mansilladev
Joined 15 years ago
Karma
403
Submissions
140
Member since
Apr 29
TW: https://twitter.com/mansilladev
GH: https://github.com/mansilladev
View on Hacker NewsRecent submissions
Comment · 1 month ago
I was @ Google I/O at live keynote when they announced/demoed it. Ironically, at the exact same time, me and the guy next to me said (in different words), "Why do we need this?" That was a collaborative Google Wave moment IRL.
Comment · 1 month ago
Almost exactly 20 years ago, two co-founders of Barndoor co-founded Mashery — enterprise cloud API management / governance. OAuth, traffic management (cloud + hybrid on-prem), data transformations, dev portal, and even API design. I was there 16 years ago and worked on developer platform / DX up to when we got acquired by Intel. It’s traded hands a few times, but alas, it’s still kicking as Boomi Cloud API Management. :)
We managed some fairly monster sized platforms through the years — Comcast, Expedia, Macy’s, Starbucks, Best Buy, NYT, USA Today, ESPN.
Memories!
Comment · 1 month ago
Hey HN folks, I'm Neil from Barndoor. I work on platform strategy and MCP operations. Wanted to share some engineering context behind our acquisition of Diaphora and where we’re heading with it.
If you’ve tried taking agentic workflows past the MVP/POC stage and into prod, you’ve probably run into this challenge -- prompt chaining is an easy button, but getting reliable deterministic results (especially when/where you need them). Once your agent starts executing multi-step workflows across external tools, maintaining state, handling failures gracefully, and enforcing strict execution boundaries becomes a real mess. And leaning on a model for every step, even when inference isn't needed, can lead the work down a rabbit hole, and burn a boatload of tokens. And you've probably experienced the limitations of SKILL files, in terms of consistency from run to run, or even jumping from model to model.
Diaphora built their commercial engine on top of Frags (https://github.com/fragshq/frags), an open-source workflow runtime designed specifically around state management and predictable execution of workflows/tasks. Bringing the Diaphora team and tech into Barndoor is about closing the gap connecting policy enforcement directly to the runtime layer (from the security lens), and to help keep innovating + supporting a rock-solid AI workflow/automation engine for the ages.
What happens to Frags? It stays open source. We’re not re-licensing it or reducing any capabilities that are already in the repo. And we aren’t turning it into an abandoned open-core shell. We’re actually putting engineering resources behind it, including work that will improve performance that runs locally / behind the firewall.
Why Frags over existing workflow engines? Most traditional workflow tools ship LLM-backed steps that end up as magical blackboxes, and a lot of modern agent frameworks lean too heavily into unbounded recursion. Frags was built from the ground up to treat state progression, tool context, and execution boundaries as first class citizens.
Frags OSS vs Diaphora (commercial) Diaphora’s enterprise features can be integrated with the Barndoor platform for teams that need governance, auditability, and managed deployment of workflows. Those that want to run it on their own stack, Frags stands on its own as an open-source runtime.
Comment · 1 month ago
Ask the AI chat agent on the website what a BobCoin is. Because, I really want to know.
> I’m not able to find a definition for BobCoin in IBM’s content. If you have another term or topic you’d like to explore, let me know!
Comment · 2 months ago
For me, stock Spotify client on Mac is using ~200MB RAM (no videos / canvas). I probably need to try a leaner build on Fastspotify (built it myself from source), bug the damn thing was using ~350MB RAM. Both RAM counts were based on initial load + playing one song.
Comment · 3 months ago
It’s almost ans if all of the big AI players have an arsenal of model releases and stories ready to volley back if a competitor gets too much attention :)
Comment · 4 months ago
Working on AI security and governance (MCP + LLM gateways, fine-grained access control, DLP, monitoring, etc.) It's a lot more fun than it sounds. :)
Comment · 5 months ago
"It requires the people losing their jobs not burning the system down before the system has time to redistribute the gains."
Who, and how, will the gains be redistributed?
Comment · 6 months ago
I was really hoping this was something I could find on CPAN from the author username perlJam.
Comment · 7 months ago
Anyone else still remember old BBS phone numbers? I’ve got several stuck in my head from 40 odd years ago, but I can barely remember my kids’ cell numbers.
Comment · 8 months ago
I remember reading gog setup instructions, and thinking, "Create oauth app/client? That's bonkers." And as cool+useful as this project looks, it's quite a bit harder to get going, especially if you're not familiar with Google Console and OAuth (or not a dev).
Reading lots of comments about "MCP vs CLI" -- reminds me a bit of the "agent vs. script/app/rpa" debates. It's usually not one or the other, but rather, both.. or the right tool for the job (and that can shift over time).
Biggest complaint we have about MCP is bigger context windows and token spend. Tools do exist that address this. I have just one MCP endpoint with a half dozen tools behind it, including Gmail, Google Calendar, Docs, Github, Notion, and more. Uses tool search tool (ToolIQ) with tiny context footprint. Give it a whirl. https://venn.ai
Comment · 1 year ago
“…our cognitive abilities and creative capacities appear poised to take a nosedive into oblivion.”
Don’t sugarcoat it. Tell us how you really feel.