#green_software — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #green_software, aggregated by home.social.
-
🎙️ SL5 Aura v0.43 is out – The Navigation, Performance & NixOS release!
100% offline, privacy-first voice assistant framework. No cloud. No data leaves your machine.
✨ What's new:
• RAM-based recognition history (faster, less disk I/O)
• Experimental NixOS support
• Advanced FZF navigation shortcuts
• Regex-powered self-tests
• ⚡ Full CI/CD via GitHub Actions — 39 core logic tests run automatically on every push, at just ~2.87 J per test (39 tests @ 0.34s avg · measured with [Eco-CI](https://metrics.green-coding.io/ci.html?repo=sl5net/SL5-aura-service&branch=master&workflow=261851628)) · no cloud compute👉 https://github.com/sl5net/SL5-aura-service/releases/tag/v0.43.00
👉 https://codeberg.org/seeh/SL5-aura-service
#OpenSource #VoiceAssistant #Privacy #NixOS #GreenSoftware #green_software
-
🎙️ SL5 Aura v0.43 is out – The Navigation, Performance & NixOS release!
100% offline, privacy-first voice assistant framework. No cloud. No data leaves your machine.
✨ What's new:
• RAM-based recognition history (faster, less disk I/O)
• Experimental NixOS support
• Advanced FZF navigation shortcuts
• Regex-powered self-tests
• ⚡ Full CI/CD via GitHub Actions — 39 core logic tests run automatically on every push, at just ~0.08 J per test (39 tests, 0.49s per test · measured with [Eco-CI](https://metrics.green-coding.io/ci.html?repo=sl5net/SL5-aura-service&branch=master&workflow=261851628)) · no cloud compute👉 https://github.com/sl5net/SL5-aura-service/releases/tag/v0.43.00
👉 https://codeberg.org/seeh/SL5-aura-service
#OpenSource #VoiceAssistant #Privacy #NixOS #GreenSoftware #green_software
-
🎙️ SL5 Aura v0.43 is out – The Navigation, Performance & NixOS release!
100% offline, privacy-first voice assistant framework. No cloud. No data leaves your machine.
✨ What's new:
• RAM-based recognition history (faster, less disk I/O)
• Experimental NixOS support
• Advanced FZF navigation shortcuts
• Regex-powered self-tests
• ⚡ Full CI/CD via GitHub Actions — 39 core logic tests run automatically on every push, at just ~2.87 J per test (39 tests @ 0.34s avg · measured with [Eco-CI](https://metrics.green-coding.io/ci.html?repo=sl5net/SL5-aura-service&branch=master&workflow=261851628)) · no cloud compute👉 https://github.com/sl5net/SL5-aura-service/releases/tag/v0.43.00
👉 https://codeberg.org/seeh/SL5-aura-service
#OpenSource #VoiceAssistant #Privacy #NixOS #GreenSoftware #green_software
-
🎙️ SL5 Aura v0.43 is out – The Navigation, Performance & NixOS release!
100% offline, privacy-first voice assistant framework. No cloud. No data leaves your machine.
✨ What's new:
• RAM-based recognition history (faster, less disk I/O)
• Experimental NixOS support
• Advanced FZF navigation shortcuts
• Regex-powered self-tests
• ⚡ Full CI/CD via GitHub Actions — 39 core logic tests run automatically on every push, at just ~0.08 J per test (39 tests, 0.49s per test · measured with [Eco-CI](https://metrics.green-coding.io/ci.html?repo=sl5net/SL5-aura-service&branch=master&workflow=261851628)) · no cloud compute👉 https://github.com/sl5net/SL5-aura-service/releases/tag/v0.43.00
👉 https://codeberg.org/seeh/SL5-aura-service
#OpenSource #VoiceAssistant #Privacy #NixOS #GreenSoftware #green_software
-
Uber Migrates 1T Records from DynamoDB to LedgerStore to Save $6M Annually
https://www.infoq.com/news/2024/05/uber-dynamodb-ledgerstore/
#ycombinator #uber_dynamodb_ledgerstore #Architecture_Design #DevOps #ML_Data_Engineering #Green_Software #Dynamo_DB #Big_Data #NoSQL #Database #AWS #Cost_Optimization #Cost_Savings -
Uber Migrates 1T Records from DynamoDB to LedgerStore to Save $6M Annually
https://www.infoq.com/news/2024/05/uber-dynamodb-ledgerstore/
#ycombinator #uber_dynamodb_ledgerstore #Architecture_Design #DevOps #ML_Data_Engineering #Green_Software #Dynamo_DB #Big_Data #NoSQL #Database #AWS #Cost_Optimization #Cost_Savings -
Uber Migrates 1T Records from DynamoDB to LedgerStore to Save $6M Annually
https://www.infoq.com/news/2024/05/uber-dynamodb-ledgerstore/
#ycombinator #uber_dynamodb_ledgerstore #Architecture_Design #DevOps #ML_Data_Engineering #Green_Software #Dynamo_DB #Big_Data #NoSQL #Database #AWS #Cost_Optimization #Cost_Savings -
Uber Migrates 1T Records from DynamoDB to LedgerStore to Save $6M Annually
https://www.infoq.com/news/2024/05/uber-dynamodb-ledgerstore/
#ycombinator #uber_dynamodb_ledgerstore #Architecture_Design #DevOps #ML_Data_Engineering #Green_Software #Dynamo_DB #Big_Data #NoSQL #Database #AWS #Cost_Optimization #Cost_Savings -
What Is Green Software and Why Do We Need It?
https://spectrum.ieee.org/green-software
#ycombinator #app #climate_tech #green_software #software_engineering #website -
An 8-year-old Laptop and Go
It seems like a law of nature that software evolves to suck up all the reserves of the most current hardware. As a result, we have to buy new hardware every two to three years just to keep the software running at a bearable speed.
Now I have the opportunity to prove that old hardware can still be used productively.
Recently, I bought a dirt cheap, used Macbook Air from 2015.
4G of RAM, 128G of SSD.
And some old Intel CPU with two physical, hyperthreaded cores.These are definitely no specs to brag about.
To get a speedy OS and GUI, I decided to install Manjaro Linux with the Sway window manager. I heard that Manjaro works well on old Macbooks. And indeed, it works like a charm. (But I had to tick the "proprietary drivers" checkbox when installing Manjaro to make WiFi work.)
Now the interesting part:
Go 1.21 compiles itself (that is, the compiler, the whole toolchain, and the complete standard library, without running tests) in 2:33 minutes. Not bad!
Go seems ideal for developing code on low-end machines, if neither OS nor sluggish GUI tools don't get in the way.
This test wasn't just for killing time.
In the face of climate change and environmental pollution everywhere, circular economy becomes a crucial factor in shaping a future that is worthwhile for everyone to live in.
This includes using hardware for as long as possible.
Don't dump, reuse.
(Maybe I'll run the same test on a Raspberry Zero. How low-end can hardware be and still be a decent platform for developing and running software?)