Did-core: Do we really want to recommend polling for changes?

Created on 31 Jan 2020  路  6Comments  路  Source: w3c/did-core

Section 10.5 Notification of DID Document Changes https://w3c.github.io/did-core/#notification-of-did-document-changes contains:

Self-monitoring. A DID subject can employ their own local or online agent to periodically monitor for changes to a DID document.

Do we really want to recommend polling for changes? That hardly seems like a reliable or scalable practice for constructing systems using or implementing DIDs!

PR exists editorial

Most helpful comment

I don't think the wording is harmful because of the preceding text (noted in https://github.com/w3c/did-core/issues/179#issuecomment-600140403). It's also non-normative.

I don't think the wording is necessary because it's not like developers needed to be reminded of the existence of polling as an option. It's not like it's a very DID-specific or totally out-there solution. If someone wants to poll, they will think of it themselves without the spec's help.

In the interests of just giving people less to read overall, I'm +1 to removing it.

All 6 comments

This is akin to subscribing for updates regarding the use of your SSN... if you don't know who is using your SSN, you don't know that someone else is using it....

Also similar to credit reporting subscriptions...

Or monitoring the activities of a specific ledger address...

https://etherscan.io/address/0x17f6c741c9509c6db4695cca662bd5ff36fbb63b

Can whoever wrote this text speak up and say why this text is even there?

Note that the section right above this one (listed first) says this:

Subscriptions. If the DID registry on which the DID is registered directly supports change notifications, this service can be offered to DID controllers. Notifications could be sent directly to the relevant service endpoints listed in an existing DID.

I will review this and probably create a PR removing the suggestion for polling. That way we'll see if there's actually support for this and why.

I think a subscription service that sits over the top of a DID registry would be the more efficient way of handling this and would be in favor of removing the suggestion to use polling. I think it's suffice to say if polling __really__ is the only way to discover an update, it's likely the obvious answer and therefore wouldn't need to be included in the text.

I don't think the wording is harmful because of the preceding text (noted in https://github.com/w3c/did-core/issues/179#issuecomment-600140403). It's also non-normative.

I don't think the wording is necessary because it's not like developers needed to be reminded of the existence of polling as an option. It's not like it's a very DID-specific or totally out-there solution. If someone wants to poll, they will think of it themselves without the spec's help.

In the interests of just giving people less to read overall, I'm +1 to removing it.

Was this page helpful?
0 / 5 - 0 ratings