Ian Macartney

@tobeycodes good call - I opened an issue

@ianmacartney tested this and my hunch was correct. needs authorization check

hey @ianmacartney i am trying convex for the first time. shouldn't these functions verify the note was created by the user before deleting them? #L63" target="_blank" rel="nofollow noopener">github.com/get-convex/tur…

@JcCastaneda05 @mikeysee @robelestifanos_ @convex have you tried setting the env var CONVEX_AGENT_MODE=anonymous?

hi, quick one about setting up cloud env for convex, i find claude code's cloud environment so handy on quick fix or edit but when i tried "npx convex..." stuff it requires authentication in terminal and for some reason the agent is not able to authenticate, do you have plans on supporting cloud env like claude's?

Awesome hanging out with @robelestifanos_ and @ianmacartney chatting about Replicate and the potential future local first “holy grail” on @convex

@MeshanKhosla @huntjordan22 Thanks! @ballingt did a lot of the heavy lifting too -esp. on the bundling and internal capabilities

The authoring API is great btw @ianmacartney @huntjordan22 and other who worked on it. I had a fun time using it

I made a @convex timeline component that allows for easy undos, redos, and checkpoints! Check it out github.com/MeshanKhosla/c…

@GorkaCesium "Cooling the ASICs would be trivial" -> you sure? Lacking conduction/convection leaves only radiation, which sounds tricky for heat dumping

The real mining would be to mine bitcoin in space. Cooling the ASICs would be trivial and it can be powered by solar panels.

@tiagoefreitas @kiwicopple I went with the prosemirror OT model over Yjs CRDTs. I started exploring CRDT sync in Convex via the automerge-sync component, but for prosemirror-sync I found the rebase approach to be more correct. But if automerge has a lighter weight server impl that would be great to know

@kiwicopple @ianmacartney looks promising as a scalable solution to add to the prosemirror component?

we've released a CRDT extension for Postgres (experimental and open source) why? A CRDT is a special data type used for collaboration - think of a Google Doc. The changes are merged even as people are typing simultaneously these CRDTs are then serialized on a server and then stored in a database. with a Postgres extension, we give the database a more "native" experience - they can be merged on the database, indexed, and searched. In the future you should be able to do something like this: create table blog_posts ( id uuid primary key, title text, body crdt default '{}' ); then you can post the changes to the database and it will merge automatically (rather than overwriting eachother's changes) you can use websockets (like our Realtime server) for very rapid changes between users, batching the changes to your database

@tiagoefreitas @mdhruvill @Cloudflare @tan_stack @convex @coderabbitai @Sentry I've been thinking about what a git-compatible VFS Component might look like, cool to see this impl Not sure how hard it'd be to do with Convex Workflow / etc. Why not embrace the DO approach? @Cloudflare is very complimentary to @Convex IMO esp. on the hosting/r2 front

@mdhruvill @Cloudflare @tan_stack @convex @coderabbitai @Sentry does convex not provide an equivalent to durable objects the could have been used here? workflows wouldn’t work? @ianmacartney

Introducing GitVex - a fully open-source serverless git hosting platform. No VMs, No Containers, Just Durable Objects and Convex. Code: github.com/mdhruvil/gitvex Built using: @Cloudflare @tan_stack @convex @coderabbitai @Sentry Demo in reply 👇

Long time coming, but you can now do v.nullable(...) as shorthand for v.union(..., v.null())

#reusing-and-extending-validators" target="_blank" rel="nofollow noopener">docs.convex.dev/functions/vali…

Handy validation utilities landed in @convex 1.29: v.object({name: v.string(), age: v.string()}) .pick("name") .extend({id: v.id("users")}) .partial(); I'll be working on the core API in addition to convex-helpers & components

Grok did a decent job, but to make it more concrete I've built two components to do this. Workpool - durable functions with steps, suspension until time/event & more Workpool - general-purpose queues w/ parallelism limits, retry configs, exactly-once execution of onComplete

@essonam_m @convex @jamesacowling @jamwt @rasmic I pushed a new version that supports the string version of models - previously you had to use the gateway provider

I need help! @convex @jamesacowling @jamwt @rasmic @ianmacartney

@lgrammel I'm adding to @​convex-dev/workflow * to improve HIL, incl. AI SDK loops- adding pause/resume- akin to promises / signals / condition variables for Convex fns. Planning to ship soon - would love to chat / coordinate! *Durable async serverless functions akin to termporal

Continuing to prototype tool execution approvals. This is how the UI components for tools could look like with approvals:

..of course sometimes skipping meaningless stack frames is good, and nothing is absolute

try { return foo(); // async function } finally { // executes before foo finishes* } * code in foo prior to the first await may run first

I feel clever subbing `return await fn()` with `return fn()` But gotchas don't seem worth it anymore • Breaks try/catch/finally expectations • Loses stack trace context My default is now to always await. Change my mind.

@ZubairIbnZamir @convex It's up to date! There's also an ongoing effort to support zod4, which will be backwards compatible (using the same clever trick zod used of a separate /zod4 entrypoint)

@ianmacartney is this still the way to go, or kinda outdated? stack.convex.dev/typescript-zod… #convex @convex

@grinich The slides from the talk on Async are here: convex.link/async-agent

@ianmacartney can you post your slides from tonight’s demo?

Big updates to the Agent Component: - Abort an async stream remotely - Pending, failed, and aborted streamed messages auto-save - AI SDK v5 support - Easier to use alongside AI SDK vs. wrapping - Type-safe custom context passed to tools - <SmoothText>msg.text</SmoothText>

@vishyfishy2 You can do either - http streaming has always worked via the `toUIMessageStream` & co. Delta streaming is opt-in and streams only the changes to all subscribers (and can be resumed). Doing both is possible right now - simplest is to read both the deltas & HTTP & de-dupe tho

@ianmacartney or is it using a resumable streams architecture now

@maxjean__ The `ai` npm package underlies some of the Agent component - and with their v5 change a lot of the data model changed, requiring some changes to translate to/from the database schema the Agent uses, along with changes in behavior for streamText / etc.

@ianmacartney What does ai sdk 5 support mean ?