Run it on Vite first, with dummy data. Build for production, remove the dummy data, merge with your Worker

I love the sound of that, just need to figure out what's involved. So far I'm not using any cloudflare services.. but it's not obvious how to get a remix.run project that targets cloudflare running in a node environment

@HardAtWork I guess I could probably do a vite string replacement for

@remix-run/cloudflare

that would get me close

not sure what I would do about cloudflare only stuff like KV backed session store, I would somehow need to avoid importing it when building in vite and use cookies as a fallback

maybe importing it wont cause issues and I can just do an

if

..

I know. That why I like an outbound spam filter

I just had email contact with them. They told me (I am adding this for others) that they will send a bounce email to the sender if they bounce the email. Also I was able to send the Return-Path and I think they respect that. I also got an error on the api when the message was rejected. For me it seems to work!

What would be the best way to achieve this? A worker where people can connect to using WebSockets. I'm using Durable Objects for that and it does work. The WebSocket "server" then needs to fetch another (external) API periodically and then broadcasts it to all connected users. The issue is, that those fetches count towards the subrequest limit so eventually the worker dies. Which is not what I want in a WS environment. Is this just something I cannot do with CF Workers or is there another way?

Do the fetch in the Durable Object instead?

I am already doing it there

Not sure if it's just a limitation of Miniflare since I got the subrequest warning locally

If you use Alarms, it should refresh your subrequest limit.

I am using alarms, still got that warning 😄 I'm fairly new to Durable Objects, so maybe I'm just doing something wrong

Not 100% sure, but that seems like a Miniflare issue...

So, if I deploy it to Workers, it should run without any issues? There probably still are limits, or not?

For DOs, I believe the limit is 1k subrequests per invocation, which should refresh on an Alarm trigger, afaik

Okay, quickly reading into alarms, it seems like that's what I want to use, let's see if I can make it work. Thanks! 😄

hello o/ is there any way to make wrangler work with a workerd instance that's hosted in a docker container?

Careful, I think you end with duplicate headers

No. It overwrites the return path. I checked. Also bounces works. Just need to make a mail parser for the bounce emails

hmm weird, I wonder which header it was that I could duplicate..

There are a couple that are not allowed or maybe it was a bug

If I have a cron worker on cron.worker.dev and I want to fetch another worker from it for example fetch.worker.dev it would fail since they are on the same zone worker.dev, right? But if fetch.worker.dev would be fetch.tld it would work? I can't use service binding because it need sometime fetch website that isn't worker

Hi. I'm trying to use KV through dashboard. I've added a namespace and bound that to a worker settings variable. But when I try to reference it in the editor, it says it is not defined. I'm attaching screenshots

Are you using the Module Worker format? (

export default ...

)

If so, it'll be available in

env

(second argument to your

fetch

method, and not as a global.

Thanks, that was it. Is this documented somewhere? Looks tricky to me

https://developers.cloudflare.com/workers/learning/migrating-to-module-workers/ and https://developers.cloudflare.com/workers/runtime-apis/kv/#kv-bindings mention it

Hey, is it possible to schedule worker programmatically from another worker? I can only find examples using wrangler. EDIT: looks like durable object alarms is what I need.