Ask HN: Who wants to collaborate?
This thread is similar to the monthly "Who is hiring?" and "Freelancer? Seeking freelancer?" threads.
But this one is for people who don't want to work for money and are not looking for people who want to work for money. But for people who want to work together on cool projects.
For free to make the world better or to start a startup.
If you do, please post your project or your skills!
513 comments
[ 3.0 ms ] story [ 338 ms ] threadThe idea is to put together a project that gives an overview of how to set up a minimal viable web application from scratch via all the different frameworks.
For each framework the project features a self explanatory shell script that builds a web app with routing, templates and user accounts. So there is no ambiguity of how to reproduce the results. And it is even possible to just copy&paste the steps into a docker container and see the framework in action.
Here is the repo:
https://github.com/no-gravity/web_app_from_scratch
So if you want to compare how the frameworks do templating, you can look at the "Let's use templates" part and have a quick overview of how it is done in Django, Laravel, Flask, Symfony, NextJS...
I wrote the beginning of the Django script and two developers contributed Laravel and Symfony scripts. So far we have routing and templates. All three still need user accounts.
If you are experienced in a web framework, feel free to add to one of the scripts or a new one and send a pull request!
The main difference is the "no ambiguity" approach I take. When you look at the Django part of the project you linked to, you see it starts with 8 manual steps and then a bunch of optional steps in case it does not work.
I want to have no ambiguity. No manual steps. That is why every framework is handled by a script that is guaranteed to work on Debian 11 and leads to a running web app.
https://twitter.com/marekgibney
So on the sidebar you have components...one is a sidebar right? so you click on it..on the right pane it shows implementations on the top of the pain is a dropdown so you can switch from say. MUI to AntD to Vuetify, etc... some are narrow scoped frameworks (Vuetify for example is only Vue), some are Wide... Tailwind Kit for example or Daisy UI... they target tailwind but are JS Framework agnostic...
So under each implementation of each component might be code for Rails, Laravel, Django, views and maybe even some sort of api integration where you can have a webhook that can either connect to your dev and push the code to the right places, OR it could put it in a queue if your local dev can't receive incoming normal http requests, you could just use a cron job or manually import the code from your localhost backend...
I'm imaging this sort of being like a mixture of low-code meets Storybook with already developed ui components, and you could even import full templates, etc... So for backend devs w/ an idea 90% of the UI stuff is just drag/drop etc.... you could even have a way to combine elements somehow so you can build up a whole page of elements...and some sort of marketplace where people could share designs and maybe charge like $1 for the code snippet MIT license, but access to the code is $1.
Edit: It'd also be cool to extend this to data-layer..so say I am designing my migrations/models... and I want an Employees table....well maybe I type in employees and it organically lists all the columns previous people have used for employees tables and ranks them by how often it was used... so you can basically have most of the data points that are pretty common across similar apps be shared and almost done for you...and then you can export to django, rails, prisma, hasura, laravel, etc...
So if anyone has an idea that want to develop or a project in Go already let me know.
I'd be interested in trying this out again.
If you're interested, please reach out. My username here is the same as my username on the Gophers and Kubernetes slack groups.
(You're of course welcome to just go pick up an issue in the repo if you'd prefer)
I commented about my current project above that is in need of some Go expertise. I am currently working to get into the summer YC class, so if that interests you, feel free to use the email in my profile!
I'm starting a static web host where every page is remixable by default.
One click and you have a copy of any web page hosted on the platform.
Not only that, but the front-end code (HTML, CSS, JS) is treated as the source of truth. Whatever changes the owner makes to a page (through DevTools, some custom JS, or clicking a button that does something) is saved permanently to the DB.
This makes it possible for front-end devs to create interactive applications just by adding a JS plugin to the page.
I'm hoping projects like this idea will lead to a renaissance of blogs, personal websites, and remixable apps that push our normal conception of what a website is to the breaking point.
(Email in my profile if you want to discuss. This is a new project I haven't started yet.)
If I want to see an older version of the page I like: click click click I have a (like really) permalink
If I wanna modify a page that someone else made, fork and edit!
It's like Github, but for our pages!
(too many exclamations smh)
The web has gotten boring.
This is so true. Before it was boring, it was hard to use, though. I think that's why the federated/decentralized/indie stuff still struggles to find mindshare.
There still seems to be a lot of interesting creative content out there, but it doesn't play by google and social media's rules, so what you find is listicles, blogspam, click funnels, various forms of low effort crap made in bad faith.
Here are just a few links to websites I've found that I thought were interesting (I'm not affiliated with any of them). Tell me, when was the last time you saw that type of stuff on google, facebook, or reddit, or twitter, or HN?
http://sod.jodi.org/index.html
https://dreamcult.xyz/
https://www.floppyswop.co.uk/
http://godxiliary.com/
https://dannarchy.com/
https://www.toiletpapermagazine.org/
I think it should be possible to build something genuinely useful in this space. I'm not exactly sure how or what, so I'm still experimenting with various ways of bringing audiences to websites off the beaten path, but it feels like I'm closing in on something.
I've been approaching this in a sort of independent R&D fashion, just trying things and seeing what sticks. If anyone is into this general space, maybe shoot me an email. Not really sure what's a good format for collaboration though.
I called it MIXINT for remixable interface. It's abandoneware now but IMO a pretty good vanilla-JS approach to window panes with class-sensitive menus. see github for code and manic dreams in the readme [0] and video of the interface swapping out one class of window for another [1]
But what I really wanted was to expose the guts of the code in a way that made it easier to edit, and as a protest to JSX I threw together a unified JSON syntax for templates that get rendered into HTML/CSS so that I could write javascript code that returns objects and just build up DOMs that way - but when I started implementing variable substitution and array mapping in my JSON templates, I realized I could do away with the javascript altogether and write entire web pages as declarative JSON, just using a few prefixes that trigger macro expansions, replacing one branch of the tree with another until no more prefixes are left (the eureka moment came when reading about macro expansion in the TRAC language - I was just going to stick to feature parity with Handlebars before realizing I had a bone fide programming language on my hands)
This is all off and on the last 4 years. I'm about to overhaul the years-old homepage with better examples but you can see the concept here [2]. Maybe some of the code would be useful, I'm certainly on board with the mission, I just wish I was better at avoiding those pesky rabbit holes :)
[0] https://github.com/jazzyjackson/MIXINTvNegative1
[1] https://youtu.be/URMHbhK0zrs
[2] https://lookalive.software
The product seems less ‘the web has gotten boring’ and more ‘we are low code development’ — a happy fun place between no code development and the normal overly complex web development of today.
Which funny enough, I think your audience/clients are best for actually big boring development via some enterprise clients who require stricter security rules or giant less skilled employees with occasional custom requirement (imagine a big painful US gov type organization). I don’t see this getting hobbyists but easily could be wrong.
Cheers and congrats. Really cool stuff! Excited to see where you land.
Demos: https://www.reportmill.com/snaptea/
Repo: https://github.com/reportmill/SnapKit
There is a lot of benefit to this - fast, traditional desktop development with great deployment. It also has built-in developer tools and online access to a UI Builder. I hope this could make Java Client development interesting again.
I need collaborators to help write demos and improve the kit.
Jeff at ReportMill dot com
No major benefit of using a dht except it's kinda cool that the entire thing runs sans centralized database.
If you're interested, the space is what my startup is working on but the cdn specifically is mostly a fun tangent to learn about kademlia for me. It's always fun to work with more people in the live video space since it's a small community overall.
In principle I think BitTorrent can be adapted to this, but essentially the output "file" doesn't hash correctly so I think it would be a lot of hacks to the point where it's avoiding actually using most of the parts of BitTorrent anyways. In particular this is evident on the DHT, my understanding is the DHT actually stores fragments of data and BEP44 verifies integrity. In my implementation, the DHT stores some metadata for who can publish the data because the media stream is live and we don't want to broadcast every packet over the DHT.
Of course, I'm new to BitTorrent so that's my understanding. I'd love it if I could use more BitTorrent components directly :)
I also know PeerTube has a Live feature now, but I have no idea how they do it. I think they use WebTorrent (at least for the non-live videos).
[1]: https://arstechnica.com/information-technology/2016/05/bitto...
My goal for the CDN is to make it lower latency that it could be used for video conferencing. In practice I have more of a one-to-million distribution in mind but it's a cool puzzle.
If you're interested, I'd love to talk more! I'm in NYC if you happen to be too, I love meeting people interested in live video :)
I'm not a Swift programmer, or Kotlin programmer, and haven't touched Java in a decade. I would love to collaborate with people on this project. I can offer lots of enthusiasm, and if you like programming without hassle I'll handle all the user-facing stuff and managing GitHub issues and support requests, project admin and writing documentation :)
There is much to do (and users really want video recording on iOS), but I've not got the coding skills to start. Picking up one language would be a nice challenge; keeping the iOS, Android and web versions at feature parity... I can't do it on my own.
At the moment we have a HR platform that one client is using and we're looking to build other stuff - we have some ideas in the making.
If anyone wants to join
This sounds (at least) tangentially related to your HR work, so I'd love to connect and share ideas. You can contact me: hello at marigoldapp dot-com
Email is in profile. Thanks!
Recently there was a post reminding people about the excellent "polling" feature on HN. Suddenly there were a bunch of polls with great discussions that followed. It would be helpful if this page (https://news.ycombinator.com/submit) included tips about how to create a poll.
My goals are:
- Restore the ability to use git repos as packages to the frontend ecosystem
- Simplify tooling needs by encouraging devs to check in a readable version of their product compiled down to standard javascript
- Provide the first interoperable way to package components, even those with styles
- Make it all Just Work as long as you run `macrome watch`
Always looking for others to join me or have a chat with.
Emails welcome, check my profile.
Email in profile :)
There are some nice and simple to do list apps (eg. Notally on Android) that have no syncing. Then there are some platform specific solutions (eg. the Apple's Notes app) that work well on platform.
I've been looking for something just to share a shopping list with my SO, and all I could find were elaborate SaaS productivity suites with projects and deadlines and calendars and project management features, but nothing just for syncing a single to do list between an iPhone and an Android device.
(I've been trying to use as few Google services as possible, but it seems very hard on Android.)
If your use case is really just groceries there is even an app just for sharing shopping lists with others (some friends use it and like it, personally for me it felt overkill): https://www.getbring.com/en/home
Todoist sounds more like a "Get Things Done" style task tracker (like OmniFocus). Due dates and assignments and priorities are way beyond what I want a todo list to do.
As for Bring, yeah that looks like it's way overengineered.
Oh yes. Much to the frustration of even the phone OEMs. Hopefully WiFi is easier to handle than 5G lol
I'm personally limited on cycles currently (N.B. I am a cofounder / direct of eng. of an ISP that builds custom 802.11ax-over-mmWave last mile access networks and CPEs in-house), but since I have enough experience in that world I can probably jump in at points to help as I free up.
This means something as simple as changing the network name or password requires changing it on every single access point manually, and even worse if your mesh system relies on sharing frequently-changing state between devices.
OpenWISP tries to address this problem: https://openwisp.org - I suggest you check it out and solve the configuration management problem first.
The actual "mesh" part is actually relatively easy. Most commercial systems use basic Linux networking tools, HostAPd (sometimes with custom improvements, but this all ends up upstreamed or reimplemented upstream given enough time) and custom glue code to tie them together. A "mesh" system is typically a user-facing network being broadcast by all APs (with shared settings such as name and password) and an invisible, "backhaul" network each AP hosts (either on a separate interface or on the same interface as the AP - I believe some wireless cards can act both as AP and station as long as the channel is the same) and the other in the path connects to, and the glue code handles configuring all of that. 802.11s is also an option that can be used, and I'm pretty sure all of this is already possible to configure manually in Linux - what's lacking is the "glue code" to set up & manage all of this automatically.
Another mesh community is Thread, which seems to have Zig-bee, Apple & Google involved https://en.m.wikipedia.org/wiki/Thread_(network_protocol)
I haven’t looked into mesh networks in a few years. Back then, Germany seemed to have a few open source mesh projects.
Seeking other pilot-programmer-contributors: [redacted]
https://awayto.dev -- Check out the video
https://github.com/keybittech/awayto
If you like making tools for developers, contractors, and the business world. Come check us out and join the discord!
I'll ask my wife who is a BCBA to share any studies related. The premise is symbols to help a kid reach from using symbols to letters/words. Right now they print a ton and velcro them to a board and communicate a complicated topic. There are apps out there but they kind of are lacking IMO and hard for a kid to figure out.
Also what kind of art do you make? Im an artist working as a dev as well. I know we're out there but we're a bit rarer than I was expecting.
As for art I love watercolor, gouache, ink and brush. I trained professionally to be an animator so I love cartoons.
Have thought a lot about this space, too, and identified similar needs. Please connect via email in profile.
I contribute to Optikey and was involved in OpenVoiceFactory in its first incarnation. Optikey is primarily QWERTY based but does supported the Communikate pagesets - more general OBF support would be a welcome PR! Coughdrop is probably a better fit for your needs, and is open source so free to self-host, though they do offer hosted plans for $.
https://github.com/CoughDrop/coughdrop
Disclaimer: yes, I'm affiliated :)
Make sure to do a Show HN when you get it far enough!!!
I noticed that a very simplistic LSTM neural network model is able to learn very quickly all the rules of pronunciation ex:
- grapheme "EA" in “please” or "heat" is pronounced one way
- but grapheme "EA" in « death » or " bread" is pronounced in a different way
- but grapheme "EA" in "great", "steak", "break" is also pronounced in a different way In fact the neural network is learning the rules and able to guess the pronunciation of word it have never seen before.
This made me try to see
1- what is the minimal number of example the Model need to be trained with to learn all the rules.
2- What is the optimal sequence that rules must be learned ... This can all be discovered and measured easily and accurately by training the model with different training set. I believe this have never been done before, because experimenting with real kid is too slow.
We make the assumption that is something is easy to learn by an LSTM model it will also be easy to learn by a human. (turn out to be true)
There is a lot of design decision that need to be made about how to visually display "grapheme" and let the kid interact with them.
This is just a side project for now but I am in strong need of a UI programmer to partner with!
But in another way this could be a kind of overlay of information. Maybe it shows how words are split, maybe it's to disambiguate different pronunciations of "ea", ... I'm not sure what's the most important added information. But overlaid on the words it might be a kind of scaffolding that makes it easier for kids to successfully read words and gain the practice that makes it possible to remove that scaffolding.
There's a lot to think about there when it comes to reading (not to mention a ton of pedagogical knowledge)... I'm not sure I have the space to be a solid collaborator, but I do find this stuff interesting to talk about sometime...
I have a 12 year old autistic son. He attends ABA therapy. I am also looking to create tools and apps that can help him communicate better. Also, he's very musical, so I am looking at better ways to help him learn piano (he loves Simply Piano but I want something that let's me define a custom lesson plan based on MIDI files of songs he loves).
"Relevant xkcd" is a meme that's been a part of online communities for as long as the comic has had online notoriety. [2] I wanted to build a bot to see how true that was. I've got a lot of the hard parts completed (data collection / curation, initial models), sitting in my TODO box for quite some time and would love to pair up with a collaborator to get it across the finish line.
The idea is an automatic xkcd recommendation bot that takes advantage of the latest and greatest in NLP advances (a fine-tuned hugging face model). [3] I've already got the training data (reddit comments that mention "relevant xkcd") and each individual xkcd's notes from the xkcd wiki. [4, 5]
Feel free to reach out to me via email (in my bio).
[1] https://github.com/adithyabsk/relevant_xkcd
[2] https://knowyourmeme.com/memes/subcultures/xkcd
[3] https://huggingface.co/models
[4] https://www.reddit.com/r/bigquery/comments/3cej2b/17_billion...
[5] https://www.explainxkcd.com/wiki/index.php/Main_Page
I’ve worked on a compiler for a HM type checked language for a while. Since September last year I’ve been stuck trying to implement tagged unions (such as Maybe). I’ve got all other features.
I’m looking for a tutor who would write a type checker with me in a screen sharing session. I’m located in Dalarna, Sweden atm, but can do any timezone that fits. I want to pay for this as I am really really curious to find out how it works! It’d probably be no more than around 100 LOC for just the type inferrer.
No payment necessary, but I'd be happy to do a videocall with you. I've implemented tagged unions once, albeit with a very different syntax: https://akkartik.github.io/mu1/html/033exclusive_container.c... (this is a Literate format that eventually tangles into C. The `scenario` blocks show what code in the language would look like. Though none of these is a good example. They all sidestep the type checker to some extent since they're creating raw objects in memory.)
Anyways, I'd be happy to chat with you :) I suspect it'll be a short call. There's a likely a single sentence somewhere that will get you unstuck.
Thanks for the offer! Sending you an email :-)
HM = Hindley-Milner
My email is in my profile please feel free to get in touch regarding type checkers!
Here is a type checker, lacking inferring, for tagged unions. What I need to figure out is how integrate this with the unification algo http://www.martinjosefsson.com/ocaml/compiler/interpreter/ty...
I'm not super tech focused at the moment either, but work as a product manager / ba for the past 7 years and building a surf community sounds like fun.
https://github.com/lukew3/congol
I could use someone to help out and keep me motivated.