Echo Music Version 1.4 is out now !
https://github.com/iad1tya/Echo-Music

https://redd.it/1nptitl
@r_opensource

Evaluating Apache Pulsar pros, cons, and license (my xp for data ingestion use case)

Background: I had been successfully using Postgres for the event streaming use case, scaled to 100k events/sec. It provides the best performance/cost ratio for our use case (collect customer events data from various apps/websites and route to hundreds of product/marketing/business tools api and warehouse), thanks to these optimizations. But it is a never-ending effort to continue optimizing as the product scales. By exploring alternate approaches, I wanted to avoid my blindspots. So I and my team started experimenting with Pulsar. I experimented with Apache Pulsar for ingesting data vs current solution - having dedicated Postgres databases per customer (note: one customer can have multiple Postgres databases, they would be all master nodes with no ability to share data which would need to be manually migrated each time a scaling operation happens).

Now that it's been quite some time using Pulsar, I feel that I can share some notes about my experience in replacing postgres-based streaming solutions with Pulsar and hopefully compare with your notes in order to learn from your opinions/insights.

### What I liked about Apache Pulsar:

No more single points of failure (data replicated across bookies): Data is replicated in at least two bookies now. This made us a lot more reliable when it comes to data loss.
Tenant isolation is pretty good, auto load balancing works well: We haven't experienced so far a chatty tenant affecting others. We use the same cluster to ingest the data of all our customers (per region, one in US, one in EU). MultiTenancy along with cluster auto-scaling allowed us to contain costs.
Maintenance is easier: No single master constraint anymore, this simplified a lot of the infra maintenance (imagine having to move a Postgres pod into a different EC2 node, it could lead to downtime).

### What I wished to be better:

StreamNative licensing costs were significant
Network costs considerably increased with multi-AZ + replication
Learning curve was steeper than expected, also it was more complex to debug

Would love to hear your experience with Pulsar or any other Open Source alternative. Please do share your opinions or insights on the approach/challenges for my use case.

P.S. I am a strong believer in keeping things simple, using the trusted and reliable tools over running after the most shiny tools. At the same time, I am open to actively experiment with new tools, evaluate them for my use case (with a strong focus on performance/cost). I hope this dialogue helps others in the community as a learning opportunity to evaluate Open Source technologies and licenses, feel free to ask me anything.

https://redd.it/1npvppn
@r_opensource

Open source project ESP32 Bus Pirate - A Hardware Hacking Tool That Speaks Every Protocol
https://github.com/geo-tp/ESP32-Bus-Pirate

https://redd.it/1npz7ac
@r_opensource

An in-depth look at TEN, an open-source framework for real-time voice AI that just hit v0.10.

I wanted to flag an open-source project that seems to be doing real-time voice AI the right way: the TEN framework. I spent some time digging into their new v0.10 release, and the engineering is genuinely impressive.

Most voice AI projects I've seen are a nightmare when it comes to latency. TEN seems built from the ground up specifically to fix that. Their new "main extension" is a really smart design, it gives you a clean entry point for your own logic without risking the stability of the core C++ pipeline.

The first-class Node.js support is also a huge deal. You can orchestrate the high-performance C++ and Python components from a familiar JS environment, which is a massive win for web developers.

The project is transparent and well-documented. Honestly, if you've ever struggled to build a voice agent that doesn't feel laggy, you should check this out. It’s a great piece of community-driven engineering solving a hard problem.

You can find the repo and docs here: `https://github.com/TEN-framework`

https://redd.it/1nq0z3t
@r_opensource

Swetrix v4 [OSS Google Analytics alternative] - new UI, OIDC, project sharing and more!
https://swetrix.com

https://redd.it/1nq375w
@r_opensource

Proxmox-GitOps – "Everything-as-Code" Container Automation
https://github.com/stevius10/Proxmox-GitOps

https://redd.it/1nq36mu
@r_opensource

Testlemon is now Open Source – API Test Automation Tool

Hello everyone!

I’m excited to share that after 1.5 years of development, testlemon is now Open Source. All code for the engine, Docker image, MCP server, and GitHub Actions is publicly available in our repos here: https://github.com/testlemon

The SaaS app will still be available for paid users, with a free trial here: https://app.testlemon.com/

Testlemon helps you automate API testing. It supports testing response status codes, response time, and body content without coding. You can also do test chaining, manage variables and secrets, and—recently added—automatically generate tests from an OpenAPI specification.

Generate tests from OpenAPI spec example:

docker run --rm itbusina/testlemon -c https://api.apis.guru/v2/openapi.yaml

Run tests from a test collection:

docker run --rm itbusina/testlemon -c "$(<collection.yaml)"

You can find full details about test collections, validators, and integrations in the documentation: https://docs.testlemon.com/

Give it a try and let me know what you think! Feedback is super welcome.

https://redd.it/1nq5oz6
@r_opensource

Interactive React app to design custom grid maps and visualize solutions using a pathfinding algorithm.
https://github.com/Azizham66/React-Maze-Solver

https://redd.it/1nq59sm
@r_opensource

An open-sourced, decentralized operating system, aka world computer.
https://anttp.antsnest.site/2bf0fe581569230260a9123a7853d1b42ecaece71ec13b06e582cb8af079c294/wasm-os/wasm-os/index.html

https://redd.it/1nq96tr
@r_opensource

Introducing Newsletter Support in Blogr - A Rust-powered Static Site Generator

I'm excited to share that **Blogr**, a open-source static site generator built in Rust, now includes comprehensive newsletter functionality.

Blogr is a fast, lightweight static site generator designed specifically for blogs. It offers Markdown-based content creation, a built-in terminal editor with live preview, and one-command deployment to GitHub Pages. You can see it in action at [https://blog.gokuls.in/](https://blog.gokuls.in/) which is built entirely with Blogr.

# Newsletter Features

**Subscriber Management**

* Email subnoscription collection via IMAP integration
* Interactive approval interface for managing subscriber requests
* Import/export from popular services (Mailchimp, ConvertKit, Substack, etc.,)
* REST API for external integrations

**Newsletter Creation**

* Automatically generate newsletters from your latest blog posts
* Preview before sending

**Reliable Delivery**

* SMTP integration with rate limiting
* Test email functionality
* Batch sending with progress tracking

# Key Commands

# Fetch new subscribers from your email inbox
blogr newsletter fetch-subscribers

# Launch approval UI to manage requests
blogr newsletter approve

# Send newsletter with your latest post
blogr newsletter send-latest

# Import existing subscribers
blogr newsletter import --source mailchimp subscribers.csv

# Start REST API server for integrations
blogr newsletter api-server --port 3001 --api-key secret

# Setup

Newsletter functionality integrates seamlessly with your existing Blogr blog. Simply enable it in your `blogr.toml` configuration with your IMAP/SMTP settings, and you're ready to start collecting subscribers.

The system works by monitoring a dedicated email address for subnoscription requests, providing an approval interface, and then sending newsletters using your SMTP configuration.

Check out the project at [https://github.com/bahdotsh/blogr](https://github.com/bahdotsh/blogr)

https://redd.it/1nq91la
@r_opensource

Who owns freshcode.club, freshfoss, etc?

From what little I've been able to dig up, whomever owns freshcode.club (and others) has been running this on his/her own dime and it's just been sitting there, slightly neglected for a while. It's now down, and there's no more information.

Does anyone know who is running (at least) freshcode.club? I'd like to help fix it, or at least get an idea of what's going on.

https://redd.it/1nqbrzy
@r_opensource

Apache Iceberg 1.10
https://opensource.googleblog.com/2025/09/apache-iceberg-110-maturing-the-v3-spec-the-rest-api-and-google-contributions.html

https://redd.it/1nqdj79
@r_opensource

Now more than ever, location sharing privacy is important.

Hey folks!

Our names are Chandler & Fatima and we've been working on an app called Grid (mygrid.app). We built it because we got tired of location sharing apps brazenly exploiting user location data (think Life360 and location sharing services selling user location data to data brokers, federal/gov agencies, etc.). We wanted a way to share location without having to compromise on our data privacy.

It's an open-source project that's fully self funded. Because it's meant to be a tool that helps the overall cause, we want to make sure it's the absolute best version it can be: the most useful, valuable and private version for users.

Here’s what Grid is:

Location sharing with end‑to‑end encryption (profile photos are also E2EE), using Matrix Synapse for the backend. Only people you choose to share with can see your location.
Self‑hosting options: you can run your own backend server and host your own map tiles. If you do this, you take on risk and maintenance.
Minimal data collected: phone number (for verification - we're working on alternatives/foregoing phone numbers altogether), username. No tracking, no location data stored in decrypted form by us.
Sharing features: 1:1 or with groups, shared durations/expiration, you control when to stop sharing.
Map tiles are by default Protomaps via Cloudflare; unless you self‑host, map tile fetching involves some metadata/logs by the map tile host (i.e. they can see what tiles were requested)
All core features will remain free. Cosmetic/nice to haves options will be paid (currently we have satellite maps) in order to continue to fund development and work on the project!
Points of Interest: Drop points on the map of locations that are of interest to your group (meet up points, restaurants, etc.)

Where Grid still has work to be done:

If you self‑host but mix with other Matrix use, there are warnings: Grid isn’t fully tested in federated settings. Could be bugs.
The phone number for verification: We're working to move away from this.
The map tiles’ privacy: Protomaps routed through cloudflare, some metadata/requests may leak. Looking into alternatives and offline maps.
UI, and edge case bugs need polish. It’s relatively smooth in performance, but not “mission‑critical proven” in every context. We're only a two-person team so our workload capacity is limited.

Here’s how people in the community are value added to the project:

Test it in real conditions and tell us where it fails.
Audit us. Grid isn’t built for the lowest common denominator but for security and privacy. Check our github out, help us identify where the gaps are so we can close them.
Ideas for improving self‑hosting security, map privacy, or making it usable on phones without Google services. We SO welcome contributions!

Let us know what you all think!!

https://redd.it/1nqdjw8
@r_opensource

Alien vs Predator Image Classification with ResNet50 | Complete Tutorial

I just published a complete step-by-step guide on building an Alien vs Predator image classifier using ResNet50 with TensorFlow.

ResNet50 is one of the most powerful architectures in deep learning, thanks to its residual connections that solve the vanishing gradient problem.

In this tutorial, I explain everything from scratch, with code breakdowns and visualizations so you can follow along.

 

Watch the video tutorial here : https://youtu.be/5SJAPmQy7xs

 

Read the full post here: https://eranfeit.net/alien-vs-predator-image-classification-with-resnet50-complete-tutorial/

 

Enjoy

Eran

 

\#Python #ImageClassification #tensorflow #ResNet50

https://redd.it/1nqgb25
@r_opensource

Audo download new video

Hello.

Is there any tool to auto download a new video when posted on a channel ?
I am looking for something that will simply check is there any new video, if yes, download the audio stream only in a specific folder.

Thabk you.

https://redd.it/1nqi67j
@r_opensource

Favorite semi-obscure Open Source Software?

What is your favorite semi-obscure Open Source Software (even if you aren't using it at the current moment)?

https://redd.it/1nqcpdj
@r_opensource

Microsoft offering support for ReactOS? Yes.
https://x.com/reactos/status/1970758174388187225

https://redd.it/1nqkr4h
@r_opensource

How to analyze Git patch diffs on OSS projects to detect vulnerable function/method that were fixed?

I'm trying to build a small project for a hackathon, The goal is to build a full fledged application that can statically detect if a vulnerable function/method was used in a project, as in any open source project or any java related library, this vulnerable method is sourced from a CVE.

So, to do this im populating vulnerable signatures of a few hundred CVEs which include orgname.library.vulnmethod, I will then use call graph(soot) to know if an application actually called this specific vulnerable method.

This process is just a lookup of vulnerable signatures, but the hard part is populating those vulnerable methods especially in Java related CVEs, I'm manually going to each CVE's fixing commit on GitHub, comparing the vulnerable version and fixed version to pinpoint the exact vulnerable method(function) that was patched. You may ask that I already got the answer to my question, but sadly no.

A single OSS like Hadoop has over 300+ commits, 700+ files changed between a vulnerable version and a patched version, I cannot go over each commit to analyze, the goal is to find out which vulnerable method triggered that specific CVE in a vulnerable version by looking at patch diffs from GitHub.

My brain is just foggy and spinning like a screw at this point, any help or any suggestion to effectively look vulnerable methods that were fixed on a commit, is greatly appreciated and can help me win the hackathon, thank you for your time.

https://redd.it/1nqblp2
@r_opensource

Browser automation using recorded user sessions

Hey Everyone! I created a tool that can record user sessions on a website and will convert them into playwright browser actions. The initial idea was to use this for QA, but I thought maybe this could be helpful for other browser automation use cases as well. Here's how it works:

1. Developer can add our js snippet to their html
2. It records clicks, fills and selects. This can be extended to more actions
3. User can generate automation workflows by leveraging the user sessions recorded and use AI to determine how to complete a task

Here's a video of how it works and my website. If people are interested, I can open-source the recorder -> playwright automation!

https://www.loom.com/share/caa295aa921f4e71bb10e0448838a404?sid=ce02e0d5-61b7-4ba9-b635-8bc5bbdcc70c

https://redd.it/1nqnldl
@r_opensource

I rewrote Minecraft Pre-Classic versions in plain C

Hey folks, I’ve just finished working on a project to rewrite Minecraft pre-classic versions in plain C

Rendering: OpenGL (GL2 fixed pipeline)
Input/Window: GLFW + GLEW
Assets: original pre-classic resources
No C++/Java — everything is straight C (with some zlib for save files).

Repo here if you want to check it out or play around:
github.com/degradka/mc-preclassic-c

https://redd.it/1nqpc9l
@r_opensource

An opensource math competition website (still a WIP, but has functionality)

Hello,

I hope this is an OK place to post this: I am working on an opensource math competition site. It is called https://conjecscore.org/ and it currently only has 1 problem. But even though there is only one problem, the problem does not have a known solution. It's an open math problem. Instead of having a known solution it has a "score" function that determines how "close" you are to solving the problem (informally). It still has a lot of rough edges but I was wondering if people were frankly even interested in the idea. If so, I could try finding more open problems and give them a score function too. Additionally, I could continue polishing the site too. Lastly, and most importantly, the source code is here: https://github.com/thyrgle/conjecscore

Thank you for your time.

https://redd.it/1nqowou
@r_opensource