$ ch rg -Mt
rg - recursively search the current directory for lines matching a pattern
-M NUM, --max-columns=NUM
When given, ripgrep will omit lines longer than this limit in bytes.
Instead of printing long lines, only the number of matches in that
line is printed.
-t TYPE, --type=TYPE
This flag limits ripgrep to searching files matching TYPE. Multiple
-t/--type flags may be provided.
Also, developed another app to reduce my Twitter usage. The reason why I created this one is that even though there a lot of Safari Extensions out there, they can access to your Safari history but the app that I've developed can't because it is just a content blocker: https://apps.apple.com/us/app/not-for-me/id6451205605
For retirement calculators monte carlo simulations seem like the best way to model risk, but all existing implementations are several orders of magnitude slower than they should be. I made https://james.darpinian.com/money/ to prove that it's possible to simulate a million scenarios in a few milliseconds. You can drag sliders and see simulation results update in real time. Even on phones.
Because I made it for my own use, the UI is terrible. The simulation result you need to pay attention to is the % success at the bottom, which changes as you drag the sliders. The simulation assumptions are reasonable and sophisticated (more so than many existing FIRE tools, especially the use of actuarial tables) but not documented except in the code comments. It would be cool if one of the real FIRE calculators out there with a real UI would implement a fast monte carlo simulator like this. I'd be happy to share how it works.
I built this web + mobile app (PWA) written in React Native + TypeScript that does simple revenue forecasting for a SaaS that uses Stripe.
Stripe's mobile app and others kinda do this already but some of their numbers can be inaccurate (as detailed in the repo's readme) so that made me open-source + solve an issue my own SaaS[1] has with Stripe.
A tracker thing for my son's in their first year. tracks sleep, food, awake, poos. easy hand off between me and their mum so we know when it's time for a nap or feed etc. got a cool radial data Viz where easy ring is 24h and the rings stack up. let's you see sleep patterns really easily. sorry no demo or screens as the data is all private and there's no security on there!
I needed a weekly report for my contracting clients to prove their web host sucked to the point where it was costing them significant money.
They were paying for cheapest tier WordPress hosting at the time, and didn't believe me when I said random 5 min blocks of downtime throughout the day were adding up.
I built a dirt-simple form that takes a URL and sends a notification when the site goes down/up, with a weekly summary email.
Then, I kept adding features every day, 2 hours at a time, even after I stopped being a contractor.
I recently wrote a simple Linux "app" to render the output from a common USB infrared camera designed for android phones on a big screen using a raspberry pi zero: https://github.com/jcalvinowens/ircam-viewer
I like the idea! Sad I almost left macOS (to Linux), so I would love to see something similar for Linux. Maybe you researched this before and know some similar alternatives? Could be anything, even a terminal app.
I was annoyed at how parasitic recruiters started contacting me about jobs in crypto, so i spent the last year building a fully automated job board that ingests all job data automatically and structures it using openai, while also enriching it with relevant company financials to combat information asymmetry.
I’m now using it to find my next job :)
nice, very solid! I've just started picking away at my own version of this (mainly to procrastinate from job searching), although not focused on crypto.
From a quick glance at the TG group, are you using just Github to scrape for jobs?
Thought about doing this as I'm fed up with Linkedin and the other crap out there. Would also love to know how this is done (where to get the job data).
I've also started down the road of rolling my own personalized job board. I'm also curious about where you're sourcing data from. I resorted to scraping sites like Crunchbase, layoffs.fyi, etc.
Related to this I put together a script to use ChatGPT to generate cover letters. Given the name of a company and the job title, it generates a fancy cover letter using Latex.
I created a note-taking app which works in every* browser, with and without JS and CSS, has bookmarklets for clipping, PKI-backed user accounts, completely portable data format, threaded conversations, completely auditable data structure, labeling, queryable with SQLite, web of trust for the user identities, and support for adding and running code.
* Mosaic has issues, and I have not tested on many pre-Mosaic browsers. I've never tested with WorldWideWeb. I've done extensive testing with Netscape and IE. Older versions of Chrome are extremely flaky for some reason, and often don't even run. Most old and retro browsers work, however, including sessions, posting, voting, etc.
Yes, the first browser was WorldWideWeb, written by TBL himself. Then, there was a small ecosystem of other browsers before Mosaic came into the picture. Mosaic was the first browser with a "must-have feature" which allowed it to become very popular compared to the others. This must-have feature was inline images, using the <img> tag. Before this, if you wanted to include an image to a page, you linked to the file with an <a> tag. After Mosaic became popular, the team who developed it started Netscape, which was unofficially called Mozilla, i.e. Mosaic K*lla. This was sort of the "zeroth" browser war. Netscape's must-have feature at the time was frames using the <frameset> and <frame> tag. After that, came IE and the "first" browser war, between IE and Netscape...
(I wasn't there, but I've done a lot of research into this.)
decision matrix web app. features
* automatic weighting & scoring
* draggable criteria/choices interface
tag-based file manager CLI app
* keeps all files/directories in a "~/t/multi./tag./structure./"
* directory structure is kept in a hierarchical order based on tag frequencies - most common tags are always at the top level directory.
GDO: Garage door Opener, a passkey enabled web interface to trigger opening my garage door. It's basically just a 2 button website, 1 button opens and closes the door with a delay, the other just opens/closes the door.
It uses a passkey auth proxy which allows me to toggle user access.
in theory some publications are available as torrents, in theory they are in batches zipped together and in theory one could, by requesting certain parts of the zip file using a crafted torrent client, download only the publication they are looking for, in theory
I got fed up with using loads of different apps to keep track of my flights while doing long term traveling trips so made a simple site to track them using the AirLabs free API.
792 comments
[ 3.7 ms ] story [ 480 ms ] threadDistributed test automation for the browser.
A WebSocket client and server.
A code beautifier for a great many languages.
A diff tool.
Some accessibility tools.
For example:
---See also: "Ask HN: Most interesting tech you built for just yourself?" https://news.ycombinator.com/item?id=35729232
Also, developed another app to reduce my Twitter usage. The reason why I created this one is that even though there a lot of Safari Extensions out there, they can access to your Safari history but the app that I've developed can't because it is just a content blocker: https://apps.apple.com/us/app/not-for-me/id6451205605
Because I made it for my own use, the UI is terrible. The simulation result you need to pay attention to is the % success at the bottom, which changes as you drag the sliders. The simulation assumptions are reasonable and sophisticated (more so than many existing FIRE tools, especially the use of actuarial tables) but not documented except in the code comments. It would be cool if one of the real FIRE calculators out there with a real UI would implement a fast monte carlo simulator like this. I'd be happy to share how it works.
I built this web + mobile app (PWA) written in React Native + TypeScript that does simple revenue forecasting for a SaaS that uses Stripe.
Stripe's mobile app and others kinda do this already but some of their numbers can be inaccurate (as detailed in the repo's readme) so that made me open-source + solve an issue my own SaaS[1] has with Stripe.
[1] https://Last10K.com
They were paying for cheapest tier WordPress hosting at the time, and didn't believe me when I said random 5 min blocks of downtime throughout the day were adding up.
I built a dirt-simple form that takes a URL and sends a notification when the site goes down/up, with a weekly summary email.
Then, I kept adding features every day, 2 hours at a time, even after I stopped being a contractor.
That app was OnlineOrNot (https://onlineornot.com).
Since then I also added status pages, like https://hackernews.onlineornot.com/
As well as cron job monitoring, to ensure database backups and whatnot run when I expect.
1. https://github.com/hamon-in/invoice/ was a command line invoicing program that I wrote and used for 2 years before moving to something SasS based.
2. https://github.com/nibrahim/Calligraphic-Rulings is a command line (and later web based - http://calligraffiti.in/rulings) tool I wrote and use regularly while to practise calligraphy
3. https://github.com/nibrahim/Hyde And emacs mode to manage Jekyll/Octopress blogs which I use for my personal site
A bunch of smaller scripts for daily work (e.g. mini pomodoro timer, Emacs scripts to manage client conversations etc.)
https://github.com/Bogdanp/remember
I’ve been using it every single day since I wrote it.
Check it out at: https://jobstash.xyz
From a quick glance at the TG group, are you using just Github to scrape for jobs?
Related to this I put together a script to use ChatGPT to generate cover letters. Given the name of a company and the job title, it generates a fancy cover letter using Latex.
* Mosaic has issues, and I have not tested on many pre-Mosaic browsers. I've never tested with WorldWideWeb. I've done extensive testing with Netscape and IE. Older versions of Chrome are extremely flaky for some reason, and often don't even run. Most old and retro browsers work, however, including sessions, posting, voting, etc.
Just to confirm: you are referring to this link in your profile, correct?
https://github.com/gulkily/pollyanna
Edit: glad to see w3m in the supported browser list!
(I wasn't there, but I've done a lot of research into this.)
tag-based file manager CLI app * keeps all files/directories in a "~/t/multi./tag./structure./" * directory structure is kept in a hierarchical order based on tag frequencies - most common tags are always at the top level directory.
I created it originally for myself. By now it has a few hundred thousand users.
https://sr.ht/~mrms/mmre
A tool which allows me to embed shell scripts into text files. I use it to generate my website.
https://git.sr.ht/~mrms/mmtt
https://developer.mozilla.org/en-US/docs/Web/API/Web_Authent...
My guess is it's some JavaScript on the website, it would trigger the browser to ask the OS to prompt the user to provide their credentials.
It has a lot of niche features, but my favorite is the ability to preview or jump to references even when they are not linked in the PDF file.
GitHub: https://github.com/PaulSinghDev/personal-flight-tracker
Public site: https://personal-flight-tracker.vercel.app/no-key
Software: Swift library to read and modify ZIP archives: https://peakstep.com/claquette/ Screen recording/video/GIF creation app for macOS: https://github.com/weichsel/ZIPFoundation A small CMS to maintain my websites (unreleased)
Hardware: A Time Machine compatible NAS based on a Raspberry PI (PoE powered, 3D printed mountable enclosure)