk/snd
I looked into pre-built systems like Synology and QNAP, but they seem too proprietary and limited, and somewhat expensive for the hardware they have, so I decided to build my own, but trying to decide on all the HW and SW is a real task: btrfs vs ZFS, Intel vs AMD, RAID levels, etc.
I've always found adding custom metrics and monitoring to applications to be a big hassle, so I'm experimenting with one that uses the log stream instead of agents/daemons.
I basically want that to be open and free. I'll have an app to easily create albums, and have the ability to connect to any storage or montage service you like. You could even make your own app connect to your own storage and montage service, so it can be completely diy. I have a few write-ups but I'm transferring it to notion at the moment so I won't link it just yet.
My other idea is a "bill hamper / consolidation" service, that I'm doing for my sister. She pays me a flat fee each week and I pay all of her bills for her. Gives her peace of mind and allows her to save some money without stressing on paying random bills
My part has been to generate news headlines and weather maps using current news and weather information. The most interesting part for me has been deciphering the 1980's era graphics format, NAPLPS, which Prodigy uses and making a library to write files in the format. I treat the file format as a data transformation, taking the NAPLPS file generated so far and appending more commands to the end of it. The commands naturally pipe into each other in an idiomatic way in Elixir.
I would love feedback on the new site's direction - https://www.getflowcode.io/ (still WIP) Here's the old one, for reference https://flyde.app/
I’d love to get feedback on what can be done to improve the website to feature prominantly for local search when people need rubble removal near them.
My quests (goals) with this digital garden are:
1. Publish more than I did when I just had space for essays, which hopefully leads to…
2. Getting more input from people on my ideas
3. And have fun futzing with my digital garden technically
So far, so good!
My app is called Project&Cut and can be found at https://projectandcut.com.
This is my latest article, just published 2 days ago:
https://www.saintbeluga.org/follow-the-evidence-wherever-it-...
I found it amazing the importance of the small details in old woodworking tools and how toolmakers solve problems and improve tools. Particularly in very simple tools.
Outside of that I am also building a gym up for myself which I could use without glasses (all apps I tried have tiny buttons or controls that I cant see in the gym)
I am releasing everything for free and the first "instrument" and sample pack with previews/videos is already available here: https://neuromorph.gumroad.com/l/alien_vocalization_study_1
It's a demo of an idea. It could be an app too, but I'd much rather it be a CoreOS service that is user controlled.
Looking for organizational and privacy first support soon.
I properly animated a model with blender, and am able to move the parts I want accordingly. It was not easy but I made it.
I also managed to implement the recoil algorithm I want.
I modeled guns.
I will probably struggle to implement server reconciliation to mitigate lag, I found 2 modules that does it, one in C++, one in gdscript.
Obviously the gdscript one is slower, but it's hard to know how much slower, since obviously a game like counter strike is very fast paced.
I also want to implement a panini/wide angle camera, also found a module for it.
Moving slow with bad mental health and anxiety, but having fun.
So I spent some time trying to make an app that allows you to do that, but can't shake the thought that such an app would work better as a Blender addon/plugin rather a standalone app. And I also am trying to figure out how people work with 3D currently, to see if such a tool would even be an improvement over existing tools.
They're driven by what problems I see at work and while playing with other weekend projects so coverage may be asymmetric (some parts are well covered, others that should be are absent <<ie. documentation>>).
I find definition of success as just "working on it" very pleasant.
It is an experiment for the exploration of the free aspect of blockchain storage, where the book library is now permanently hosted on the blockchain, and it's content voted upon by its members.
The licence is based on creative commons, and enforces that the data is decentralized and doesn't need any account, or wallet, to use.
There is also the creation of a new label, the D-Safe label, for a safe experience across generations.
A fun side project - which I have been working on for several years now - rewriting it fully already 10 times - and been restructuring my mental health around it.
Evy is an app for collectors, both professional and casual, to help them keep track of their items and share it with the world. Currently there are either overly complex inventory management systems, which are overkill for casual collectors, or generic social platforms, which are "fine" for everything, but great for nothing. My goal with Evy is to fill this gap by offering an easy way for any collector to manage their items.
In short, it is an asset management system tailored specifically for collectors.
A demo should be available very soon, meanwhile you can wishlist it on Steam: https://store.steampowered.com/app/2945950/Dice_n_Goblins/
It's turning into a full time side gig that's paid off in handful of batches (~$25K and running)
I'm trying to become a loyalty vendor API.
Would really like to work on my own interconnected systems of blogging tools and social media. Kinda a blend between having your own site and myspace? Not really targeting a market or anything like that, just think it would be cool to do for personal use and sharing with family and friends
We're offering online memberships, event management, and a member database packed with features. Membership management is a crowded space, but it's also a low-tech space with lots of sleeping giants not willing to iterate on their product.
It's been a really fun project so far and even more rewarding to see clubs using embolt for their daily operations.
It's built using C# and WPF, and a related project I work on is an open source MVVM framework called UpbeatUI for making WPF apps that behave vaguely like mobile apps. It's for apps that have a main bottom layer and modal popups that float above and can be closed by clicking/touching the background. Pulselyre uses UpbeatUI, and I actually originally extracted UpbeatUI from a much older version of Pulselyre.
Bungee is unique in its quality, performance and controllability. Every media player, DAW, video editor should use something like Bungee for smooth scrubbing and audio slomo.
Try it with one click: https://bungee.parabolaresearch.com/bungee-web-demo
So far we have several licensees of the Pro edition with more on the way.
This is a problem that every refiner solvea to select raw msterials. They traditionally solve with local optimization methods (LPs) that rely on initial guesses and return suboptimal solutions.
For the same refinery we can find raw material acquisition solutions with 2-5% higher profits.
We made a cloud platform for it and hoping to change the landscape.
I've been working on an exciting project called DataSignal, where I'm developing a sophisticated Named Entity Recognition (NER) and enrichment system. This system is designed to identify and extract key entities from blocks of text, such as names, locations, or organizations, and then provide rich contextual information about them. The goal is to transform raw text into meaningful data that users can easily understand and leverage in various applications.
The real magic of DataSignal lies in how it delivers this contextual information. It can be seamlessly integrated into your workflow through a browser plugin, which instantly highlights and explains entities as you browse the web. For developers, DataSignal can be embedded directly into a webpage via JavaScript, enhancing the content on the page with dynamic insights. Additionally, it offers a REST API, allowing for flexible and customizable integration into other software systems.
Whether you're a developer looking to enhance your application with smart text analysis or someone who wants to streamline research and data gathering, DataSignal is poised to bring a new level of insight and efficiency to text analysis. This project aims to make complex data more accessible and actionable, transforming how we interact with and interpret information.
Currently I've built a call/response RPC abstraction with it to use it for another project that I'm working on. Eventually I'd want to add a streaming message system beyond the basic call/response pattern. I also might like to ship some easy setup to bridge between these protocol abstractions and ZMQ, JSON-RPC, be able to behave like inetd, etc.
It's functional enough to use in non-production environments but not completely seriously yet. I'm working out the kinks and improving the DX by using it for that other project. There's still some easy improvements around the byte encoding that I still haven't gotten to yet.
— I’m working on a second newsletter that’s more discovery oriented but still focused on the personal web space
— I’m setting up a private, invite only discourse forum to create a space where people can hang out and connect in a more meaningful way
— I’m working on a new studio site in collaboration with a friend
(Source code at https://bgammon.org/code)
Happy to talk more about this with anyone. If you're at all interested in "software to make your team run better" I'd really love to hear what you're having trouble with and how your team runs, maybe I can figure out a way to help you, even if it's not at all related to what I've built already.
It's been sitting there for years, and I finally found health, time, and dedication to withdraw it.
On cs.money I had $35k and on Steam balance I had $12k.
The only way to get money from cs.money trade mode is to buy skins and sell them somewhere else.
On Steam, on the other hand, I buy Steam Decks via Steam balance and sell them locally.
I have sold 3 in the last 2 days.
Last 2 days I've been taking a short break from coding a bot for cs.money to recover some health back (I have RSI).
[0]: https://ai.meta.com/research/publications/revisiting-feature... [1]: https://en.wikipedia.org/wiki/Q-learning [2]: https://arxiv.org/abs/2402.04494
* Analyzes candidate information
* Examines job descriptions
* Generates unique CVs and cover letters for each job
* Answers specific questions that recruiters ask
* Automatically applies to jobs
There are already a lot of services to right-size servers, but I want to focus on small teams that might not have a full-time devops yet and lack the time to check every server manually.
The app will be simple with an educational part, it should be like having access to continuous audits on our infrastructure, just like the CI/CD tools we use on our codebase.
It already helped me at home to reduce the load on my oldest Raspberry Pi by shifting some services to a newer one that was mostly idle and right-sizing some VMs in the Proxmox cluster of the rack in my homelab. Having a green check next to each server name is a good motivation!
Another thing that is scratching my own itch is server discovery (especially in a multi-cloud environment) and integrations because it's just so easy to forget some servers in a cloud when you do a lot of experiments and keep paying for them for months.
The MVP is getting ready to launch, I'm making the landing page right now, and then I hope to find some early adopters to iterate on the app and make it useful to other devs. I'm also thinking about open-sourcing it.
I'm interested in having conversations to discover the pain points of other people in that space, so feel free to reach out to me!
- Analyzing grocery pricing patterns for 7 Canadian grocers. I took care of the data acquisition. Now it is time to speak with economists / data analysts who can make sense of it. Are you interested? Please get in touch!
- Gathering and archiving all graphical material from the "Printers International Specimen Exchange". The gathering is done, now working on a writeup.
- ... various side quests from the above. Incl. a tutorial on making mobile-friendly imagemaps
Basically a pretty+easy editor! I'm really enjoying building the small features https://owri.netlify.app/share/25a7d985-7cd5-4217-9766-f5296...
Feel free to tell me if there's anything you want!!
New idea: So currently I'm trying to learn leetcodes, but thinking of alternative/structured ways to do it. What if there's an AI-powered book that's basically like your personal Wikipedia into anything?
You start with a topic, it generates you outlines/courses -- and what'd be great is if this was social, like you can see what everyone else is learning about, what nodes they're expanding...
and make notes on topics and ask questions
The only difference here is that the program will switch on the fly between different algorithms depending on which one that can compress file smaller.
It can compress 1 GB file (enwik9) down to around 230 MB. Pretty good I guess for something that I worked in my spare time.
I'm not publishing it yet, since I'm still experimenting with it a lot.
I’m currently polishing the logs and reports as much as possible, and then I’ll add support for more tools (Kafka, Rabbit, Nats, etc.). I have tons of features in mind to improve UX, speed, and bring more value to the tests.
Down the line, I want to find a business model for this tool and sell it, but I need to do a lot of thinking on this side since I’ve never done this before.
The set of puzzles is really tickling my fancy at the moment, for some reason.
It's a fun challenge, because I need to figure out how to reliably determine the matching game for files like /pub/johnny/midi/games/sc2kmenu.mid, and also to show them in a way that fits nicely in the UI.
For $100 we ship a vending mechanism for you to put on your own custom vending box.
We then provide online marketplace and payment services (Shopify for vending machines) so your customers can use their phones to buy products from your vending machines.
Great alternative for products that don’t justify big $10k vending machines.
I'm still contemplating what I want the final form to look like, but currently it's a 5 kg load cell and a floating idler and an rp2040 reading values via hx711.
https://github.com/fedi-e2ee/public-key-directory-specificat...
Think: Encrypted DMs for Mastodon. I wrote several blog posts about the project and why it matters.
https://soatok.blog/2024/08/21/federated-key-transparency-pr...
https://soatok.blog/2024/06/06/towards-federated-key-transpa...
https://soatok.blog/2022/11/22/towards-end-to-end-encryption...
Eventually I plan on doing a "Show HN" post when it's built and close to feature-complete.
An automated dubbed translation service. Supports translation into English, French, Spanish, German, Portuguese, Chinese, and Japanese. The other options on the market often have jarring audio artifacts and glitches, which Audiomatic avoids.
My friends and I are currently working on improving the voice quality and adding new features.
It is implemented in Go, Python and JS, with support from Postgres, TimescaleDB and Kubernetes.
This is an open-source alternative to paid services like:
AssemblyAI Deepgram Gladia Google Cloud Microsoft Azure RevAI
With trintAI you can build your own custom workflows & integration for speech-to-text transcription.
The main features are:
- Speech-to-Text Transcription: Converts audio files into accurate, readable text in real-time.
- Summarization: Provides concise summaries of long audio files or transcripts. This feature extracts the most important information and key points from the text, allowing you to quickly understand the main takeaways from meetings, calls, or any extended audio content.
- Sentiment Analysis: Detects emotions within the transcribed text.
- Language Identification: Detects the language spoken in the audio file and can transcribe in multiple languages.
- Diarization: Identify and distinguish between different speakers within an audio recording.
Give it a try! it is open-source. Your feedback is very appreciated!
Cheers
Works great with physical maps, screenshots of maps or downloaded pictures of maps. NavigateAnyMap.eu (Android only, sorry!)
Free as in beer, will tell you is uses ads, but have not added them yet (and will make a paid, non-ads version first).
We’ve been hand coding full-service integration workflows for 2+ years on a large B2B agency model, and in parallel have been building a platform using what we’ve learned to build better abstractions.
We’re at an interesting point where almost all new workflows are being built on the platform, and we’re serving over 100 clients.
We’re prepping for public release in the next few months.
In 2022, I wondered what it would be like if my manager could leave comments directly on the live website for all the edits, rather than sending a bunch of screenshots and videos. That’s how I came up with the idea for a "simple" browser extension that lets you leave comments on any site, anywhere on the screen.
There were several challenges along the way:
Attaching Comments to the Correct HTML Element: I initially struggled with ensuring comments were attached to the right element, as relying on X/Y coordinates would not be responsive. Now, I use a combination of element IDs, classes, outer HTML, and attributes. This approach works correctly 95% of the time. Do you know of any other methods to find the correct element in the DOM?
No Third-Party Libraries: Extension development only allows pure HTML, CSS, and JS, with no external scripts from CDNs. Building a text editor from scratch was one of the most challenging parts.
Real-Time Functionality: Keeping the extension's background page active was tough because it deactivates when not in use, making it almost impossible to maintain a socket connection. I wrote code to wake up the background page and reconnect the socket whenever it goes to sleep.
I stopped all my freelance projects, and now three other people and I work on this tool full-time. We recently became #1 on Product Hunt: JustBeepIt.com
There are still many issues we're working on, such as using it with iFrames or inside scrollable objects, but we're tackling these challenges one by one.
Currently experimenting with a game build around the old classroom method of Total Physical Response — you can see a very early prototype here: https://kolja-sam.itch.io/the-tpr-game-gala
Besides that our team is always working on cool stuff for our clients. Lots of interesting work these days “in” AI, plenty of cool stuff going on in health tech.
Got something cool you need devs for? Always happy to talk shop — wyatt at apsis dot io.
There's a gap for email monitoring. How many times did you have to awkwardly explain why the notification emails in your app stopped sending a month ago? Or that time when the weekly summary report did no go out on Friday, and on Monday you were chewed out by the boss? Or when Jerry went on vacation and neglected to send the weekly marketing campaign email.With
[I'm building https://wasitsent.com - you get an alert when your emails *stop* sending. It's like an uptime monitor for email sending. If something breaks your email sending process, we'll let you know.
If you would like to see such a service, or have feedback (positive AND negative) about it, please let me know. If you heard of other companies offering something like this, I'd love to know as well.
Our approach is a bit different since unlike making a better copilot or a chat experience we are building workflows which encourage engineers to work alongside AI and not just offload tasks to AI.
I've been spending a lot of time in México over the last year and have been working on a photography project surrounding the VW Beetles (Vochos as they call them) you'll find all over the country. I'm inspired by the range of character of the Vochos and how they've become ingrained in the country's cultural landscape.
I realized I had a sort of hyper-niche problem where I was locating Vochos I'd like to shoot when I wasn't carrying my camera. I was getting annoyed with using Google Maps to list locations I'd like to return and shoot so I spent a weekend putting the app together and turning it into a game.
After using the app for a few weeks I've realized the spotting or hunting of the cars is the part I enjoy the most about my project and when I have the time I'd like to publicize it so others who find themselves in Vocho dense area can check it out if they'd like!
I’m using it personally as a transcribed voice-note system.
https://athens.winterdelta.com/playlist/536a9778-cd9d-4491-8...
I need help finding a full time job in Boston so I can move there.. Facing a catch 22 with apartment applications. I have a tooth brush and will scrub makefiles for a living wage, I'm not picky! (resume in profile)
Custom 3d engine, opengl / webgl, Lua scripting, voice chat. Mostly in C++.
Decided on a GPL/AGPL/ODBL license for the scraper/website/dataset. https://blr.today/license
My Core-thesis[0] on why I think this is worth building:
1. The vast majority of events in aggregator websites are low-quality, and often filled with spam. The aggregators make money by listing lots of events.
2. Small venues hosting cool events will not always publicize them on the aggregator platforms.
3. The best UX for event discovery is your existing calendar app.
4. Events are highly structured data - but this is often not captured.
I've been wanting to build this for almost 6 years now, finally getting around to it.
I updated my ideas repo last week with some more of my ideas.
https://github.com/captn3m0/ideas
If any of these sounds like fun, take a look:
* A physical variable fuzzy clock
* A curl impersonation proxy
* A Whisper UX Design Pattern
* Mobile App Traffic RE Platform
* One-Page (RSVP|EventHosting) Platform on Edge Compute
* Price Index for Indian Grocery Websites
In the best case, the developers of e-commerce website create unit-, integration-, and functional tests to make sure, they do not break existing functionality.
But a website is not only code. Its also data and configuration.
Since these types of configuration (e.g. prices, legal texts, bonusses) can change quite often and can get quite complex from the business perspective, it does not make sense to have developers create automatically executable tests to make sure, the website is configured as it should be.
And its often not the developers who change the configuration. Maybe its a product manager, a marketeer or the ceo.
These people often do not know if they configured something correctly. In additiom, they often don't get any notification, if an process, that is important for them, breaks due to a misconfiguration or some bug in the code (not every code base has a test coverage of 100%).
So, I am creating a No-Code Black Box E2E Monitoring Tool, that the process owners/configurators can use to regularly check, if everything is working fine.
Happy to elaborate if anyone is interested, I will also write about it on my blog (https://watwa.re) at some point
Spec generation from request logs, automatic schema generation and validation, test generation (eventually), totally offline, no accounts or cloud sync necessary!
Been taking longer than I hoped but should be released soon (next week or two)
- TheFile.Ninja, which is a file manager with the Everything indexer built-in for extremely fast information retrieval. This allows you to quickly run file queries, and these queries can then be saved or even added as folders in the file system. When you open the folder, it automatically fills with the contents from the query. I have also created a service with LLM/AI that translates plain text into Everything's query language. This enables you to build very complex queries directly from plain text. For example, you can ask if any directories contain a certain file, and if this file is found and contains specific text, it will be displayed. You can learn more about the project at: https://thefile.ninja or watch https://youtu.be/JREufgkf5pk. This summer, I also built three smaller games (not fully finished but almost):
- ThrustMe!, a space/underwater cave exploration game (https://youtu.be/M0d7CSpEJ1E). It works on Android, Web, Windows, and possibly soon on iOS.
- MergeQuanta, a Tetris-like game where you merge matching blocks (color + shape) to make them disappear, taking surrounding blocks with the same shape along with them. There are also cement blocks and bombs. The game works well with touch and stylus but also on regular computers (https://youtu.be/VXvpzhi8ySE).
- Flip the Maze, a simple multi-dimensional maze game (https://www.youtube.com/watch?v=qOVRB0uAPIE).
What do you think about them?
You can see some of my world-noodling about aliens that don't need to breathe at: https://rznicolet.com
I've been plotting a series of hard sci-fi novels entirely from the point of view of very not-humanoid aliens. In this case, hard SF = all real physics except for FTL. First manuscript complete. Humans not included in the main series, except as passing footnotes. Let's just say that when I originally started, the pandemic had me feeling a touch misanthropic. We'll see when/how I actually publish these.
I write fantasy stuff, too, though since the blog is relatively new, I haven't started adding that in yet. I'm currently wrapping a sequel to a finished fantasy work (again, publication approach TBD) before switching back to relatively hard sci-fi.
Initially also on Emacs as a minor mode, the concept is to write in a shorthand of greek letters, mapped like to a Dvorak distribution, that stand there as a code, as variables or latent values, expanded upon a key combination to one of several languages: German, Hebrew or Yiddish. Text can stay code-mixed indefinitely. The greek letters are a preform, condensed by layers of translation and transliteration, a hybrid con, and the keyboard layout itself, dynamic and not static, is dual: for every greek letter in the layout correpsonds a hebrew letter or diphtong or diacritic mark etc. It is a merge and disambiguation of both Dvorak/Programmer Dvorak and Yiddish Klal.
I really don't know if it may seem useful to someone, but I've been itching writing a bit jargon in a system like this for some weeks.
I'll come back to it once I'm farther along in my language learning journey and want to take my foot off the gas pedal a bit.
Also working on my homelab. Recently got DocuSeal, changedetection.io, and Actual Budget running.
https://x.com/seanw_org/status/1815442179361317022
There's lots of tools in this space, but the key features of this is it lets you easily check the contrast of any color pair (not just against white, so you can check e.g. your text colors will contrast on off-white shades), it's for creating a full palette of colors vs a handful of brand colors (you always end up needing lighter/darker variants for things like borders and backgrounds), and you can alter how the hue/saturation/lightness varies across a whole swatch of colors with a few clicks (being able to visualise these curves also makes picking new colors really easy).
Feel free to reach out if you think this might be useful to you!
All that to say I "worked" on a chrome extension to let you browse a hacker news post via arrow keys: https://github.com/humishum/hacker_news_keys.
(While typing this I found that the arrow key actions are still active while in the textbox, time for claude to provide me a fix!)
Rich text editor, run PHP and NodeJS on device, manage Git repos, and view your projects in a built-in browser that includes dev tools.
Made it with React, Vite, firebase via Typescript. It’s a really different world for sure.
Right now it's a collection of a few tools:
• AWS Resource Explorer - a lighter-weight version of the AWS console where everything is just a sortable/filterable/searchable table.
• Access Denied Debugger - paste an "AccessDenied" message and get back a stack-trace style UI showing all the resources involved, reason for the error (e.g., which policy is missing a permission), recent changes via CloudTrail, etc.
• AWS Organizations / SCP Viewer - generates a tree-diagram style UI showing all your AWS accounts, which policies apply to them, etc.
Still working on merging these into a cohesive application (mostly just been scratching my own itches so far). I'm trying to consider privacy/security carefully, so everything is client-side, using the AWS JavaScript SDK, and creds/data are only stored locally.
Not a sexy idea, of course. However, I couldn't find any solution that fit my needs for my SSG blog. Services like Cloudinary, and Imagekit are too developer-focused, and popular image sharing websites like Imgur don't allow backlinking.
[1]: https://i.magecdn.com/d2b508/574114_cleanshot_2024_08_25_at_...
[2]: https://i.magecdn.com/d2b508/574114_cleanshot_2024_08_25_at_...
To help Londoners find their perfect home by filtering for proximity to parks, river, PoIs, commute time, etc.
The website is not optimized for mobile and customers have to pay to access the functional filters/properties.
Stripe has a "developer friendly" and "easy to use" reputation. But, since 2019 the EU regulation has stepped in and made things more complex with 2nd factor authentication for payments (3DS/SCA). This made payment integrations way more complex. Integrating Stripe subscriptions now easily takes 6 man-months (I'm being very optimistic here).
Also, there are some basic scenarios that are hard to get right:
- Creating a paid subscription and ensure a customer always has a card on file (this one is almost impossible). - Upgrading a free ($0) plan to a paid one. - Upgrading paid plans, eg $10 to $100/month and ensuring immediate payment. - Guarantee customer Tax location, keep the flow simple.
I made a video analyzing Stripe integration of a successful SaaS company (+$100M valuation). I first paid $30, then upgraded to their highest $2000/month plan - for free!
https://www.youtube.com/watch?v=YuXp7V4nanU
The gem I'm working on is intended to be used with Ruby on Rails apps. It covers all the above mentioned hard cases, and irons out many more Stripe API kinks. And yes, it handles the basic-but-impossible "ensure customer always has card on file if they start paid subscription".
After buying the gem you can hand-off the task to the junior developer (it's that simple). They follow the integration tutorial: follow the steps for Stripe dashboard config, do the local env config, done.
The product (Ruby gem itself) is ready. I'm now working on the web app, tutorials etc. If anyone wants an early, guided access, please email me - contact is in the profile.
Keeping myself marginally sane by building a custom browsing experience for my “graphic design reconstructed from dreams” project: https://dreambrief.presteign.com/
Why not Nix? I just don’t like the Nix language. Plus, it seems like they ended up reinventing part of Lisp with their Flakes, which made me even more convinced in using GUIX. The GUIX folks didn't disappoint me, as they utilized the full power of Scheme to design a (internal) DSL that is clean and almost reads like JSON/YAML. I feel quite confident that it’s a better language for building devenvs for most people. The only downside is that it can get really verbose, but, hey, at least you have full control.
If anyone’s looking to share their story, I’d recommend IndieHackers, but for hardware-focused startups/projects, check out howtoware.co. (I’m not affiliated with the site’s owner in any way, but I really like the concept and want to help spread the word to my fellow hardwarers).
I've recently taken a new job working closely with some data scientists so wanted to skill up a little. So far I've got an elixir library which can calculate Attainment 8 and Progress 8. They're formulas provided by the DfE to understand individual student progress, and the schools ability to progress students.
It's just a bit of fun, and is letting me experiment with some of the various elixir data libraries and livebook. At the end of it, his trust is going to plug in some anonymous data and see if they can get more insights than this mega spreadsheet.
It's also nice to be doing a side project which isn't aimed at being a business.
I wanted a way to define new Arbitraries that's easier than working with .map(), .chain(), .oneOf(), and other combinators that require you to think in terms of sets. It has those combinators, but you can also write code to randomly generate one value at a time. It uses an approach to shrinking that's inspired by Hypothesis.
Along the way, I ended up adding a Domain subclass, which also does validation like Zod. Not sure where I'm going with that, but it's also useful for generating unique keys.
(The documentation isn't done and doesn't explain what's interesting about it. Caveat: it's Deno-only, and will probably stay that way unless someone wants to help.)
The one zone that's in the game is looking more polished than ever, but it's not going to be much of a game if there are only 4 levels. Next month is all about stage art, enemy design, and items to flesh out the new economy systems. Hopefully it all comes together, we'll see. It's a ton of pixel art, and I'm still learning my way around that craft.
One early achievement was buying voop.com - a very nice man in Arizona had held it since the 90s, and despite being retracted on whois, i worked out how to contact him via his custom ssl certificate on the empty holding site parked at the root domain. After a few emails and follow ups to get a response, I bought it for a not-cheap-but-not-bank-breaking amount.
FHIR is a data standard for sharing healthcare data. It's in use all around the world by healthcare, health insurance and medtech companies as well as researchers working with healthcare data.
Vanya grew out of my own frustrations as a developer working with FHIR. I wanted an easier way to see the data behind the APIs. To say Postman is not ideal for working with this data is a major understatement.
We're a few months away from a commercial release, and have an MVP available for download right now. We're very much building in public and have a strong user base already all around the world.
Just finished a Sunday morning release of the latest version. Looking forward to the next 6 months!
Currently focusing on building native apps using React Native. Finding limitations with the platform and considering dropping the whole thing and start building with Swift and Kotlin instead.
It also helps you discover new recipes that other users have saved.
Together Gift It solves the problem just the way you’d think: with AI.
Just kidding. It solves the problem by keeping everything in one place. No more group texts! You can have private or shared gift lists, and there are some AI features like gift idea collaboration and product search. But the AI stuff is still a work in progress.
I’m grateful for any constructive feedback.
I’m writing the backend in Go, with a standard API layer to allow custom frontends. I will also be building “official” frontends for mobile apps, desktop apps, and browser extensions.
The main reason I’m doing this is because I don’t like the UI of other self-hosted password managers, and I hate relying on the security of cloud options.
Seeing as I just started work on it today, I don’t have much (not even a real name). If you want to follow along, heres the Github, https://github.com/dickeyy/passwords
My goal for this project is to provide teams and individuals with the ability to secure their passwords while also providing a clean and elegant experience.
LMK what you think!
After years of having the New Year’s resolution, “make a personal site.” I finally got past my procrastination and did it! It’s live and I’m filling it out with things I’ve done over the years.
It is a layer on top of the qtile window manager that allows for more advanced window arrangements - especially subtabs. It was an itch to scratch and I use it myself.
I am now diddling about with Rust and thinking of trying my hand at a PDF parser library. It's sort of nice when there's an established specs document - like all the requirements are clearly laid out and you don't have to spend time thinking about the scope and product design.
I can implement as much as I want to, have tests, and come back another time when I feel like it.
I am almost done with the landing page and don't think I'll put much more into it, and will continue working on the API itself. But I still plan on adding a playable game into the 3D scene just for the sake of it and to learn. But so far I'm struggling at projecting a 3D camera render to a texture inside an already existing 3D context.
[0] https://the-project-e609e.web.app/ (temporary domain for development)
Pretty standard CRUD flows + plenty of opportunity to get comfy with iterating and deploying software on iOS.
This app will basically be the playground for me to work on: - Frontend / UX - Computer vision (scan in cards w/ Camera) - recommender systems (what cards are good suggestions to add to this deck?) - Clustering (what decks are similar to this deck?)
Websites like TappedOut and Archidekt are great products, but as someone who plays casually and doesn't do much to keep up with releases I find assembling fun Commander decks from singles to be quite a time investment, and I'd like to see if I can focus in on that UX specifically for myself.
Most of my days are spent reading the news and working on LLMs, which has been a blast. As an example, here's a dashboard that tracks major supply and demand shocks to various commodities around the world: https://emergingtrajectories.com/c/commodities
Besides building the ideas I have in mind around how UAT could be improved for both devs and clients (users). My other goal is to encourage more Malawians to consider Go as a programming language for shipping solutions in, my Go evangelism has been working slowly but I think people need to see something they can appreciate to get more people onboard.
Also working on a personal finance app in Flutter for young Africans, but that's not public yet :)
It allows querying multiple resource kinds via their relationships (i.e. replicaset owns pod, service exposes deployment…) and easily crafting custom response payloads.
Lately I’ve introduced aggregation functions and the ability to visualize query results using ascii art.
This is a daily driver for me and I’m now mostly focused around features that will make it a complete kubectl alternative.
It's meant to be a scripting language to be embedded in another program, generating HTML. Eventually I hope to make a CMS and this be my solution for easy templating.
Technical details: Written in C, the language has an x86_64 JIT compiler, but falls back to a bytecode VM so it should work on any architecture. The language itself is dynamically typed and garbage collected. Currently at 20K LOC (not counting blank lines or comments), with good test coverage and checking for memory leaks. It's been quite enjoyable!
Personally, I'm building a website which lets you quickly setup personal online dashboards, because I wanted to have an overview of all the sites I refresh often on one page: https://frankendash.com/
Basically it is an AI computer. It is nothing like we have seen before.
May open source it soon. With working zero hallucination (need additional evaluations).
Works with all devices. Created a MVP, air-gapped for military use, but it can be used for anyone including businesses/consumers who need an on-prem AI computer.
It has zero lag or zero second delay in responding your questions, and does not use any third party AI cloud platform. This is critical, because some of our customers have restrictive requirements for their data and they can't have anything of their customers to be placed on these cloud platforms. (eg. OpenAI, Google, Apple).
Adding additional features including Self-Teach, Cloned Personality, the holographic UI Wear, makes calls on behalf of the user, uses voice imprinting, edit videos, autonomously, for content creators and remembers 100 years of your life. And, more.
The UI Wear uses the exotic sensors I used in military to make it light and performant unlike Apple Vision Pro. Imagine interacting with something like "Jarvis" from Ironman. Quiet similiar but much more grounded in a sense that it's simpler and logical. Also, you can take it with you to play fast action sports or play games like Dragon-Ball-Z type game with your buddies. And it lasts a day, not two hours, on a full charge.
You can interact with the Personal AI Computer without the UI Wear, but it takes you to another level of experience.
Currently, doing it solo and looking for a cofounder who has experience in large distribution, logistics, creating polished commercial hardware, etc. Even if you don't these experiences, please contact me, if you want more info, interested, and/or want to work for this startup.
Some sample photo to help you on imagine. https://docs.google.com/forms/d/e/1FAIpQLSdwsdyvjkIcTk41JFXK...
There is still a lot to do and learn (especially in the marketing department), but we have plans for a new product in the privacy space. I don't want to say too much about it until we've started working on it, but it's in the compliance space and fits quite well with our existing product. I think it's always a good starting point to solve your own problems.
It's very alpha at this point with no styling. I'm also working on getting addresses, websites, and instagram links gathered and added. Plus collapsing down the entries, so a place with multiple awards only shows up once, with all of its awards. Right now it's just 1 results per award.
I'm currently road-tripping through New England and it helped find the amazing bakery Norimoto.
It’s a powered 55-seater auto gyro - an electric helicopter during take-off and landing and a conventional turboprop regional “plane” during flight. We are at a “works on paper, let’s start prototyping” phase: basic assumptions are verified by people from industry, a development plan is in place.
My co-founder is an experienced high-power electric propulsion engineer and he’s currently developing our power train. I am in charge of the business/operations/relationships side. We are currently looking for a third co-founder to lead aircraft design.
On the technical side, it's a desktop application coded entirely in java with a javaFX interface. No AI or online data uploads, just the Lucene search engine and PDFium for PDF rendering.
Here is a very short demo : https://youtu.be/CGo9JRUByGA
Very excited.
The primary feature that I'm implementing p2p synchronization (similar to syncthing, although the prototype use tailscale); append-only storage for easy backup and an "api" for easy integration/synchronization with external sources like RSS feed for example, all working offline.
Right now I'm using various software to manage my digital life (Zotero, Keepass, Syncthing) but i want to consolidate since I'm having trouble keeping all properly synced and backed)
https://github.com/willswire/checkd
This project has been an exciting way to explore better securing my iOS apps. I'm looking forward to refining it further and would love to hear any feedback or suggestions from the community!
There are many out there, who are intimidated by trivial things, such as sending a text, or even scrolling or switching between apps (what is even an app?).
So this course is ready made just for our parents. Or just for the ones who have no patience or capacity to teach these things step by step to their loved ones.
8+ hours and 100 lessons: https://smartphonehowto.com
I hope it helps many of you!
I started it a couple of months ago, I was tired of reading the same story posted across different news outlets.
Over time I added new features like summarization, the ability to hide articles that have specific words in titles, etc.
Currently working on adding custom themes and fonts for the reader mode and various customization options.
It's a fun little side project, still much to do to improve it, I'll be busy for a while!
Edit: Two from prior published here: https://github.com/jaronilan/stories
Currently working on adding manga and YouTube support, and figuring out how to expand it to all languages
https://github.com/jonroig/usBabyNames.js
We're engineers, we want to make data-driven decisions about what we name our children. My app won't necessarily help you choose a name, per se, but it can assist in eliminating a lot of possible names, giving you a much smaller set of choose from, each of which you can research more. So... like filter by name origin, length, popularity, etc..
Uses the scenedetect Python library to locate scene breaks and then uses perceptual hashing to find points for alignment. Also will include a small webapp to debug and adjust things manually in difficult cases.
Clace already supports most python based apps (wsgi or asgi), any other language works with a custom Dockerfile. Plan to add support for automatically shutting down idle containers, allowing for scaling down to zero for each app.
It's a simple site for streaming local radios from my city, Mendoza (Argentina): https://www.radiosdemendoza.com.ar
It was built with WordPress, and the new theme is using Tailwind and AlpineJS. I hadn’t used AlpineJS before, so this project was a great way to experiment with it.
I don't have any monetization on it, and I'm not sure if I'll pursue that for this site.
Next up, I plan to build a Chrome extension. I have limited experience with that, so it should be a fun challenge.
Been working on this for nearly 2 years now and it’s been cool seeing people use it daily, some definitely a lot more than I do. I still feel like it’s not even close to its full potential; infinite more things to add but most often a joy for me to work on.
For work I'm working on AI implementation. Because the vendor does all the technical goodness I'm kinda bored. It's also quite a crap solution IMO and we're kinda forced to be all positive about it because the vendor is closely watching and they have a lot of buyin from the top.
I set up a local LLM server too just to feel like I'm actually doing something.
Then, I started working on an interactive TUI app for Python beginner exercises. I've previously written such apps for CLI tools (grep, sed, awk, coreutils, etc): https://github.com/learnbyexample/TUI-apps
AnyStack Pages is aimed at making it easier and simpler to host static websites with free SSL protection and CDN. Websites can be hosted by uploading a zip file of the site contents or a single html file. It's currently in closed-beta for limited users every week.
Next Steps: Full launch with support for adding custom domains and premium features.
Happy to chat about this if you are interested in getting early access or for any questions and suggestions.
I'm a designer, and this is my first Vue project. Super happy and proud of it. This is a way for me to learn coding. I'm working on improving the project by adding a simple website builder where people can edit the components and templates directly.
It's an online development platform similar to other things like Replit, but it's mainly for static site hosting. You can create projects and make custom subdomains for them. You can add libraries to your project in a click of a button, so you don't have to add scripts or CSS files to every single file of yours. I've recently been doing a lot of work on it and I hope to launch it pretty soon.
If you know a mortgage loan officer, tell them to email tyler at lightningestimates dot com and mention HN. I'm just trying to fix an industry problem at a reasonable price.
I noticed that sometimes, I book accommodations in a new city without realizing I'm very far from the nearest vegan restaurant, having to walk for a really long time to get food, and sometimes getting there after the kitchen is already closed.
The planner helps me find hotels/airbnbs that are within 15 minute walk of at least 3 different vegan restaurants - makes my life so much easier!
The setting is like retrofuturistic version of 1998 with survivalist elements.
I’ve been enjoying flutter, and Have found it easier than other cross platform GUI technologies I’ve used. Dart is pretty easy to learn too and I’m considering it for other non flutter based tools as well
I've spent a lot of time in analysis paralysis and this has given me the kick in the pants to get me going again.
As far as new ideas go, I've already spent time learning Verilog, and hope to get a chip design through the TinyTapeout[3] before too long.
Also, I found out that it only takes a few lines of Python and a lot of time to have an AI rate all my photos locally. That's a work in progress, should be less than 100 lines all told with niceties.[4]
[1] https://mikewarot.github.io/Bitgrid_C/bitgrid_sim.html
A delightfully simple blogging-by-email app, Pagecord (https://pagecord.com). $20/yr.
A delightfully simple, all-in-one RSS reader (web and PWA), Feedgrab (https://feedgrab.net). FREE!
Very early days (built quickly!) but both products are fully functional with customers.
I have a lot of ideas around content discovery for Feedgrab.
Give them a spin, I’d love to know what people think.
I like to mix different materials into the designs I cook up. For the tops I use cut stone which is prohibitively expensive for most hobby stuff but works really well for commission work because of all the different types (marble, granite, composite) and styles/ colors you can get. It gives it a very high end look and feel when combined with a sturdy metal frame. I’m also really into 3D printing so I try to incorporate that into the work as well. Using different pattern schemes to make inserts and accent pieces.
I’ve also experimented with incorporating IKEA furniture pieces into metal frames because I love the idea of doing modular furniture that can be swapped out for different colors and textures. I use the same idea for my 3D printed work.
I love the creativity and flexibility this gives me and my welding skills have ramped up significantly. It’s been years since I welded seriously and I really lost the techniques so getting that confidence back is great. I’ve been approached to do some big 5-6 figure home design jobs through the connections I’ve made acquiring stone pieces but I’ve only consulted on advice. I still feel a bit insecure about putting myself out there in structural work. I see how successful different design firms locally are building this stuff for tech companies and I’d love to eventually have my work featured like that but taking it slow for now.
As a founder, I really needed a simple place to centralize all those business metrics. Couldn’t find anything that suited my needs (everything was way overkill) and so I ended up building Polynomial.
Still tweaking it, and it could use some better instructions. But the thread's here now so why not.
If you see confetti, you answered correctly. Otherwise the points are added or subtracted in a "hot or cold" fashion based on your answer's distance from the actual pitch.
My motivation has been some recent jobs/roles in which no-pipeline-feature-gets-denied YAML scripts have grown unmanageably large.
I'd be interested to hear of any pitfalls that any of you might foresee. Other than this is a pretty hefty undertaking - that much I can already see on my own. :-)
My race yesterday: https://app.see-sailing.com/b/nmea/2024-8-24-11-34-21
The open tooling is quite lacking in the sense that one has to create a lot of stuff manually. Especially for the evaluations, even though most of them are quite common in the actuarial realm.
Also I wanted to tinker with Jax. So I am basing all on Jax, polars, seaborn/matplotlib and scipy.
I already have lift plots and Poisson regression in Jax. I’ll keep adding to it over the next months as time allows.
Note: this is a WIP, so if you have any experience working with .fcpxml, I'd love to hear from you!
It's a way to share messages, files, code snippets, links, without worrying about distracting others, or quitting the app yourself to avoid notifications. It uses around 10x less RAM than Slack/Discord/Teams and prioritizes fast moving teams over enterprises. High quality software over features. Features that work rather than half-backed list of smart/AI/notes/wiki/etc bloat.
The most recent site is https://computerdisplayprices.com
When I’m done, I’m probably going to make a blog post or two just talking through the automation I figured out so far. LLMs are amazing force multipliers.
To learn more about SEO, digital marketing, and other areas I hadn't explored in much depth yet, I tried bringing it to market with modest success. It's been a fun journey so far, and I’ve documented the first legs here: https://medium.com/@desmat.ca/a-genai-haiku-journey-part-1-a... and https://medium.com/@desmat.ca/a-genai-haiku-journey-part-2-f....
I'm also experimenting with a similar format for limericks (https://limericks.ai) but I haven’t cracked that nut yet.
But in order to get there, I need to rework a couple of mirror grinding machines in the club according to specs by the professional optician teaching us.
This means building the right set of mechanical thingies to safely hold glass and tools from 6 to 16-inch diameter optics. According to our optician, we need an aluminum disc the same size as the mirror with a reverse threaded hole in the center of the aluminum disc. This disc will be screwed onto a coupling which is attached to the shaft of the turntable motor.
We are currently working on sourcing the parts, machining, and doing a trial run with a 6-inch mirror. Then I get some blanks, get the curve I need generated to save me hours upon hours with 80 grit hogging out the glass before he can settle down and let the combination of grit, water, moving glass and time work their magic on the surface, every minute bringing me closer to optical eh-its-good-enough".
We have enough members in the club building telescopes that the effort of making these mirror grinding and polishing machines work is well worth it. With any luck, you'll see me at Stellafane next year.
But lately I took a break and worked on running (GPX run file analysis) and habit tracking. The eventual goal is to package it all together into one integrated PWA / mobile app.
Transparency and non-executable export format is a serious thing, take a look at https://embracethered.com/blog/posts/2022/machine-learning-a....
After version 0.9 release, PyMilo became feature-completed with full support of scikit-learn models, now it's time to move on to PyTorch and then Tensorflow. But we decided to add the "ML Streaming" feature before getting into PyTorch, in order to provide an easy way to smoothly stream your ML model. By using the "ML Streaming" feature you can easily deploy your model into the remote server, connect to it from the client side, and choose the working mode, either delegation or local mode, through delegation mode your requests will be relayed to the remote server and you can easily work with your remote model from any devices without any further dependencies, and finally, you can download model for local use.
We will release the 1.0 (tenth) version of PyMilo around Sep 16th, this release will be the first release to have the "ML Streaming" feature with support of REST API, and we will next add other protocols such as Websocket.
Here is PyMilo: https://github.com/openscilab/pymilo Thank you for your time
I'm doing so by making it easier to mint certificates for your pods in k8s (https://gitlab.com/gauntletwizard_net/kubetls/-/tree/master), by writing documentation on how to create good root certs with cheap HSM backed keys, updating cfssl to work with name constraints (https://github.com/GauntletWizard/cfssl/tree/ted/constraints), and building tools to issue short-lived certificates to developers.
Outside of that, I've been blogging a lot more (https://www.jacky.wtf/essays/ - August looks so full, ha) and now I'm writing about things I'm reading too (https://www.jacky.wtf/links/). Been doing this to try to ween off social media and rely on places like this to share stuff.
It comes with a set of tool for your IDE (currently only VSCode) where you can draw sprites, compose music, make SFX right next to your code.
It's still not in v1, it's also my first "serious" C++ project, and life got in the way lately -- I haven't made significant progress in the last month or so
You can even buy the app itself on it. It’s going to be my starter for all other projects too, so I’m adding all the common CRUD app features. Also it aims to be lowest cost possible, which is free for now. I try not to tie it to specific technologies when possible to make it as easy as possible to switch providers in the future.
I used to work for a media monitoring company but was instantly struck by how old fashion everything was. And the total reliance on boolean searches meant only experts could find relevant information. This still appears to be the case for most players in the industry.
So I'm building a platform that finds what is important before you look for it. Novel entity linking, and sentiment analysis plus speaker tracking. It has come a long way from the proof of concept. Focused on audio media at the moment as it is the hardest to index compared to news articles in my opinion. And the hypothesis is audio media such as podcasts can contain so many juicy insights.
Next steps are converting pilot customers to paying customers, testing more markets (based in a tiny market now), raise a small pre-seed (bootstrapped at the moment), and quickly evolve the product based on feedback.
Since I'm just one guy with 100 ideas, I have to try to apply my skills where they can have the most impact. My current project is converting existing circuits from Fritzing format (which is usually displayed graphically) into a textual format. I wrote about it in these three blog posts:
1. Premise, ways to represent circuits: https://blindmakers.net/posts/accessible-schematics-1/
2. Why Fritzing? https://blindmakers.net/posts/accessible-schematics-2/
3. Actual output examples: https://blindmakers.net/posts/accessible-schematics-3/
I'm also working on a 3D printed single character Braille cell that can be added to any Arduino project, and has only 3 simple parts to assemble. Unfortunately I haven't posted about that yet.
We were both tired of asking AI for the same code again and again, so we thought, why not build something where we can instantly save code from our AI chats
The idea is to create a public library of AI-generated code that anyone can search through, find what they need, and either use it directly or customize it further.
For example, if you find a function that almost fits your needs, you can add it to your AI chat with just a click and tweak it, instead of starting from scratch.
So far we have -
1. Chrome Extension - To save your code or find what you need from the library and add it to chat. Currently only works with claude, working on making it compatible with v0 and chatgpt
2. VS code extension to find whatever you need right in there and an AI that will understand your project's context and integrate these snippets in seconds ( AI part is in progress )
3. Website to do other stuff - build your profile, rate other snippets ( which helps us to maintain quality )
Tech Stack -
- Frontend: NextJS, Tailwind, Shadcn
- Backend: Node.js, Express, MongoDB
- Extensions: TypeScript (VS Code) & React, TypeScript, Tailwind (Chrome)
It has been a wild ride. Often frustrating but rewarding. Some days I may spend 10 hours solving an API for which app is frontmost at a given second. Sure, there's a system API for that, but it doesn't even remotely cover edge cases that exist. And many times with no answers out there on how to do something, it feels like I'm the first, so finally solving it is a great endorphin rush.
Part of the complexity is I wanted the app to be completely pluggable, kind of like Obsidian. The app communicates events to a local socket server in full duplex, which enables cross-plugin communication (but also cross-app!). Plugins access the app's JavaScript system bridge API for pubsub and system calls through a Webkit interface. Edits are hot reloaded and instant, no compile time necessary. The first time I could change my "Start" menu in real time through JavaScript/CSS was quite a feeling.
Sadly, I couldn't leverage existing tools like Tauri or Electron. They don't have adequate system bridge APIs available, and would be more work to leverage instead of less. They are too general, whereas this project only builds to macos. Therefore it can be much more complicated (and useful) by design.
I originally set out to be more productive in macos. But I've also spent time making prototypes for fun. A desktop widget system, real-time system color theme set from Spotify album art, live video desktop backgrounds from Twitch/YT, a Destiny 2 macos system theme, etc.
I plan to open source it and build a community around it one day.
Some key features we're working on: - AI-powered analysis of in-match player actions to detect anomalies - Customizable rules engine for automated responses to toxic behavior - Visual replay system for reviewing flagged matches
Check out our website at https://www.getgud.io and watch our detection video at https://www.youtube.com/watch?v=5EhTpfEzh1M to see Getgud.io in action.
We support server-side integration for popular platforms like Unreal Engine and Unity. For integration guides and SDK references, visit our docs at https://github.com/getgud-io/getgud-docs.
Happy to chat more about game analytics and cheat detection if anyone's interested!
I'm knee deep into coding 6 days per week as I'm building a startup and working right now on an "easy" solution for merging xls (yes xls) from some external company with our internal json, and I needed to let my keyboard rest for a day.
It's convincing me that our company should allocate a yearly budget to donate to OSS for stuff like the xlsx package.
It doesn't save anything to the cloud. It supports markdown. It currently doesn't using an LLM but I'm sure that will sneak in at some point. One day, I'd like to have it generate wireframes using AI.
the goal is to be a lightweight alternative to pgadmin/dbeaver. it has vim-like keybindings, shortcuts to preview tables, and session history.
it only supports postgres at the moment, but sqlite and mysql support are on the roadmap.
https://github.com/thebigG/Gunner
It's got long ways to go before being "complete", but I'm enjoying the heck out of working on this. I like working on things that aren't tied to money/serious job because they remind myself of the joy in programming :)
So now I’m working on learning Golang and loving it! Going to build a backend with it
[0] https://chromewebstore.google.com/detail/ui-copilot/hgaldpfd...
You can talk to it over a e2e encrypted reverse proxy (using Signals double ratchet protocol) so you can access it on the go safely and privately.
I've gotten quite carried away with a web interface and custom on-disk (on-sd card?) storage format based on Facebook's Gorilla paper.
I've realised that where it'll be deployed it will only have access to a public guest WiFi, so a remote server hosting the dashboard makes sense. So now it sends the compressed time-series over a TCP connection to a server component, hoping to bypass WiFi captive portals. Might need to use UDP to make it look like DNS or perhaps NTP. I haven't tested it on-site yet.
It has been a very fun and rewarding project so far. Looking forward to deploying it and getting remote updates working if I can get it to work with Tailscale on the guest WiFi. If anyone has any tips on circumventing captive portals and sending really small amounts of data through, I'd love to hear it!
I built a bunch of consultant led tools (Jupyter notebooks mostly) using LLMs to help with this, and saw how customers reacted to what they could do. So I decided a few months ago to bundle them together into a SaaS app.
I’m just about to release the beta of https://www.portage.so to my waitlist in the next couple of weeks.
The main use case is using a virtual whiteboard (ReactFlow) to string together a series of nodes that do a different task through the strategy development process. For example, you can generate scenarios, then analyse the impacts, then create courses of action in response to potential disruptions and so on.
I do a fortnightly dev diary as well, the latest one shows off creating the final output: https://youtu.be/nD_FhWkREhE?si=JK60fVlT7wmmONW3
Still got a big list of improvements to make but I'm quite happy with how it's coming together.
Ideally a way to let people lay out strategies and see them executed by a team, go on - prove you're the best IGL!
It's slow going between a full time job and family life but really great to expand my mind into something that isn't "web dev".
Working on the open source alternative to Heroku - https://ptah.sh (https://github.com/ptah-sh/ptah-server)
Already have some success: we use it for our other 3 projects, thus saving us some cash. :)
I have the 1-click apps marketplace, backups and, soon, monitoring.