Ask HN: What are you working on? (April 2025)
timtimmy 2 minutes ago A v2.0 update for my biology education app. I'm adding the ability to walk around cell models with billions of atoms on the Vision Pro.I'm designing the content browser right now. I'm trying to achieve something really immersive like Apple's new Spatial Gallery app.
tetris11 17 hours ago A tree cutting tool.Take photos of the tree from 6 different angles, feed into a 3D model generator, erode the model and generate a 3D graph representation of the tree.
The tool suggests which cuts to make and where, given a restricted fall path (e.g. constrained by a neighbors yard on one side).
I create the fallen branches in their final state along the fall plane, and create individual correction vectors mapping them back to their original state, but in an order that does not intersect other branch vectors.
The idea came to me as a particularly difficult tree needed to come down in my friends yard, and we spent hours planning it out. I've already gotten some interest from the tree-surgeon community, I just need to appify it.
Second rendition will treat the problem more as a physics one than a graph one, with some energy-minimisation methods for solving.
vintagedave 16 hours ago This is the kind of thing that makes me love HN. An idea I would never have thought of, with an immediately obvious use in multiple ways (fall path plus ideal lumber cutting?), probably very difficult, yet being tackled with one implementation already... and spoken of quite humbly.cacheorbit 1 hours ago Testing this is real pain in the ass, you gotta cut real tree to see if it works in various situations :(Xmd5a 2 minutes ago I got filtered by the Ent arc of LOTR and dropped the book.
chaosharmonic 10 hours ago Funny, one of mine also involves trees -- but is mostly outdoor cleanup. The kind that involves decades' worth of it, thanks to what I'll just say is a lot of maintenance that wasn't done over a long time. There's an extensive amount of brush, leaves, etc of varying ages that could maybe be shredded up into something useful, invasive vines I'm still trying to deal with, and more old trash than I've fully figured out how to properly dispose of.It's turning into various DIY rabbit holes, actually, with the next one (outside of various related landscaping stuff) being to gut a basement.
firesteelrain 5 hours ago You should branch out (hehe) into flower and plant pruning suggestions with your app. ChatGPT can do this now if prompted.defterGoose 10 hours ago I would love to have such a model tell me how to prune my fruit trees as they grow up. Should be a fairly straightforward supervised problem with the right front end for the graph generation.pbhjpbhj 7 hours ago When i read OP this is what I thought it was going to be - these branches are going to be apex competitors, these are crossing or going to cross, this one shows signs of disease, this one interrupts air flow through the centre, etc.toss1 4 hours ago You can start right now with an algorithm I learned from an expert when I was working in a landscaping business.It is a very simple three-pass plan: "Deadwood, Crossovers, Aesthetics".
So, first pass, go through the tree cutting out only and all the dead branches. Cut back to live stock, and as always make good clean angle cuts at a proper angle (many horticulture books will provide far better instructions on this).
Second pass, look only for branches that cross-over other branches and especially those that show rubbing or friction marks against other branches. Cut the ones that are either least healthy or grow in the craziest direction (i.e., crazy away from the normal more-or-less not radially away from the trunk).
Then, and only after the other two passes are complete, start pruning for the desired look and/or size & shape for planned growth or bearing fruit.
This method is simple and saves a LOT of ruined trees from trying to first cut to size and appearance, then by the time the deadwood and crossovers are taken later, it is a scraggly mess that takes years to grow back. And it even works well for novices, as long as they pay attention.
I'd suspect entering the state and direction of every branch to an app would take longer than just pruning with the above method, although for trees that haven't fully leafed out, perhaps a 360° angle set of drone pics could make an adequate 3D model to use for planning?
In any case, good luck with your fruit trees — may they grow healthy and provide you with great bounty for many years!
monkeywithdarts 11 hours ago I was imagining something like this for pruning fruit trees — something to help noobs like me see how to put pruning guidelines into practice on a real, overgrown tree. Good luck!javiercornejo 14 hours ago Where I live this could be very helpful becuase people is too, how to say it, maybe ignorant in safety and logic specs. Also could be usefull to know or estimate what tree are in a innminent or highr posibilities of fall with wind.Happy to help!
r0fl 16 hours ago That’s a great idea, but so much liability if the user is an amateur and follows steps incorrectlyJackFr 13 hours ago Or perhaps follows the steps correctly.conductr 8 hours ago Making this determination alone will sink you in legal feesDoes an insane amount of fine print really save you? Even if you say the model is only an aide to be used by licensed or certified professional arborists or whatever, I fear some Joe blow whose tree lands on his house will be suing you.
postscapes1 8 hours ago This is great idea - I have a huge tree in front yard that will either cost be $5-10k to come down or was going to rent lift and do it myself - A few particular branches scare me though in terms of how they will come down... Bonus points for where to tie things off.1970-01-01 13 hours ago Do consider the value of the wood in relation to your cuts. A well-placed cut not only guarantees safety but will also take the maximum board feet from the tree.boogieknite 11 hours ago i work a lot with NVEL for this. one time even tried porting nvel to wasm for fun and client accessibility. we "virtually buck" trees which seems like could be applied to your proposed use case. if op wants to go down this path: https://github.com/FMSC-Measurements/VolumeLibrary/tree/77d4...dyauspitr 12 hours ago Seems insignificant. What are you optimizing for- an extra foot or two?1970-01-01 12 hours ago Yes, board feet is usually measured by the inch.skeeter2020 8 hours ago with dimension lumber it's way more about the width you can cut than length; sometimes shorter is more valuable depending on supply & demand (and transport). Accounting for the fact that trees are not perfect cylinders (or cones, really) is where all the fun optimization comes from anyways.boogieknite 6 hours ago good ol conical frustum
mon_ 12 hours ago Aren't longer boards worth more per boardfoot too?jermaustin1 10 hours ago Yes, and the wider the more it costs per bf as well.I have a couple products I make that require 12" widths, which means I pay a whole lot more per bf than < 10" widths at my hardwood supplier.
trollbridge 11 hours ago Yes, but most trees are plenty tall enough.
dyauspitr 11 hours ago Then just make the cut as low to the ground as possible. You don’t need a lot of complex math for that.boogieknite 9 hours ago the right cuts at the right heights while working down the tree from a specific max height of the tree to still produce viable board feet while maximizing boards per cut. in most places, unless youre pulping the entire tree, its quite a bit more complicated than cut as low as possible.its surprising to me how little work is done to make the tools which do this accessible considering how much money and open data there is.
it gets less open and more complicated is when you consider certain mills only can make certain cuts, produce certain products, and accept certain logs. then factor in distance between mills and the products they can make, and also log lengths accepted by the trucks which can travel those routes.
its all solvable and should be, but its so niche and that i still think there isnt an accessible solution
oslem 10 hours ago No offense, but this comment is very reductionist. The problem isn’t nearly as simple as you’re making it out to be.
pm2222 7 hours ago Perhaps an opportunity for weed control for lawn as well.boogieknite 9 hours ago i work in forestry software and am curious about your methods. is any of this open source? any intention on supporting growth modeling?tetris11 9 hours ago I plan for a time-bomb license (closed source for 10 years, make my money (if any), GPLv3 after that).My methods are all over the place. Tree is taken as-is on the day, and cuts calculated on the fly, no future growth-modelling if that is what you're asking
mon_ 12 hours ago How does the graph representation help you solve the problem?tetris11 10 hours ago I was mixing methods, sorry. My initial rendition for solving the cuts would initialise a somewhat sparse network from tree to ground, and solve for non-overlapping paths.This became convoluted and I just opted for a far easier method of solving vector intersections.
Its also not perfect since I haven't factored in rotation origin very well, and I'm now pursuing a far simpler physics-based approach
ddahlen 11 hours ago I am about to begin a PhD in astronomy. Until last month I was working at Caltech for 3 years on code which calculates orbits of asteroids to high precision. This code is being used on several NASA telescopes now to predict when they will image known asteroids (NEO Surveyor, SphereX, maybe Roman eventually). I was allowed to open source it and I am planning on making it the basis of my PhD research:https://github.com/dahlend/kete
It can predict the location of the entire catalog of known asteroids to generally within the uncertainty of our knowledge of the orbits. Its core is written in rust, with a python frontend.
jxjnskkzxxhx 10 hours ago Ever thought of making a presentation about this subject and putting it on YouTube? :-)It sounds really impressive.
ddahlen 10 hours ago I've never really dabbled in youtube. I have several projects/papers I am working on using this code, I have thought about writing some blog posts as I publish those. But a PhD is going to be a major time sink, we will see what happens.dang 8 hours ago Do you want to post it as a Show HN soon, before the PhD sucks you in altogether?(If so, email hn@ycombinator.com and we'll put it in the second-chance pool (https://news.ycombinator.com/item?id=26998308), so it will get a random placement on HN's front page.)
ddahlen 7 hours ago Thank you for the offer! Unfortunately the PhD has already sunk its claws in. I should have some flashy stuff to show off in a 2-3 months (I have a conference talk coming I have to prepare material for).dang 6 hours ago Drat! Well, if you ever have some cycles to share your work on HN, contact us at hn@ycombinator.com and we'll be happy to help.More importantly, good luck with the PhD and we all hope it goes swimmingly!
Intralexical 8 hours ago I like the daredevil asteroids going for the close dive of the star emoji sun :)Would it be appropriate to communicate on the README which telescopes this is used for? You see these very niche, very professional-looking repositories on GitHub now and then, and it's never clear how much credibility they have and whether they come from a hobbyist, student, experiment, or are in operational use.
juxtaposicion 13 hours ago I’m working on Popgot (https://popgot.com), a tool that tracks unit prices (cost per ounce, sheet, pound) across Costco, Walmart, Target, and Amazon. It normalizes confusing listings (“family size”, “mega pack”, etc.) to surface the actual cheapest option for daily essentials.On top of that, it uses a lightweight AI model to read product descriptions and filter based on things like ingredients (e.g., flagging peanut butter with BPA by checking every photograph of the plastic or avoiding palm oil by reading the nutrition facts) or brand lists (e.g., only showing WSAVA-compliant dog foods). Still reviewing results manually to catch bad extractions.
Started this to replace a spreadsheet I was keeping for bulk purchases. Slowly adding more automation like alerting on price drops or restocking when under a threshold.
abdullahkhalids 3 hours ago I don't think I have the time to go to different stores to buy different things based on what is cheap. I have one fixed one.However, what I would like is a product where I upload my shopping receipt for a few weeks/months from the one store I go to. The application figures out what I typically buy and then compares the 4-5 big stores and tells me which one I should go to for least price.
juxtaposicion 2 hours ago Yeah, I agree. It is a pain to search product by product instead of sticking to one store. Also popgot.com can only do what's online & shipped to you -- so really just the non-perishables / daily essentials that are not fresh groceries. But even when limited to consumables I save ~$100/mo by basically buying by unit price.Uploading a receipt to see how much you can save... that's a good idea. I think I can find your email via your personal site. Can I email you when we have a prototype ready?
abdullahkhalids 2 hours ago A one time email is fine.However, I am in Canada. So can only test it once you expand there. Thanks.
I don't know how things are in the US, but it does seem like the grocery store oligopoly is squeezing consumers a lot, so tools like this are valuable for injecting competition into the system.
KerryJones 13 hours ago I like this idea a lot -- feels like there's a lot of room to grow here. Do you have any sort of historical price tracking/alerting?And/or also curious if there is a way to enter in a list of items I want and for it to calculate which store - in aggregate - is the cheapest.
For instance, people often tell me Costco is much cheaper than alternatives, and for me to compare I have to compile my shopping cart in multiple stores to compare.
mynameisash 7 hours ago > For instance, people often tell me Costco is much cheaper than alternatives, and for me to compare I have to compile my shopping cart in multiple stores to compare.A few years ago, I was very diligently tracking _all_ my family's grocery purchases. I kept every receipt, entered it into a spreadsheet, added categories (eg, dairy, meat), and calculated a normalized cost per unit (eg, $/gallon for milk, $/dozen eggs).
I learned a lot from that, and I think I saved our family a decent amount of money, but man it was a lot of work.
juxtaposicion 7 hours ago Glad you guys mentioned Costco -- I happen to have written a blog post on exactly that: https://popgot.com/blog/retailer-comparison Surprisingly, Costco does not win most of the time, and especially if you are not brand loyal. Costco has famously low-margins, but it turns out that when you sort by price-per-unit they're ok, but not great.@mynameisash I'm curious what you learned... maybe I can help more people learn that using Popgot data.
mynameisash 6 hours ago One thing to call out is that costco.com and in-person have different offerings (& prices) -- but you probably know that already.I just dusted off my spreadsheet, and it's not as complete as I'd like it to be. I didn't normalize everything but did have many of the staples like milk and eggs normalized; some products had multiple units (eg, "bananas - each" vs "bananas - pound"); and a lot of my comparisons were done based on the store (eg, I was often comparing "Potatoes - 20#" at Costco but "Potatoes - 5#" at Target over time).
Anyway, Costco didn't always win, but in my experience, they frequently did -- $5 peanut butter @ Costco vs $7.74 @ Target based on whatever size and brand I got, which is interesting because Costco doesn't have "generic" PB, whereas Target has much cheaper Market Pantry, and I tried to opt for that.
juxtaposicion 7 hours ago I'm so glad you like it!We have historical price tracking in the database, but haven't exposed it as a product yet. What do you have in mind / what would you use it for?
mynameisash 6 hours ago I like that you have the ability to exclude on some dimension (eg, I don't use Amazon.com). Do you have or are you considering adding more retailers beyond the four you mentioned? For example, I buy a lot of unroasted coffee from sweetmarias.com, and excluding Amazon from Popgot results eliminates all but one listing (from Walmart).juxtaposicion 4 hours ago Ah, hell yeah! My buddy on this project has been itching to add sweetmarias.com ... he just needed this as an excuse.So yeah, we'll add it. If you shoot me an email (or post it here?) to chris @ <our site>.com I'll send you a link when it's done. Should take a day or two.
unvalley 12 hours ago Cool! I hope it's coming to Japan (I live) near future.
xarici_ishler 1 days ago The first ever SQL debugger – runs & visualizes your query step-by-step, every clause, condition, expression, incl. GROUP BY, aggregates / windows, DISTINCT (ON), subqueries (even correlated ones!), CTEs, you name it.You can search for full or partial rows and see the whole query lineage – which intermediate rows from which CTEs/subqueries contributed to the result you're searching for.
Entirely offline & no usage of AI. Free in-browser version (using PGLite WASM), paid desktop version.
No website yet, here's a 5 minute showcase (skip to middle): https://www.loom.com/share/c03b57fa61fc4c509b1e2134e53b70dd
benjaminsky2 13 hours ago This is awesome! I’m work with a team of analysts and data engineers who own a pretty big snowflake data warehouse. We write a ton of dbt models and have a range of sql skill levels on the team. This would be the perfect way to allow more junior devs to build their skills quickly and support more complex models.I would recommend you target data warehouses like snowflake and bigquery where the query complexity and thus value prop for a tool like this is potentially much higher.
xarici_ishler 12 hours ago Thank you, nice to get some idea validation from folks in the industry. For sure data warehouses are the top priority on my TODO list, I picked PG first because that's what I'm familiar with.I can ping you via email when the debugger is ready, if you're interested. My email is in my profile
parrit 22 hours ago Was thinking today... not a debugger but even a SQL progess bar, so I know that my add column will take say 7 hours in advance.thenaturalist 6 hours ago Possibly look at https://duckdb.org/community_extensions/extensions/parser_to...Even if not for DuckDB, you can use this to validate/ parse queries possibly.
jeffhuys 11 hours ago This would be incredible to understand why some queries execute slow; most of the time it's one of the steps in between that takes 99% of the execution time at our company. Do you record the time each step takes?thebytefairy 2 hours ago Can you not use EXPLAIN ANALYZE to identify steps that had the highest compute time? I think most databases have some form of this.xarici_ishler 9 hours ago You're onto the original idea I started out with! Unfortunately it's very difficult to correlate input SQL to an output query plan – but possible. It's definitely in future plans
lie07 1 hours ago Following...Ni3l55 13 hours ago Cool! We're dealing with many complex CTEs and costly queries. Would be useful to have those visualized one by one.xarici_ishler 10 hours ago What database are you using? I'd be happy to hear about your usecases and hopefully help you, shoot me an email (in profile)
netcraft 4 hours ago this is very cool! Where can I follow you to see updates?IceDane 21 hours ago This seems like it could be extremely useful.xarici_ishler 20 hours ago Thanks! Would you mind sharing what would be your use cases?At my job, all of our business logic (4 KLOC of network topology algorithms) is written in a niche query language, which we have been migrating to PostgreSQL. When an inconsistency/error is found, tracking it can take days, manually commenting out parts of query and looking at the results.
alok-g 15 hours ago Am not the person you asked, but feel that it could have good value for education and learning as well, besides debugging.
anitil 1 days ago Is this postgres only? What an interesting idea!xarici_ishler 20 hours ago For now, yes, but I'll start working on adding support for all other DBs (especially OLAP) as soon as possible. The geberal approach is the same, I just have to handle all the edge cases of the SQL dialects
binary132 14 hours ago Nice one!
9dev 14 hours ago https://github.com/colibri-hq/colibri/Colibri—a self-hostable web application to manage your (and your family's) ebook library, intended as a companion to Calibre. I want it to be a friendly, simple, capable, opinionated app to review your books, add metadata to them, get them onto your reader, share them with family and (few) friends, create a public shelf for bragging, connect with Goodreads etc., and exchange comments and reviews on books.
This is explicitly not intended to ever be monetised, and I enjoy all the implications that has on the design. Colibri is as much a tool I personally want to use, as it is a study in small-audience user interfaces, and the quest to build the perfect book catalog schema.
I'm looking for fellow book-loving people to work on Colibri, to create the best personal digital library possible. If you're interested, feel free to reach out via email (in bio), or on GitHub.
teleforce 59 minutes ago Great stuff, has been meaning to create an online library of my burgeoning ebooks collections for quite som time now, this is exactly what the doctors ordered and it's based on Postgres.Just wondering about the encrypted collection of ebooks from Kindle for example. Are these ebooks supported and does it only supports metadata, what about the content search for these ebooks?
vallode 13 hours ago A fan chiming in. I'm really happy someone someone is tackling this and it's looking good. One thing: can we get a demo instance just for initial snooping? A screenshot or two is fine but to get a feel for features it would be nice to have something (even heavily limited) we can just interact with?9dev 13 hours ago That's the first thing I'm going to do as soon as it's possible! I recently refactored the code base to a monorepo, and still need to make some adjustments so it'll run stable again. Stay tuned :)
geekplux 1 hours ago Looks really cool, gave a star!nathan_douglas 11 hours ago Ooh, this looks fantastic. I'd love to help, but I'm spending almost all of my off-work hours looking for jobs right now. Maybe I'll find a good one sooner rather than later...9dev 7 hours ago Thanks you, that means a lot. And good luck for your search!
ian-g 10 hours ago This is super interesting. Where do you store the ebooks and the metadata?9dev 7 hours ago Ebooks in an S3 compatible storage bucket, metadata in a Postgres database. That has the huge advantage of being able to do full text search and kNN similarity right in the database, for example.Colibri is built around a pretty solid data schema (I hope). Check out the migrations folder if you’re curious :-)
coyotespike 2 hours ago What a lot of amazing projects!I'm working on a defense drone.
I built a garage workshop with a Shapeoko 5 Pro, X1C, soldering station, and learned CAD (ok just fusion). I have a lil drone in the air and I'm adding OpenHD for vtx, Rpi for on-edge compute (Jetson would be better but is expensive).
Haven't figured out FHSS or GNSS-denied nav yet (tbh I feel like fhss is gonna be harder). And SITL in a good sim remains to be conquered (ros on osx is a terrible experience). I'm also designing a battery pack that's modular, quick-swap, smart/telemetry.
I've shifted a lot of focus to networking (attending SOFweek in tampa) for the normal fundraising/team-building/customer discovery.
I'm also basically broke due to bootstrapping so I'm about to partner on some b2b ai saas consulting with a friend, today I got Suna up and running, pretty cool.
lnsru 21 hours ago So many cool technical projects here. But I am doing something completely different - masonry. Repairing walls in 3 rooms. It includes reinstalling dozens of falling off bricks, installing 30 or so power outlets, replacing old windows with bigger modern ones, fixing openings for the doors and plastering everything afterwards. On one hand it’s interesting, because it’s very different from the dayjob. But doing it by myself pays my newish car in cash immediately. However I wouldn’t do it for money somewhere else, it’s really really hard work.Instead of masonry I would like to work on time of flight cameras. But the day has only 24 hours :-(
redbackthomson 12 hours ago Working on a browser extension to make it easier to find content on YouTube that fits your interests.I watch a lot of YouTube videos and have found it very annoying that YouTube latches onto one or two topics that you've watched and only recommends that type of content over and over again. Even if you use their "Not Interested" tool, not a whole lot changes in your recommendations.
At the end of last year I launched Relevant - a crowdsourcing website where users can categorize the channels they watch into a defined hierarchy of categories ranging from broad topics like "Science" and "Gaming" to more specific ones like "Phone Reviews" or "Speedrunning".
Although I've had good feedback on the website, engagement has been relatively low and I think that's because it's a big ask to have someone navigate to the website to find the content. This year I decided that I'd bring the content to them by making a Chrome extension that lets users interact with Relevant directly from within YouTube.
It's still a work in progress but I'd love to get a first version out within a month or so to start spreading the idea and gathering feedback. If this is of any interest to the people here on HN then please let me know what you'd like to see most on your feeds.
Timwi 3 hours ago Hm, I tried to contribute, but it asked me to categorize a channel I've never heard of (DefendTheHouse) and no matter how many times I click “skip” it keeps going back to the same one.I also notice that you first said “browser extension” but later you said “Chrome extension”. Are Firefox users going to be out of luck?
HaZeust 8 hours ago I strongly, strongly recommend adding video essays as a category as well - they're very big now.redbackthomson 7 hours ago Yeah absolutely! The category list is meant to be dynamic as the industry changes and new forms of contents crop up. I can't stay on top of it myself, so I'm always looking for suggestions/maintainers from anywhere.If you have a look at the category tree, where do you think video essays would go in that?
_Chief 17 hours ago https://mysukari.com - A Diabetes management platformI got diagnosed with type 1 diabetes in Feb (technically LADA as it's late onset). I'm the first in my family with it so I had zero info on it. I tried getting some CGMs to use but most don't work in Kenya as they are geo-locked, and even apps for measuring carbs like CalorieKing are not available in my region. I was really frustrated with the tech ecosystem, and started working on My Sukari as a platform of free tools for diabetics.
I mostly get time to work on it on the weekends, so it's not yet ready for public use, but I've fully fleshed out one of the main features: Sugar Dashboard - A dashboard that visualises your Glucose data and helps you easier analyse it.
To help with demos, I've shared my Sugar Dashboard here: https://mysukari.com/tools/sugar-dashboard/peter
I'm really passionate about this and getting as much free, practical tools in the hands of patients (it honestly shouldn't be this hard to manage a disease)
jeeeb 7 hours ago I just wanted to say I had exactly the same experience this month.I was diagnosed with LADA type 1 diabetes. First in my family to have it.
My immediate reaction was wanting to put together something to track my diet, blood glucose weight and so on.
Thank you for sharing your experience.
kakoni 8 hours ago Great stuff!> I tried getting some CGMs to use but most don't work in Kenya as they are geo-locked
Are you familir with xdrip? (https://github.com/NightscoutFoundation/xDrip) It works directly with various cgm sensors (dexcom etc.)
_Chief 6 hours ago yes, came across xdrip+ when looking for an android app I could use for Libre 2. I don't think Dexcoms are sold in Kenya, and even the Libres around are UK ones so you need 1) a VPN to setup, 2) an iphone. Both things being a challenge for most - I had to buy a my first ever iphone for this. Anyway, found xdrip a bit of a challenge to setup and a bit too technical to suggest to others; needs sideload and manually disabling a lot of Android defaults.I had a lot of success with Juggluco[1] which is available on the Play Store and provides easy to use APIs to interact with supported CGM readings. Juggluco has an inbuilt xdrip web server but I haven't tried it yet.
Will definitely look into xdrip+ further.
westpfelia 16 hours ago Is this just for Type 1 or would type 2 work well also? Seems like it would?_Chief 15 hours ago All types. The sugar dashboard allows import of data from different glucose apps, so its goal is to allow you visualize and analyze your data. I hope to integrate with cgms directly if I get some that allow it, and also source from Health connect. Sharing with specific people eg doctor is also a big ask that I'm working on. The other WIP tools will be fore general health, not just diabetes, like carb counting from a photo via AIjekude 13 hours ago Also recently diagnosed and just open sourced how I'm using AI to count carbs + get insulin doses [1]. Biggest issues I've seen to starting a legit business is not having sanctioned access to real-time blood sugar values (the APIs are all one hour behind), and dealing with the FDA. Love the idea of more tech-enabled diabetes management, good luck!_Chief 12 hours ago Love this! Thank you for sharing! My backend is also in Go so this is a godsend. Will see how I can incorporate and let you know if I do!> not having sanctioned access to real-time blood sugar values (the APIs are all one hour behind)
Ah, I didn't know this. One of the prospective tools I had in mind was real time alerting in case of drastic drops eg ping doctor or relative. I think will have to be limited to the apps/tools that do support realtime.
jekude 11 hours ago Technically there is unsanctioned access (someone reverse engineered the real-time APIs [1] which I ported to Go). I think the FDA does not want easy access to real-time values so that folks can't easily recommend insulin dosing without oversight. I am personally of the opinion that it is our right to have programmatic access to the real-time data and do with it what we please.Would love to get in touch to hear more about your long-term vision for the project!
pbhjpbhj 6 hours ago >so that folks can't easily recommend insulin dosing without oversight //Is there genuinely a consideration here beyond not allowing activity without paying money to the hegemony?
jekude 2 hours ago Insulin is lethal at higher dosages, so there is definitely an argument. My counter would be that someone who has to self administer this drug 5+ times a day should have the right to make determinations about dosing
selimthegrim 8 hours ago I used to work for another diabetes management platform (NuMedics), great to see more entries into the space especially from LDCsshiggaz 13 hours ago That's so cool! Nice work!! Are you happy to share how you built and host it? How long has it taken you to get it to this point?_Chief 12 hours ago Thanks! I started out with a Nextjs full stack on Vercel, with db on Turso but ended up with a React frontend (next on vercel) and Go backend (selfhosted on vps).Decided to port the backend to Go + postgres (on a Hetzner VPS), and retain the frontend on Nextjs - A lighter weight client, moving most of the compute to the backend API. Few reasons for the port: I've had a lot more success/stability with Go backends, Turso pulled multi-tenant dbs which is what I mostly wanted them for, Nextjs is getting too hard for me.
Go backend is just the std lib (1.22+ server with the nice routing) - I mostly write all the lines in this
Frontend is textbook modern react: React19,next15,tailwind4 - AI mostly writes the code in the frontend (Cursor + Cline + sequentialthinking + context7 + my own custom "memory bank" process of breaking down tasks). AI is really, really good at this. I wrote this https://image-assets.etelej.com/ in literally 2 days 2 weekends ago with less than 10% of code being mine (mostly infra + hono APIs)
kavalg 22 hours ago I am working on the sunflower plant density estimation problem. The goal is to be able to estimate the germination rate as early as possible. Farmers benefit from such information, because:- there are lots of expenses still to be made (fertilizer, pesticide, salaries), which may not be worth it if germination is under certain threshold
- if detected early, there is still time to plant another grain or to fill up the missing plants (requires precision seeders and seeding maps)
- is a very good proxy for yield estimation (farmers often trade futures even before they have harvested)
For the purpose I have created a dataset (a collaboration between my employer and Sofia University) and published it in order to enable scientific collaboration with other interested parties. Still working on the dataset annotations.
https://huggingface.co/datasets/su-fmi/sunflower-density-est...
tringuyen_cse 4 hours ago Hey, this is interesting. I used to work on a somewhat similar problem. Our problem was more general, but one usecase is to predict the number of interactions between flowers and pollinators, given some initial counts. As these initial counts are obtained manually (by going to the fields, taking pictures and count, like number of bees within a frame), those count numbers are likely to be lower the the actual numbers. We addressed this under-counted issue using low-rankness and Poisson mixture model. Take a look if you're interested: https://ieeexplore.ieee.org/document/10888717firesteelrain 5 hours ago Feel like this basically enabling the use of ANOVA? (Compares yields across different treatments (e.g., irrigation methods, seed types).ragebol 19 hours ago Interesting, I'm also involved in a project to do yield prediction, but with a ground-vehicle with camera's on top to drive between strawberry and blueberry plants.Yield prediction is huge indeed, because overshooting your prediction means seller stuff for a lower price. Undershooting means paying for someone's product to make up for the difference. Probably there's quite a bit of matchmaking in between those under and overshooters and someone making a good buck out of that too.
Simon_O_Rourke 18 hours ago Very cool, what type of parameters are within your control if detected early?
Cyphase 19 hours ago Myself.Been a freelance dev for years, now going on "sabbatical" (love that word) imminently. Just moved to reduced hours, still in the transition and unwinding phase.
Planning to do a lot of learning, self-improvement, and projects. Tech-related and not. Preparing for the next volume (not chapter) of life. Refactoring, if you like, among other things.
I'm excited.
nicbou 16 hours ago Will you write a bit about it? I find this sort of move really interesting. I'd be curious to see what you've learned by the end of that sabbatical.Cyphase 6 hours ago I do want to write and publish more, generally. Perhaps some of it will be about the sabbatical itself.I'll let you know what I've learned by the end when I figure out what "end" means. It's not my goal to go back to what I'm doing now.
I'll post about this again the next time (end of May presumably) this "What are you working on?" thread is posted, for anyone who wants to follow along. Also, email address in my profile; use subject, "What are you working on?".
abound 13 hours ago Depending on how long you're planning on sabbatical-ing, you might consider applying to the Recurse Center [1]. It's basically a (free) 6- or 12-week self-directed learning program that you do with others.I had done a half batch last year and really enjoyed the experience.
Cyphase 6 hours ago Ha. It's long been on my someday list, since before the rename from Hacker School, and something that I've been explicitly considering for my sabbatical. I've glanced at batch schedules more than once in the past weeks. Glad to hear you enjoyed it!9 years ago: https://news.ycombinator.com/item?id=13778951
alok-g 14 hours ago I am in a similar stage.In brief:
25 years of experience including FAANG. Recently got divorced (complex litigations, fought hard, very satisfied with the outcome).
Now rethinking myself. Want to do something useful for humankind in the rest of my life. Having big ideas for the future. Trying some research-focussed 'side' projects. Considering writing a book. Learning new things. So on.
keepamovin 19 hours ago This sounds very cool! Do you have any travel planned? A move within country to more of a "reset" locale? :)Cyphase 6 hours ago I do want to travel more, in general.I'm eyeing some tech conferences in the next months, which would involve varying levels of travel and sightseeing.
I went on a ~three-week roadtrip once, which I saw as a shorter/practice version (it was originally planned for two weeks) of a longer one I'd like to do one day. So I might do something like that.
cookiemonsieur 19 hours ago I like the concept of "refactoring" one's life.kridsdale1 8 hours ago When I move cities it feels like a rebase.
nicksergeant 5 hours ago I'm building a platform for small-city business directories called TownWire: https://townwire.comWe launched our first location last week and have had a great response so far from business owners and residents alike: https://canandaigua.com With the current tech climate of AI and big tech, things have gotten so impersonal for the majority of actual people. I'm betting on communities and the small businesses that serve those communities in the medium and long term.
raybb 4 hours ago I love this idea! Do y'all push/pull the data to OpenStreetMap at all or the users have to add it all manually?nicksergeant 2 hours ago Not at the moment but that's an interesting idea! I didn't want small biz to have to lift a finger to be on the site, so we manually populated 550+ biz to start, and then I paid a photographer to take shots of about 80% of the spots around town.I really like the idea of contributing those photos back to OpenStreetMap actually... right now I have Google Maps on profile pages but only because we absolutely need the Google Places API for accurate hours (that's really the only spot businesses update current hours at the moment). But I could see swapping for OSM at least for maps in the near future.
tonyonodi 11 hours ago I launched my web-based notepad calculator, https://numpad.io/, a few years ago.Right now I'm working on a version 2 that has user accounts, multiple documents, markdown support, and document exports. Everything is local-first and it uses CRDTs to sync documents.
It looks like this: https://i.imgur.com/Plk1DQ4.png the calculator is mostly the same for now, with a few improvements. It's unstable right now, so I don't want to publicise the dev url, but if you'd like to become a beta tester email me at contact@numpad.io
thunkle 17 minutes ago I see this on the example: 6 ft 2.0000000000000000004 inchesrogeliodh 2 hours ago I would love a pad where i could dump any text with numbers and it will sum/avg all of themch33zer 9 hours ago I think this is super cool, but feel like the killer implementation would be an extension for VSCode. Just a thought.GlacierFox 11 hours ago That 'per' and 'to' syntax you got there is chefs kiss.tonyonodi 11 hours ago Thanks! Though I can't really take credit for it, Soulver did it before me.
kevinqi 7 hours ago super cool
barrell 22 hours ago I have been working on http://phrasing.app - a language learning & acquisition tool for polyglots. I’ve been using it to study ~12 languages (5 on maintaince, 2 seriously studying, 5 casually “studying”) and it’s starting to feel really good. If anyone is learning/maintaining several languages, please reach out! I’m looking for beta testers in as many languages as possible (it supports 120+).In what I believe is still the spirit of the question though, I discovered Maltese these week and have added it to my casual study. It’s a Semitic language (closely related to Arabic), written in the latin script, with about 40-50% of its vocabulary being Italian/Sicilian based. It’s become my new obsession
franklin_p_dyer 3 hours ago As someone who is both an avid language learner and a software developer - what’s the value added in this platform, for someone who is already pretty comfortable as a programmer and autodidact?It would take a lot to convince me to pay that much for a product like this. True, it can be inconvenient trolling around for content in your target language, but as a software dev I am pretty experienced with finding obscure things on the internet by finessing search queries. And there are plenty of other apps out there that do spaced repetition for you, and open source tools and data sets that can be used to help you scrape/process vocab (again, if you don’t mind spending some time debugging, which I personally do not). Besides that, I really don’t find it that inconvenient to manually write down words/phrases from books or movies and copy them into my SR deck. On the contrary, I think this overhead actually helps the phrases stick better!
So how would you sell your site to someone in my situation? What would I get out of it?
bogdansolga 12 hours ago I have just created an account and tried to use the app; either I am dumb (at this hour), or the app has a very not intuitive UI :-)I (think I) managed to create an expression, but: 1. it takes forever, I don't know what it's doing and it's still not done 2. I have no idea how to use it, onward 3. does not seem that I will be actually able to use it, as the app requires to be subscribed to use it...
Looking forward for a how to use manual / page + a real trial period. If the app requires a subscription with this UX experience, I will be gone :-)
barrell 11 hours ago I am the first to admit that the UI does not make sense to people who have not seen it in use. I have done some user onboarding sessions where you just “let them ask questions and click” and it goes terribly if it’s any consolation, I have never gotten that feedback after showing people how to use it (quite the opposite actually!)Wrt to the expression, all expressions created today succeeded, so if you’re still seeing a progress bar let me know as that’s a bug. It’s possible something failed with the live updates, or it does take several minutes to create an expression (depending on servers, it can take up to 10 minutes at times, although the typical timing is 2-4 minutes depending on the expression)
If you click on any of the review methodologies, it will start reviewing any of your successful expression. From there, the experience should be a lot more explorer-friendly :)
What it’s doing is: analyzing the sentence, splitting it into phrases, aligning it across all languages, tagging all of the gender/case/tense/etc, researching pronunciation, generating audio, aligning the audio, prioritizing the words (across several axes), and generation explanations/dictionary for each individual word
_puk 20 hours ago Sounds good.What languages do you support?
Learning Latvian through Anki flashcards, but it's not well supported by the main platforms, and there's not a huge amount of content out there for learning.
This alongside a couple of the usual suspects.
As a side note, on a Pixel 4a 5G (old phone , but functionally not ready for e-waste) the homepage bleeds all over. Some components into each other, others off screen. Might want to check that.
barrell 18 hours ago Oh no, the website is brand new, it should be working everywhere. I'll have to dig up an older android, I should have one somewhere.Languages below, if you know their alpha 3 code. Currently having some issues with Thai and Zulu though, so they're temporarily disabled until I have time to fix them.
I have not ~tested~ verified it for Latvian, I would be curious to hear your thoughts. It has been working pretty well for Maltese, Albanian and Macedonian though, which should be lower resource than Latvian!
As mentioned elsewhere, the first time user experience is abysmal. If you reach out though we can hop on a call and get you set up - or in a few weeks I'll have a video done and up. In the meantime, you should be able to create an expression (in the nav bar for desktop and mobile) fairly intuitively.
afr, amh, ara, ara-are, ara-bhr, ara-dza, ara-egy, ara-irq, ara-jor, ara-kwt, ara-lbn, ara-lby, ara-mar, ara-omn, ara-qat, ara-sau, ara-syr, ara-tun, ara-yem, asm, aze, bel, ben, bos, bul, bxr, cat, ces, chu, cop, cym, dan, deu, ell, eng, est, eus, fao, fas, fil, fin, fra, fro, gla, gle, glg, glv, got, grc, guj, hbo, heb, hin, hrv, hsb, hun, hyw, iku, ind, isl, ita, jav, jpn, kan, kat, kaz, khm, kir, kmr, kor, lao, lat, lav, lij, lit, ltc, lzh, mal, mar, mkd, mlt, mon, msa, mya, myv, nan, nep, nld, nno, nob, ori, orv, pan, pcm, pol, por, por-bra, por-prt, pus, qaf, qpm, ron, rus, san, sin, slk, slv, sme, som, spa, spa-arg, spa-bol, spa-chl, spa-col, spa-cri, spa-cub, spa-dom, spa-ecu, spa-esp, spa-gnq, spa-gtm, spa-hnd, spa-mex, spa-nic, spa-pan, spa-per, spa-pri, spa-pry, spa-slv, spa-ury, spa-usa, spa-ven, sqi, srp, sun, swa, swe, tam, tel, tha, tur, uig, ukr, urd, uzb, vie, wol, wuu, yue, zho, zht, zul
EDIT: I have tested it for Latvian, I know it technically works. I however have not had any Latvian speakers review it's quality
helenite 16 hours ago The UI was a little unresponsive on mobile, and when I opened the "Media" page on desktop I got multiple block rendering errors. Opening the console reveals a syntax error (missing ] after element list) and some type errors.Also, it looks like you have to get the subscription to use it in any way? It's hard to gauge whether it is for me or not if I have no way to trial it. I found the UI a bit confusing too, I was not sure what I was supposed to do after logging in. As another commentator mentioned, it's asking me to set a reference language but I see no way of configuring it.
barrell 16 hours ago Block rendering errors on the media page is new to me. I will look into it.The reference language error should not be shown (I mean it’s not incorrect, but there is a “no expressions error” that should take precedence).
A video is coming :) I didn’t expect so much interest from a comment in this thread. If you get in touch, I can walk you through it personally, otherwise check back in a couple weeks and there will be a video overview.
yurishimo 20 hours ago Since you're in Amsterdam, I'm curious how well you think it performs for learning Dutch? I'm a native English speaker with a B2~ in Dutch and just looking to progress more. I've not used spaced repetition up to this point in my learning journey (almost 3 years).barrell 18 hours ago It does really well, 95% of the time. The application was built to jump in at any levels - find a movie you want to watch, align the subtitles, see the most important words, create expressions with words you don't know, and the SRS should focus on the words most important to you.For my Dutch (which was probably once a high B2, now probably a low B1) I only use the audio review when walking my dog or cooking. It plays the audio of the cards in a playlist, so I practice hearing and repeating them.
It's not so self serve at the moment, but if you get in touch I can get you up and running.
KerryJones 13 hours ago Great job on the design. I like the idea here, but the app was unresponsive on Windows > Chrome to most clicksMumps 17 hours ago I was quite eager to check this out. As some polite feedback, a few things turned me off quite strongly:1. I want to get confirmation that the language I want is covered (Hungarian). "120+" doesn't confirm it for me, as Hungarian seems fairly rare for language apps. Can we not just have a "search your language" field?
2. I need to see what the app actually looks like, how it proposes it'll teach me.
I'm one of the eager-to-pay people, because Duolingo is frankly dogshit (ok. Mostly polite) at teaching languages (doubly so ones that it doesn't care about like Hungarian). But I'm so suspicious of language apps, due to being burnt a dozen times.
barrell 16 hours ago Thanks for the feedback! I agree with you completely.1. I just started the marketing website a few weeks ago, and if you can believe it, I didn't readily have that information. One of my tasks last week was to compile a list of languages that could work, write some tests for all of the languages, and get a list of supported languages. I have that list now, I just need to put it on the marketing page.
2. As mentioned in other comments, I'm working on a video. I'm preferring to fix glaring issues before making the video, although at this point I'm verrrrrrry close. I have started scripting it, but it takes a lot of time to make a good video (1-2 full days if I don't want to edit it).
Your feedback is completely valid, and they're both reasons why I'm not really marketing the product yet. This thread seemed like a good middle ground though as having some people using all the languages would be really helpful. Also, I've genuinely been loving using it and want to share.
It's just me working on it, so these things are coming, but everything takes a while! Hopefully these didn't sour you on the project permanently :)
EDIT: And yes, it supports Hungarian :)
Mumps 12 hours ago Thanks for addressing, really!Nope, not soured. And don't worry, I totally get that things take a bunch of effort and time (doubly so as a solo project). I'll give it a re-look in a little while :)
barrell 16 hours ago And fwiw, I've added the languages to the marketing page now in the FAQ section. I'll add a more prominent section in the coming weeks!
android521 19 hours ago i signed up and tried to use it. The UI is very confusing. i couldn't find the place to setup what language i want to learn and what language i know (for translation). It is best if you can have a video or images documenting how to use it.barrell 18 hours ago Agreed. I’m getting close to a video to put on the landing page, probably some time next week.The first time user experience is really bad, but the app itself makes a lot of sense once you see it in action. Feel free to get in touch with me (there are several methods listed when you log in) and I can give you a personal introduction!
If not then check back in a few weeks for a cool video :)
muzani 21 hours ago Yes, please. I've been looking for something like this. Lately I've been just casually going into another language with ChatGPT and asking it to correct me. I do I like some of the old languages, things like Aramaic, which just have a different feel.I signed up, but now it's asking me for a "reference language" (which is a little ironic because it tells me this in English lol). I guess I'll play with this later.
barrell 18 hours ago Start by creating some expressions ("create" in the nav bar) and you should be able to play around with it. If you want to learn more, please get in touch. As mentioned elsewhere, I'll be adding a video tutorial soon (probably not this week, but sometime next week if all goes well).Would love to get feedback on the old languages! It's been really good for the minority languages I'm learning
pandemic_region 21 hours ago Would be great to be able to login via Google or Facebook. Creating an account is cumbersome on mobile.yard2010 20 hours ago "...are we still doing phrasing?"barrell 18 hours ago So much of this stemmed from me wanted to learn French & Italian by watching archer :D
pheelicks 18 hours ago A pentagonal geospatial indexing system: https://a5geo.org/If you’ve used H3 or S2 it should be familiar, the major difference (apart from the fact it uses pentagons) is that the cell areas are practically uniform, whereas alternative systems have a variance of around 2 between the largest and smallest cells, making them less useful for aggregation. The site has many visual demos, e.g. https://a5geo.org/examples/area
The code is open source: https://github.com/felixpalmer/a5
stevage 13 hours ago Wow, that's a really fascinating problem to work on. I work in mapping but haven't actually come across this field of what you call DGGS's.Is it essential that the cells be the same shape?
Also where does the name "A5" come from exactly? I get that 5 is because it has five sides, but why A?
pheelicks 12 hours ago The cells being the same shape is useful in some use cases and irrelevant in others. For example, see the Airbnb demo: https://a5geo.org/examples/airbnb. The H3 tiles are very different sizes in the two cities, and make it appear that there is a much higher density of listings in Malta, even though that is not the case.However the symmetry of H3’s hexagonal cells lends itself well to flow analysis, or routing - which is no surprise as it was developed at Uber.
As for the name, it follows the convention of S2 and H3, which come from group theory and refer (loosely) to the symmetry groups of the various systems
westcoast49 15 hours ago Is it based upon repeated subdivision of an icosahedron?pheelicks 13 hours ago No, it is based on applying a lattice onto the faces of a dodecahedron (technically a pentakis dodecahedron). Take a look at https://a5geo.org/examples/teohedron-dodecahedron and other examples on the website.H3 is based on a dodecahedron it is it the reason the cell areas range so much, the same is true of S2 - but this is based on a cube.
vldszn 4 hours ago I’m working on a free and open-source invoice generator with live PDF preview — fully browser-based, no sign-in required.It supports multiple languages, currencies, European VAT deductions, and more.
I built this tool for myself so it’s kinda like a personal software. Hopefully, others will find it useful too :)
FlyingSnake 21 hours ago I hacked my old kindle and turned it into a eink dashboard for my daughter’s school! Now planning to enahace it a bit further and make it easy to customise.Here’s a a detailed write up of the process: https://samkhawase.com/blog/hacking-kindle/
A4ET8a8uTh0_v2 19 hours ago Genuinely brings a smile to my face. And such a nice exposure for the kid.FlyingSnake 17 hours ago Thank you for the kind words. I’m trying to get back into writing and such feedback means a lot to me.
pranav7 19 hours ago Love it! Nice work
sci_prog 1 hours ago I am building a service to make accessing AI as easy as sending email called ThreadWiseYou can email prompts directly to your ThreadWise address and get instant AI-powered responses, essentially an always available co-worker. Another great feature is the ability to schedule recurring tasks and since the AI has web access, you can get things like:
Daily mortgage rates or airfare price monitoring
Weather and news summaries
Sport scores, jokes, quote of the day
Pull data from public APIs (and more)
So you can essentially use it as a personal newsletter, crafted to your taste.
The free tier will let you test this out for free! I am looking for some feedback/criticism, testing, and additional ideas and I am open for collaboration if you have experience with sales. Also open to hearing which scheduled tasks people would find most useful.
Why I built it: I noticed a trend online, as well as with family/friends, that people would like to have a quick access to AI in instances where they couldn't always install apps or use browser-based tools (such as in remote/low bandwidth environments). This is when it him me, email clients already have all the features needed to interact with an AI (text + attachments) and I quickly got to work.
Some of the advantages are also that since there are no new apps, or browser tabs needed, the tool is ideal for companies who don't have the bandwidth to setup full fledged AI solutions on their own. The companies can choose either between public LLMs (e.g. OpenAI) or host everything on-premise with locally run models, so no data ever leaves the premises.
Eager to hear what you all think!
perihelions 20 hours ago Reverse typesetting: reflowing page layouts where you don't have knowledge of the typesetting structure, i.e. a scanned physical book or PDF paper. Naive rules-based heuristics based on the dimensions of bounding boxes and gaps. Point is to reflow things for resizing to eink readers. (Specifically the size that fits in my pocket which I carry around. User #1 is me). Building in Common Lisp and targeting an Emacs mode for interactive execution with manual feedback.perihelions 6 hours ago Err, here's a visual explanation of what I mean by this, from my REPL:https://ibb.co/album/MDw79y?sort=name_asc
(The source example is from David Tong's physics lectures notes, that were featured on HN last week — https://news.ycombinator.com/item?id=43763223 )
Foreignborn 19 hours ago i don’t quite understand, what makes it reverse typesetting?my understanding is your typesetting books for responsive eink readers.
perihelions 19 hours ago You're inferring the structure of the document from the printed result. If typesetting takes a set of layout directives and outputs a page, this is taking a finished page and guessing what layout directives could create it. Then you can take that inferred structure and reflow the page in a new layout.froh 19 hours ago so like ocr but not recognizing characters and words but recognizing the layouted structure and transforming it into content markup and layout markup?perihelions 18 hours ago That's a way to view it!The reason I'm not falling back on OCR is because the general case is full of things, like math equations and inset graphics/diagrams, that can't be OCR'd. The only robust way to deal with those is to treat them as graphical atoms: "this bounding box can be moved around, but should not be split up into pieces".
leansensei 7 hours ago I'm working on my second self-published technical book related to Elixir. The first one has been a surprising success, but since it was about learning to use Ecto on the basis of an old pedagogical toy database (and tiny dataset), the new book is about the development of a production-grade REST API with Elixir and Phoenix.I find that many books out there are focused on documenting the "happy path" for rather small and simplistic applications without boundary conditions or business thinking behind them.
So, I thought (as with the first books) that it mix things up by also documenting the business context, the questions, decisions, and the decision-making process itself, as well as all the gotchas and "side-quests", rather than showing "here's how you do it" and then expecting the reader to suddenly make the jump from tutorial hell to actually software engineering.
Overall, it's enormously enjoyable, and I hope it goes as far as the first book, and possibly even farther.
matt_heimer 12 hours ago It tried out vibe coding this weekend. I've been using AI to get smaller examples and ask questions and its been great but past attempts to have it do everything for me have produced code that still needed a lot of changes.I hit an OpenSearch bug this week where you can't get any browser based requests to work. Its due to zstd becoming a standard part of Accept-Encoding and OpenSearch not correctly supporting it so I wanted to install a browser plugin that modified the browser HTTP request headers to my servers.
I don't know about everyone else but I love that browser plugins are possible but I hate having to find them. Its mostly due to never knowing if you can trust a plugin and even if you find one, you have to worry about it being bought out in the future. With vibe coding I was able to build a browser extension in 45 minutes that had more features than I originally planned for.
I spend more time documenting the experience than building which is wild. If you are interesting you can look at the README in https://github.com/mattheimer/vibe-headers
But I left the experience with two thoughts.
Even seasoned developers will be using vibe coding in the future.
I think in the near future the browser plugin market will partially collapse because eventually browsers will build extensions themselves on the fly using natural language.
nonethewiser 12 hours ago In my experience, "vibe coding" can produce a rich prototype very fast.Then as scope expands you're left with something that is difficult to extend because its impossible to keep everything in the LLM context. Both because of context limits and because of input fatigue in terms of communicating the context.
At this point you can do a critical analysis of what you have and design a more rigorous specification.
benjaminsky2 6 hours ago If your project is well organized and individual files are small and the dependency graph isn’t too crazy, Claude code does an amazing job building only the context it needs even as the project grows. You just have to be aggressive about refactoring for maintainability. The bonus here is that it’s easier for humans to work on too.matt_heimer 11 hours ago I don't disagree but context limits are expanding rapidly. Gemini 2.5 Pro which was used here has a 1 million token context window with 2 million coming soon. Cost will be a concern but context size limits will not.loufe 11 hours ago Totally agree, I mentioned it in another comment but Gemini was a game changer for allowing me to increase the size of the project I can feasibly have AI work on.Only issue is Gemini's context window (I've seen my experience corroborated here on HN a couple times) isn't consistent. Maybe if 900k tokens are all of unique information, then it will be useful to 1 million, but I find if my prompt has 150k tokens of context or 50k, after 200k in the total context window response coherence and focus goes out the window.
hnuser123456 10 hours ago I'd love some more innovation on increasing context size without blowing up RAM usage. Mistral small 2503 24B and Gemma 3 27B both fit into 24GB at Q4, but Mistral can only go up to about 32k and Gemma about 12k before all VRAM is exhausted, even with flash attention and KV cache quantization.theyinwhy 8 hours ago What editor are you using with gemini 2.5 pro? I really don't like their vscode extension.
falcor84 12 hours ago But my reading is that the parent wouldn't have tackled this at all without the vibe coding, and would have used an off-the-shelf extension. So in that case, it's a pure win, no?nonethewiser 12 hours ago Yeah I could have framed it better. I was responding to:>I've been using AI to get smaller examples and ask questions and its been great but past attempts to have it do everything for me have produced code that still needed a lot of changes.
In my experience most things that aren't trivial do require a lot of work as the scope expands. I was responding more to that than him having success with completing the whole extension satisfactorily.
begueradj 12 hours ago So there is no technical debts ?matt_heimer 11 hours ago That's a difficult question to answer because I don't know if I'll grow the extension in the future. Only time will tell.After I completed the extension I did try on another model and despite me instructing it to generate a v3 manifest extension, the second attempt didn't start with declarativeNetRequest and used the older APIs until I made a refinement. And this isn't even a big project really where poor architecture would cause debt.
Vibe coding can lead to technical debt, especially if you don't have the skills to recognize that debt in the code being generated.
weakfish 14 hours ago A note taking app - I know, I know, been done a billion times. But really, no software I’ve found works the way I want it to, so I’m just trying to write my own, starting with a front-end agnostic core and a TUI.If it strikes a chord with anyone, I’d love to collaborate! The concept is centered around organization bubbling up naturally from dumping info in with tags, and “typing” your tags so that when you go to a tag’s page, the layout is customized based on what it is - a project, person, etc. A project could have all relevant tasks and notes listed, whereas a person might have name, contact info, etc.
weakfish 13 hours ago Expanding some - the tags concept is similar to AnyType[0] types, but the rest of the software I’m writing is more oriented towards dump-first, tag, and let it sort itself out, whereas AnyType requires careful management and configuration of the workspace.
siruva07 15 hours ago PodSnacks (https://podsnacks.org) — Personalized AI summaries of top podcasts, delivered to your inbox.I've been building PodSnacks because I found it overwhelming to keep up with podcasts across tech, business, and science. PodSnacks uses LLMs to summarize the most popular episodes from shows like Lex Fridman, Acquired, All-In, Invest Like the Best, and more.
You choose your favorite shows, and we email you short, high-signal summaries — no audio to skim through, no endless backlog guilt.
So far:
126K+ episode summaries generated 92K+ hours of podcasts processed 48–50% open rates 2,900+ early users
Still iterating and adding features like bundling by theme, language translations, and audio feeds.
If you're the kind of person who wants more inputs without more noise — would love for you to check it out.
Always open to feedback from HN!
akudha 14 hours ago What are the costs like, for such a service? You have to spend for transcribing (even if you ran something like Whisper) and pay for APIs to summarize, correct?hsuduebc2 14 hours ago I'm also interested in this. :)
jbverschoor 15 hours ago I’d prefer summaries of my YouTube subscriptions, where I can also add my own “summary template”In currently have a prompt for it that works for me, based on the transcripts.
Problem: too much duplicate information in any type of publication and too much fluff
Problem2: YouTube/transcriptions
Probably wouldn’t pay for such a service, but would be very happy using it. Perhaps some channel promo / email based ads for discovery or recommendations.
qwertox 15 hours ago Interesting. Once signed in, how do I search for podcasts? Discover is not showing a search field, unlike https://www.podsnacks.org/newsletter/podcastsEdit: Ok, I found the search function in the hamburger menu. A bit unintuitive.
bcye 13 hours ago That's awesome! I'm really not a podcast person, but a lot of good content is available primarily through podcasts. Hope that can help with it.atlgator 14 hours ago Just signed up! Do you use different AI prompts for each podcast based on the specialty or a single prompt?quichenp 11 hours ago any way to expose api access for the keyword search feature? would be cool to build this into some topic summaries delivered via email I'm building for my team
iamthepieman 14 hours ago Creating stealth group in a huge Fortune 500 company with the blessing of my immediate boss but no other higher-ups. Trying to productize critical consulting tool sets in the utility industry so we can stop repeating ourselves for the 100th consulting engagement.Yes, customer is a special snowflake but they still need 90% or whatever every other client in this industry needs.
Feeling increasingly like this is a fools errand.
Even though we've proved this out with tool sets strung together with duct tape and safety pins, and are therefore the most profitable group within our department, we still need to be 100% billable.
It's only because we're the most profitable group that we can pretend we're all billable while I work with two other people to bootstrap this crazy project
Edit: anyone hiring? Just found out my boss is quitting.
collingreen 13 hours ago Oof. This post started so good and then got progressively more sad until the edit nailed it home. I hope your story continues and works out as a huge win, either as a new, good boss, you getting to openly lead this kind of thing, someone reading this and poaching/sponsoring you, or maybe even you working on this under your own name.Good luck and we're rooting for you!
iamthepieman 12 hours ago Thank you for the enthusiasm!It was not intentional but my post really does read like a little story vignette that ends with a gut punch.
Not looking for sympathy so much as fellow appreciators of irony and schadenfreude but here's another kicker.
I pitched this idea to my previous company and was told there was no appetite for it. Just saw on my old company's blog that they released a "digital transformation in a box" program for mid-market clients in this space which is 90% of what I pitched to them. Bad and hilarious timing all around.
nathan_douglas 11 hours ago Good lord, how many times are you going to get punched in the gut today?
pajop 12 hours ago Very interesting. Maybe we can chat and explore? DM me on my X.com - it's on my HN about page - copy-paste the HN link here for context :)toss1 4 hours ago This sounds very much like an application begging to be done as a stand-alone company supporting these F500 companies. Could be very profitable as the basis for a service-provider model while you gain enough knowledge to product-ize and package it on basically customer-funded development. It seems your company kicking you in the gut is showing you the directionGood Luck!
BSTRhino 5 hours ago https://easel.games/aboutA programming language for teenagers to learn to code by making multiplayer games. I've spent 3 years making the multiplayer completely automatic so you can just code it like a singleplayer game, then flick a switch and it just works. My hope is that teenagers will find it more engaging if they can play their games with their friends. A bit like a combination of Roblox and Scratch.
Currently trying to implement some region affinity so it doesn't just put everyone in the world in the same game. It's rollback netcode so the latency is very good even across continents, but it can't overcome the fact that the world is just too large.
spudlyo 7 hours ago My project is an ongoing one of self improvement. I've spent most of my adult life working on technical stuff, and it's taken a bit of a toll on me.I quit working about ~20 months ago, started a low-carb time restricted eating regime, lost ~230 lbs, have been doing 15-25 hours of cardio a month for the past year, started going to therapy, got an ADHD diagnosis, read a bunch of classic literature (Middlemarch and The Count of Monte Cristo are my favs thus far), maintaining a 19 week streak of Latin language learning through the killer Legentibus iOS app, and I'm playing guitar every day (trying to nail the the major scale in three different fingerings across all 7 modal starting points).
I miss my old job working with Vitess and Kubernetes a lot (Hi Sam!) but eliminating all work stress has really allowed me to take control of my life.
devilsdata 46 minutes ago I was diagnosed with ADHD last year, and am curious as to what you've learned in your time. I'm taking 40-50mg methylphenidate hydrochloride (Ritalin) daily and working a hybrid web developer role. I'm trying to increase my reading of literature and begin writing, but I find myself just watching YouTube/browsing Reddit and HN.230lbs is wild. Great job :)
gtowey 38 minutes ago Diet matters. Protein, especially in the morning. Carbs and sugar are so bad for focus.Sleep is essential. Getting a full night of quality sleep will probably help more than anything else. If you often wake up feeling tired already then maybe do a sleep study to see if there are problems there.
Exercise helps.
Developing a system for organization is key to unburdening your mind. There are various books for adults, for me I also found an executive function coach gave me the right amount of accountability and discipline to practice the systems enough that they didnt feel like a burden anymore.
coyotespike 2 hours ago Yaaayyy Legentibus! I only dipped into it but next time I return to Latin really look forward to working through all the material. Extensive reading ftw.prmph 7 hours ago Really inspiring; most of what you've done are things I've been trying to ge the wherewithal to do since like forever. I wish you well.samlambert 7 hours ago so proud of you Mike. we still regularly talk about how we miss you. i am so glad you are doing well.javiramos 7 hours ago 230 lbs in 20 months -- that's amazing. Congrats.
drev 15 hours ago I am working on a super mario land re-implementation for game boy using C. It's made with the GBDK toolkit so the ROM can be run on a gameboy. Currently the scrolling and background collision has been implemented, I am working on drawing the enemies... their is still a long way to go. As for why I am doing this is nostalgia and fulfilled a childhood dream repo at https://github.com/odrevet/marioland-gbdk/merolish 14 hours ago I remember SML fondly, best of luck.cedel2k1 15 hours ago Love it, good luck with the project!
tha00 22 hours ago I'm currently working through Frank Sikora's "Jazz Harmony" book [1] to learn to play Jazz piano.In parallel, I'm building an exercise generator "Jazzln" [2] to help me practice.
[1]: https://www.goodreads.com/book/show/54391815-jazz-harmony
Joel_Mckay 19 hours ago Here is a weird album you might find fun:https://www.youtube.com/watch?v=8OJHPWlaCrc
Cheers =3
nake89 13 hours ago Creating a modern development environment for win98. Currently I’m working on a git client. I have no patience for learning C. So backporting git is not an option. I also don’t want to use cygwin. So I’m using a server to expose git as http endpoints and coding a git client in php to use in msdos. I have Vim 7.3 and and gnu coreutils working in msdos already. So soon I will have a very nice dev environment. I want to create a one click installer which gives you xampp, vim and then all the tooling I’ve created. I’m also interested in creating SPA that works in IE5.5. But I’ll do that when my tooling is ready.joshuaissac 9 hours ago For Git, you may be interested in JGit, which runs on the JVM. From some MSFN forum posts, it seems to be possible to run Hotspot JRE 8 on Windows 98.
