Storage workflows/tech for film/TV/media. PMKTG/AI+human code and knowledge tools. Former Apple servers+storage evangelist, fmr Marine - write/code/teach/hype!
@AppleTV Silo episode before the finale… then BAM @itspetergabriel as Himself doing an absolutely harrowing rendering of Here Comes The Flood - incredible moment and a great piece of foreshadowing….
Some new retro-terminal profiles - scifi: MCRN, Rocinante, Empire and Prime Radiant - and retro: Convair - blueprint and whiteprint styles - see em all here: https://t.co/26sO8NGV0f
hard to describe the thrill of helping work on and evangelize these incredible products at the time - together with Xsan and Mac OS X Server!! to this day the best server and storage bezels ever created! Okay - we tried to top other with Active Storage bezels - but still…
Apple used to make rack mount servers and storage, called Xserve and Xserve RAID. I was hired by Apple in 2003 to be an SE (now called Forward Deployed Engineer) 😀 on Xserve. A custom SKU for a server for AI isn’t something I see as possible, it’s too specialized.
I could see:
- more server-like features in the os. (In fact in the latest betas of Golden Gate there’s a feature to auto restart on power-on in the settings UI. A small affordance to server use.)
- the old tower Mac Pro could be side mounted in a rack. I could see Apple making a rack mount for Mac Studio or Mac mini. (Or more likely reselling someone else’s product.)
The OS gap is huge though, the default macOS security and privacy settings fight server use. A “make this Mac a server” checkbox in settings would be a huge win.
This project started out as 'I liked the old w3w browser in a terminal window' - why can't that work, but be markdown?? Turns out - its a really efficient way for LLM's to drive the web, and for humans to have a quiet way to browse whats _really_ on a web page - you know.. the _content_??
Terminal browsers broke 20 years ago: they parse HTML and the web went JavaScript.
mdbrowse fronts a real browser engine: your terminal gets the semantic page. Headings, prose, links, tables. Your cookies, your logged-in web, no noise.
brew install giantravens/tap/mdbrowse
Also the cheapest web an LLM can read: raw HTML costs ~9x the tokens per fact. Bundled MCP server, deterministic captures with provenance hashes, watch sensors that fire on real change, full-text archive of everything you capture. Open source, MIT.
https://t.co/oES4wFv8oT
Built for human+agent co-work on the same files: metadata-first listings (~500 tokens for 600 tasks via the bundled MCP server), a timestamped audit log on every change, and pin doctor, which exists because agents really did corrupt task files.
https://t.co/c3fIvIsMbi
LLM coding agents have no memory so their todo lists vanish when the session ends.
punchlist: every task is a plain markdown file. Agents file work, tick acceptance boxes, log telemetry. The next session picks up exactly where the last one left off.
brew install giantravens/tap/punchlist
https://t.co/26sO8NGV0f
You can sorta fake 'old skool' crt effects in iterm2 profiles - but you can REALLY geek out if you render them in ghosTTY via webGL - fun CRT effects configurator had to include.
Retro terminals for Weyland-Yutani, Wallace Corp, etc.
★ A native graphical shell for SSH
I spent time rethinking how we should interact with servers, Raspberry Pi, robots, etc., then built it up piece by piece.
Great cause and a great org - all sales help support Scout Sniper Association!
ANNUAL RAFFLE: USMC Scout Sniper Association (SSA) | Raffle Creator https://t.co/p84zzBTEa0 via @rafflecreator
Only two more weeks left until the @MacDevOpsYVR conference begins! Join the event for Mac Admins in Montreal from June 25-26th. Secure your ticket here: https://t.co/IZD7fdD6Kr
Marines now need 30 months time in service to be eligible for sergeant, reduced from 48 months.
If you haven't completed your Math For Marines MCI, that is 18 fewer months. But the Marines say they have a reason for the reduction.
Story: https://t.co/SV2cmI3DUk