fix(svelte): Ensure client is loaded before use

[c12i]

Closes #466

This PR adds measures to ensure the client is not accessible before its initialized. This is achieved by moving the client state to a global state and and sharing it across child components.

Blocked by failing CI

[jost-s]

Lexically LGTM. Can I test this somehow or even better, is this tested on CI somehow? If not, can it be?

[c12i]

Can I test this somehow or even better

You can test this locally by checking out this repo, installing scaffolding via cargo install --path . then running hc scaffold example forum to spin up a test forum app with svelte

is this tested on CI somehow? If not, can it be?

For now, are only ensuring that the ui generated by scaffolding can be built without producing any vite build or typescript errors.

We had a discussion on testing the UI in the past on this spike #421, ready for refinement

[mattyg]

Suggested change

	import { getContext, setContext } from 'svelte';
	import { getContext } from 'svelte';

setContext is not used

[c12i]

Removed, thanks

[mattyg]

Suggested change

	import { writable, type Writable } from 'svelte/store';
	import { get, writable } from 'svelte/store';

type Writable not used, get missing

[c12i]

Fixed, thanks

[mattyg]

SignalType was removed from holochain-client-js due to changes in enum serialization: see https://github.com/holochain/holochain-client-js/pull/322/files#

I think there's a few other places in this PR that use it.

[c12i]

Removed, this was a left over change while fixing merge conflicts, thanks for catching

[mattyg]

I'm running into a few unrelated issues that are preventing me from testing this:

• When I run npm run start, I get an error from hc-spin that the request to issue an app auth token timed out:

[1] (node:976325) UnhandledPromiseRejectionWarning: Error: Request timed out in 60000 ms: issue_app_authentication_token
[1]     at Timeout.<anonymous> (/home/matt/Projects/Holochain/forum/node_modules/@holochain/hc-spin/dist/main/index.js:4576:34)
[1]     at listOnTimeout (node:internal/timers:573:17)
[1]     at process.processTimers (node:internal/timers:514:7)
[1] (Use `electron --trace-warnings ...` to show where the warning was created)
[1] (node:976325) UnhandledPromiseRejectionWarning: Unhandled promise rejection. This error originated either by throwing inside of an async function without a catch block, or by rejecting a promise which was not handled with .catch(). To terminate the node process on unhandled promise rejection, use the CLI flag `--unhandled-rejections=strict` (see https://nodejs.org/api/cli.html#cli_unhandled_rejections_mode). (rejection id: 1)

• When I run npm run start:tauri, I get an error that the conductor config is invalid:

[1] [1] 2025-03-04T23:30:50.857592Z  WARN kitsune_p2p::spawn: crates/kitsune_p2p/kitsune_p2p/src/spawn.rs:31: gossip_arc_clamping network tuning parameter is not set. This is not permitted without "unstable-sharding" feature enabled. Please choose either "empty" or "full"
[1] [1] 2025-03-04T23:30:50.857671Z ERROR holochain::conductor::conductor::builder: crates/holochain/src/conductor/conductor/builder.rs:275: Error spawning networking err=OtherKitsuneP2pError(Other(OtherError("gossip_arc_clamping must be set")))

I'll make issues in the corresponding repos. Have you been able to run this @c12i?

[c12i]

I'll make issues in the corresponding repos.

Feel free to open, I suspect the error might be related to the change in enum serialization as well. Last manual testing was done on dev.18 and things were working then prior to rebasing from main

[matthme]

I'm running into a few unrelated issues that are preventing me from testing this

I'm working on bumping hc-spin and hc-launch right now. hc-launch PR is ready for review here. Once merged, I will udpate holonix.