goodman.dev
Curious mind. Creator of https://plnkr.co. Tech lead for extensibility at Auth0 (OKTA). I will never stop experimenting.
72 posts
61 followers
119 following
Regular Contributor
Active Commenter
comment in response to
post
Excellent job on the release and blog post.
comment in response to
post
Does the team plan to offer some of the component packages as public, stable-ish modules?
That could create quite an inflection point for the go JS tooling ecosystem.
comment in response to
post
Really just the browser (monaco).
comment in response to
post
Do you have plans to continue to support runtime environments that can't run native binaries? WASM build and bindings maybe?
comment in response to
post
What does this mean for use-cases that run in browser-like environments where we can't run native binaries? Will a WASM build and wrapper be created?
comment in response to
post
On X, I had explored this space with an asyncish/awaitish syntax: x.com/filearts/sta...
These would be opt-in graph colour-buster syntax that a sufficiently good compiler could down level. But runtime support would be needed to really unlock the perf potential.
comment in response to
post
It has "A Tale of Two Cities" vibe building.
comment in response to
post
I think it's the same phenomenon that explains certain framework flame wars. Folks sometimes associate themselves emotionally with projects, VC funded or not.
comment in response to
post
It seems like that provides everything needed to hack arrays and maps into css 😁
comment in response to
post
They're feeling the years for sure but maybe they've got a few more in them 😅
comment in response to
post
I've been rocking the same slippers all day every day for almost 20 years. They were swag for a college sports team I was on. Came from somewhere in Slovenia that I've never been able to track down. Leather and cork.
Might these let me move on?
comment in response to
post
I still haven't figured out WTF nitro is and hope I never have to.
comment in response to
post
🤦♂️ thanks
comment in response to
post
Do you think it is a compelling design?
comment in response to
post
I think that I would be able to get more done straight from within Slack if I had access to tooling akin to Copilot.
comment in response to
post
As Clients!
comment in response to
post
Slack, Obsidian, Confluence
comment in response to
post
Heck yeah @joelhooks.com, I've gotta say that your sharing of your journey was part of what tipped the balance for me.
Now I pay for personal training and am convinced it has unmatched life ROI.
💪🏋️
comment in response to
post
I love the idea of topical feeds. Maybe I default to following the user but later have the option to refine the feeds I follow.
As a Bsky user, I want to use it for technical content. Others will have very different wants. Topical feeds seem like a nice way to address this.
comment in response to
post
I should not have made such a confident answer without context. You're probably right.
comment in response to
post
I would question my level of respect for the author. This would probably also make me reluctant to trust the content or quality of said book.
comment in response to
post
The discussions highlight that this is probably not something that can ever be derived via static analysis. In other words, there may not be objectively right answers in every case.
So maybe the breaking change definition itself needs to be part of the public interface.
comment in response to
post
Oh no! Chrome, Firefox and Safari are also vulnerable! Better file against them to be sure.
comment in response to
post
I've been working on a CLI that acts as a server for _many_ MCP servers.
```
mcp server install foo --env K=V # Install foo with env
mcp server install fs --mount $(pwd):/mnt/data # Install fs, exposing part of the fs
```
The MCP client would be configured with a single server: `mcp run`.
comment in response to
post
For example, docker would be a great common denominator for executing servers.
In addition, metadata could describe installation requirements like OAuth2 tokens, directory shares or other manual KV data.
WDYT?
comment in response to
post
I would love to leverage the registry via http API. Is that something you are open to? I also think it would be very powerful to add extra metadata to servers...
comment in response to
post
Is this another TLA nuisance?
comment in response to
post
I kinda wish there were some stronger guarantees than "should not" Requests during the initial handshake. IMHO, this shifts complexity from the Client to the Server in a way that seems at odds with the spec's goals.
I could be willing to make an exception for "ping" but even there, why?
comment in response to
post
Would it be considered invalid for a server to receive a request other than initialize before completing the initial handshake?
comment in response to
post
Did you build the whole chat experience? It seems like you must have to inject those beautiful twoslash-esque messages while integrating typescript tool use.
Looks like an awesome build!
comment in response to
post
Does this resonate with you?
This is me trying to look at this through a systems lens, "What is the system at work that creates the need for this complexity?"
I hope to start brainstorming by proposing some ideas but I'm in no way pretending that they're the *best* ideas. How would you tweak it?