Justin White

If there’s any low level coders in my network please check out Troy’s GitHub and his project to build a new, faster shell! Some details on Troy: ### Windows Accessibility and TTS Contributions (Red Eagle Team) - Description: As part of the Red Eagle Team (a Georgia-backed initiative with Microsoft support in the mid-2000s), Mallory developed foundational code for universal Text-to-Speech (TTS) systems and cross-platform handicap accessibility features. This included compatibility layers for screen readers, app navigation, and voice input (e.g., influencing tools like Dragon NaturallySpeaking). - Impact on Windows: Microsoft adopted and evolved his logic starting with Windows Vista, integrating it into core OS features for impaired users. If you've used modern Windows TTS, clipboard integrations, or basic accessibility navigation, elements trace back to this work. - Modern Extensions: He's currently exploring IoT/smart home integrations for disabled veterans, building on this foundation. Troy’s Post “Okay guys think tank. I maybe over complicating this or may habe to do a bit of tweaking with some C code. Rubian has been going smoothly. The new boot and dynamic system discovery deamons work great. As standalone. The new stack needs tweaking for efficiency but works and is stable. Here's the stack im working on with the prototype test. /etc/rc.local ->calls boot.rb as bg process. ->boot.rb bootstraps rubian's system discovery and array logic. User login -> rubian3.rb (repl) -> chsh -s ./rubian3.rb USER Rubian boot.rb logic. ->starts C ruby and Jruby C ruby starts the bootstrap process while jruby warms up. -> Cruby passes workload to jruby when its ready to finish boot system. The deamon system is the server-side, where shells and env are the client. (See my UNIX socket prototype here) https://lnkd.in/gf6rWs46 C and j ruby can now natively talk to each other. I know that works. The issue im having is the new login shell isnt hooking into the deamon when it loads. https://lnkd.in/g5_2ZuPH A further explanation: “Yes. I've been building from the Shell working down to the kernel. The Shell side of things mimic Bash but in ruby expression. On the logic side of things, ruby treats everything as an object. What im doing is taking all the system information and building Arrays. Each directory and file then has an index that we can iterate through with .each_with_index, .each_with_object, etc. So instead of a static file tree, i have arrays of nested file paths and directories being pattern matched. Any ruby app will be ran and communicate to it. This allows for dynamic file utils systems and more. Each array index points to a file, directory, process, thread, interpreter, shell, hardware, everything that is a defined method object.”

3

I've been building in Cursor all week, and loving every second of it, while doing so I began pondering the new advancements as they came. A Google engineer just replicated a year of her team's work in 60 minutes. Not a script. A distributed agent orchestrator. Complex multi-machine architecture. Anthropic's Claude Code v2.1.0 dropped this week and I went deep on the research. Here's what's actually happening: The IDE era is ending. Not dying, just becoming optional. Claude Code doesn't live in your editor. It lives in your terminal. It reads your entire codebase, spawns subagents, controls your browser, and loops for 30+ hours autonomously if you let it. The wildest part? ~90% of Claude Code's own codebase was written by Claude Code. We're watching a tool bootstrap itself into existence. The competitive landscape is splitting into three lanes now: 1. Claude Code for deep architectural work and vibe coding 2. Google Jules for async "set and forget" maintenance 3. Cursor for line-by-line acceleration They're not competitors. They're different jobs. I put together a full breakdown, terminal architecture, subagent swarms, the new safety loop, Chrome integration, all of it. The AI race isn't Google vs OpenAI anymore. It's a race to empower the builder. What's your current setup? Single tool or mixing them?

54

42 Comments

I don’t remember any BDFL-led open-source project changing its license aside from Redis. Well… that wouldn’t be exactly true because Redis’ BDFL stepped down (who knew it’s possible!). Last year, Redis, the company, took over the project and ended up changing it to a dual-license making it “source available”. Now Redis, the company, changed Redis license to AGPLv3. While I think that business-wise this is a good move - it’s a pretty dirty trick to pull from an open-source trust perspective. The license change from a pure open-source perspective is too little too late - trust is already broken. I think the biggest win of Valkey over Redis is that it’s backed by the Linux Foundation which essentially makes it a “forever” open-source project. Here is an year old interesting blog post from a competitor to Redis:

25

1 Comment

so many engineers saying they turning their org to AI first org, none of them talks about the fact that one of the KPIs (beside dev velocity obviously) for that effort is actually reducing amount of devs.

1

At Rackspace, a colleague once said it "freaked him out" that I could hear "my Plesk server's load average is over 1000" and immediately say "rename file X in directory Y." It wasn't magic. It was years of pattern recognition compressed into instinct. The problem: that knowledge lived in my head, not in the organization. When the customer report came in, tier 1 agents stared at "my site is slow" and wondered where to start. Below, I wrote about what AI-assisted support could actually look like, and it's not chatbots deflecting tickets, but tools that give every agent the same diagnostic starting point an experienced engineer would have. The goal isn't automation. It's augmentation. https://lnkd.in/gxT62yuj

21

There are now more & more agent skills marketplaces showing up. Thousands of skills across them. How many of those skills actually pass the spec? I ran Sensei's spec compliance checks against many of our own skills last week. Even we were missing things — no license field, no version metadata, directory name didn't match the skill name. These aren't cosmetic issues. The agentskills.io spec exists so that skills are portable across Claude, Copilot, Gemini, Spring AI, and every other platform that adopted the format. If your skill violates the spec, it might work on one platform and silently break on another. In v1.2 we added programmatic checks for every constraint in the spec:  → Name format (lowercase, max 64 chars, no consecutive hyphens) → Required fields (name, description — validated, not assumed) → No unknown frontmatter fields → Directory name matches skill name → Description under 1024 characters → License and version metadata (recommended, not required) This isn't just linting. It's a portability contract. If you're publishing skills to a marketplace, run a spec validator first. If you're building a marketplace, require one. The spec is the floor. Quality starts above it. https://lnkd.in/ewU5Ni5A

4

1 Comment

𝗔𝘇𝘂𝗿𝗲 𝗡𝗔𝗧 𝗚𝗮𝘁𝗲𝘄𝗮𝘆: 𝗛𝗼𝘄 𝗜𝘁 𝗪𝗼𝗿𝗸𝘀 I’ve been learning Azure lately, and I discovered that Azure NAT Gateways work a bit differently compared to AWS. I ran into an interesting issue. After setting up my NAT gateway, I attempted to ping a public domain from my private virtual machine, but it didn’t work. After about an hour of debugging, here’s what I learned: 𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀: You attach the NAT Gateway to a subnet, not directly to the VM. Any VM inside that subnet automatically uses the NAT Gateway for outbound internet access. Don’t try to manually provide internet access through NSGs or route tables, as that defeats the purpose of the NAT Gateway. 𝗦𝗼 𝘄𝗵𝘆 𝗱𝗶𝗱𝗻’𝘁 𝘁𝗵𝗲 𝗽𝗶𝗻𝗴 𝗰𝗼𝗺𝗺𝗮𝗻𝗱 𝘄𝗼𝗿𝗸? Well, the Azure NAT Gateway only supports TCP and UDP traffic for outbound connections, not ICMP. The ping command uses the ICMP protocol, so it was never going to work through the NAT Gateway. In other words, my setup was perfectly fine the entire time... I just didn’t realize that ICMP isn’t supported. There was no misconfiguration after all, just a protocol difference. And honestly, figuring that out was super satisfying. This is exactly how I love to learn, one “why isn’t this working?” at a time! #SRE #cloud #azure #tech

36

1 Comment