Hands-on experience with 🛠️ Dev tools · 🤖 AI agents · 🐧 Linux. Turning 2am projects into guides you can actually follow
Joined August 2011
- Tweets880
- Following106
- Followers61
- Likes121
botmonster retweeted
Quote: "an independent security researcher reported that the keys of eight eMudhra Global TLS intermediate Certificate Authorities were cryptographically weak".
Well, yeah, they were weak. A private key should look something like this:
84ac06790886c80c48d31691fd4c0fe9
ce3207f9807f2fe571d6e4ec426d7864
4298fc53df2b92566970e757ac998c34
Instead, their keys looked like this:
00000000000000000000000000000000
00000000000000000000000000000000
00000000000000000000000000000004
Global TLS Intermediate CA incident: We’ve updated our incident page with how it happened, who is affected (one US customer ecosystem), what’s not affected, and what we’re putting in place. We will publish the root cause once our investigation is complete.
👉emudhra.com/en/post/certific…
Readers added context they thought people might want to know
eMudhra minimizes the scope by claiming only "one US customer ecosystem" is affected. In reality, 8 of its CA certificates had catastrophically weak private keys (one being simply the number 4), allowing anyone to forge trusted HTTPS certificates for any website on the internet.blog.google/security/prote…
Wi-Fi presence fires your "goodbye" scene while you're on the couch, because iOS drops Wi-Fi when the screen sleeps. HA's bayesian sensor fuses Wi-Fi, GPS, motion, and power into one probability instead. How do your lights know you're home?
A $5 LD2410 mmWave sensor sees breathing, confirming a still sitter PIR misses.
Power draw above 300W is a surprisingly strong occupancy vote.
One missed signal drops the score, doesn't crash it to zero.
botmonster.com/smart-home/be… #HomeAssistant
botmonster retweeted
The Internet goes on high alert after discovery of 8 compromised TLS certs arstechnica.com/security/202…
Matter-over-Thread is not a Zigbee2MQTT replacement, and treating it as one costs you. Zigbee2MQTT still backs 5,000+ devices vs Matter's thin slice of bulbs and plugs. Run both in 2026. Which protocol runs your sensors?
Zigbee wins battery life: years on CR2032; Thread sensors lean mains-powered.
Thread/Matter latency lands 30-80 ms vs Zigbee's 50-150 ms.
Matter gear runs 20-40% pricier per device right now.
botmonster.com/smart-home/sh… #Matter
A 2012 MacBook Pro with a $50 SATA SSD drops boot from over a minute to under 15 seconds and app launches from 5-15s to 1-3s. The CPU was never the problem, the spinning drive was. Linux Mint XFCE handles the rest. What old laptop are you leaving in a drawer?
The 2012 unibody is the sweet spot: swappable RAM to 16GB and a 2.5-inch bay.
Skip the 2016-2019 Touch Bar models; the T2 chip fights Linux.
auto-cpufreq can add 30-50% battery vs the performance governor.
botmonster.com/self-hosting/…
Basal 2.2 checks a decision model's confidence scores against your own data. On new data, the setting that promised 1% wrong answers was wrong 8% of the time. After the retune, those cases go to a person instead.
Runs on your computer, and your data never leaves it.
No retraining, it only adjusts the confidence scores.
Its 31B model scores best on JevBench among open models at its price, but two small 4B models win once speed and cost count.
botmonster.com/ai/basal-2-2-…
Sonarr never downloads a byte. It decides which episodes you are missing, hands the transfer to something else, then does the boring part: renaming the file and filing it.
botmonster.com/self-hosting/…
CSS container queries are past 96% support, so a component can size itself by its parent's width. Those ResizeObserver scripts can go.
botmonster.com/web-dev/conta…
RimWorld 1.7 ditched Unity for an engine built on BRUTAL, a free C# framework. A 1,500-mod pack now loads in 6 minutes instead of 14, and most mods still work. Leaving the big engine was the speed upgrade.
It took Ludeon about five months from first call to announcement.
BRUTAL ships no editor and no renderer, so you build those yourself.
It's free for commercial games, but your game must show its logo before play.
botmonster.com/coding/brutal…
Jenkins ships a release most weeks, sits at version 2.575, and carries about 3,610 open issues at the same time. Almost nobody picks it for new work, and it still runs plenty.
botmonster.com/coding/jenkin…
AI agents built SolidCraft, a free Fusion 360 rival, in 10 days. It runs on Linux, FreeBSD and in a browser. Autodesk never shipped Fusion for Linux.
botmonster.com/ai/solidcraft…
"Hallucination" is a junk label that makes you fix the wrong thing. A faithfulness failure, where the model contradicts a retrieved chunk, needs a different cure than format drift. Chain-of-Thought alone cut errors 20-40%. How do you measure your hallucination rate?
Bad retrieval can push hallucination above a model with no RAG at all.
Gate CI on a Ragas faithfulness score; below 0.8 means weak grounding.
Best-of-5 costs 5x; two-call self-verify adds only ~1.5-2x.
botmonster.com/ai/how-to-fix…
