cleanup agoricNames publishing

[turadg]

follow-up

Description

#9477 (comment)

Security Considerations

none

Scaling Considerations

none

Documentation Considerations

none

Testing Considerations

CI

Upgrade Considerations

This code ran only in tests

[cloudflare-workers-and-pages]

Deploying agoric-sdk with    Cloudflare Pages

Latest commit:	af9e1c1
Status:	✅  Deploy successful!
Preview URL:	https://a7b2d03e.agoric-sdk.pages.dev
Branch Preview URL:	https://ta-agoricnames-cleanup.agoric-sdk.pages.dev

View logs

[dckc]

Where is that link supposed to go? I don't see a README in this dir.

Possible alternative link:

• bootstrap upgrade complicated by 5 non-durable exports #8849

[turadg]

@dckc PTAAL. To ensure chainStorage is always available, I had to mock it in chain-behaviors when there's no bridge. That's production code (or was, during bootstrap) so I'm not fully comfortable putting that in there. I could rename it from "mock" to "in-memory" or something like that.

Thinking in terms of least surprise, one could fairly expect a chainStorage capability to actually write to chain. So I think we're better off keeping it as null in the space when it won't run to chain and handling that case at the consumer sites.

IOW, drop these last two commits. I'd still land the one removing the deprecated function.

I see PTAAL now

[dckc]

Yay! The undefined case for chainStorage was only for the sim chain, I'm pretty sure.

console.warn is exactly the way I think this should be treated.

A big part of me would like to tell the type system that this thing is always there so we don't have to keep supporting the sim chain in new code. That might be more involved, though.
cc @michaelfig @0xpatrickdev

[turadg]

What would it take to remove sim-chain concerns from chain-behaviors entirely? I figure we could provide a fake bridge device here:

agoric-sdk/packages/vats/src/core/chain-behaviors.js

Lines 319 to 328 in af9e1c1

	if (!bridge) {
	console.warn(
	'Running without a bridge device; this is not an actual chain.',
	);
	bridgeManagerP.resolve(undefined);
	provisionBridgeManager.resolve(undefined);
	provisionWalletBridgeManager.resolve(undefined);
	walletBridgeManager.resolve(undefined);
	return;
	}

But I don't know what fidelity it needs or if anything downstream relies on detecting its absence.

[dckc]

What would it take to remove sim-chain concerns from chain-behaviors entirely?

I can't think of any technical obstacles. @michaelfig , maybe you know?

Seems like product (@mitdralla, @sufyaankhan ) should be involved. @sufyaankhan led the charge to hide the sim chain from docs.agoric.com .

[michaelfig]

What would it take to remove sim-chain concerns from chain-behaviors entirely?

I can't think of any technical obstacles. @michaelfig , maybe you know?

No technical obstacles. Earlier in our history, it was considered onerous to force dapp-developers to install Docker or compile agd themselves. Nowadays, we rely on Docker without batting an eye. The times have changed, and sim-chain is a relic we can safely jettison.