Ask HN: What software should I rewrite as open source?

I'll probably rewrite it in Rust, but I would like some ideas on software to rewrite and make open source.




So many people wishing for VB6. While that wish makes sense, you should give Gambas [1] a try first. It may already have what you seek - or may be a trivial addition.



Basic autonomous drone software a la Ardupilot or INAV. Look at replicating ardupilot's auto tuning for new vehicles to tune in flight characteristics. And add an API to send flight instructions to the flight controller from something like a Jetson.

You could take the best from the various drone software packages (OSS and proprietary) and make something really special that could power the next generation of autonomous vehicles that rely heavily on local AI.


Google Docs, Airtable, Apple Notes, GPT Playground, a mail client, reddit, discord


> Google Docs

and Overleaf

> discord

and Teams/Slack


For Discord, matrix could already be a good alternative.


visual basic but using react/react native or similar modern frameworks. Enable developers to drag and drop components but can also right click to edit the code seamlessly


Framer ( website builder may be good inspiration here.

They have a decent model for visual editing and letting you drop down to React.


Paint Shop Pro. iDraw. Evernote. Hashicorp Vault.

Best of luck!


Visual Basic - language + modern IDE implementation.


Have you tired Gambas ? I used a lot of VB6 in the old times (and not so old). I didn't try iGambas, but it looks very similar.


language + modern IDE but using react /react native


This is what I came to say. So glad to see it here. VB6 is RAD.


This is actually my answer too, VB6 was the epitome of GUI dev.

I'm actually starting to play around with Lazarus to see if it's a reasonable replacement.




> I'll probably rewrite it in Rust, but I would like some ideas on software to rewrite and make open source.

The Linux kernel. ;-)


1+ on this


Weird question. Rewrite whatever you feel like writing. You have to have an internal drive rather than something derived from outside of yourself.

You aren't being paid for your work, so there has to be some other incentive. Whether that is just some mental challenge, or the acclamation of your peers.


It’s more to see if I can, and maybe to learn something from it. Also I am very bad at generating ideas.


I think it's a good question. By trying to find something that others want, OP is more likely to find actual users and feedback.


I think that the biggest thing missing is an open/libre mrc.bin implementation that is written from scratch (as a clean room implementation).

Coreboot to this date still relies on the Google Chromebook's dump from more than a decade ago.

And it's probably the hardest reverse engineering challenge, because Intel isn't offering any details.


Would love to see ride hailing apps in open source, since most of them are losing lots of money and may shut down.


This is a good wish. It's going to be challenging without a central authority to be fair to both riders and drivers.


Not all are fair. Grab acquired Uber's operations in SE Asia.

They lied on the app about how long the cars would arrive, so people opted for Grab when booking on both platforms. They'd offer bonuses to drivers that were impossible to obtain (e.g. drive to a location without road access). They'd offer X in bonuses then pay out 0.7X, taking a cut from the promised bonuses and from tips. Uber would loan money to drivers and then let the drivers they banned keep the loan. Grab would withhold payment to drivers, ban drivers who got less than say, 4.5 rating, then not pay out the drivers they banned.

All fair game, all legal. The rough play helped them win, or so they think.

I've had requests from 3 people who wanted to build a ride hailing app. It's not worth the time and I'd discourage anyone who thinks they could stand up to Grab with just implementation ability. But if there was a code base, I'm sure some would be happy to take the mantle.


That's really about the business model and not about the tech being OS, no?


It's that and more. The tech is a moat. The business model is another moat.

Part of the problem is the business model allows for some shady practices like selling $1 services for $0.80 to kill competitors (like the taxi industry and other startups in the same space). Then raising prices to $1.30 over 20 years.

I don't think it works though. It just ends up with everyone dead and us buying enshittified taxis at 2-3x the price later because nobody else has a ride sharing app.




yes. Unity but with Python ,typescript or a lisp like language


Isn't godot already an replacement?


Godot needs a whole lot of work to be a visble alternative even for a mid-sized team.


> Godot needs a whole lot of work (..)

There you go!


go help with freecad or something - we need as much "the blender of drafting" as possible :)


Serial comms tool with buttons to send pre-defined commands. One of the only tools I miss from windows.


Thats a really interesting product. In my past job I used a lot of CAN analysis tools and most of them were windows only. We ran win VMs on our dev laptops to do real time debugging. Germans really dominate in this space


I miss FoxPro; I learned it while studying CS and created applications that could empower mom-and-pop stores. It could manage their inventory, handle invoicing, and generate basic sales reports for them. FoxPro was simple yet remarkably powerful.


MATLAB add-ons or toolkits, any of them depending on your interest. As easy option would be to adapt them to Octave .




Substance Painter. Blender's texture painting capabilities are rudimentary at best.




NDepdend please, it is hard to make myself Roslyn working code to do most exiting architecture tasks every time.