Meaning of `lastUpdate` for "strange" dates

[Schnark]

The current example for .well-known/gpc.json is

{
"gpc": true,
"lastUpdate": "1997-03-10"
}

For me, this actually raises more questions than it answers: What is this supposed to mean?

Following the definitions above the site claims to abide to GPC requests, but only according to the standard as it was defined in March 1997. But the standard didn't exists then. So I assume this actually means that the site does not abide to GPC requests.

If my interpretation is not correct, then the standard should clearly state what the expected meaning really is.

But if you agree with my interpretation, then I think it is very awkward to represent non-compliance by "gpc": true. In this case the date should be limited to a sensible range, e.g. state it must not be before 2020-10-07 (initial version) or even 2021-05-13 (when the lastUpdate field was added). I think it also makes sense at the same time to state that the current date is the upper limit for lastUpdate.

[AramZS]

I will make a change to the spec so that the date in that example reflects the last build date of the document which is a better example and--at a fundamental level--reflects implementation advice to the reader of the spec at the time they read and decide to comply with the spec.

[dmdabbs]

FWIW, none of the three .well-known/gpc.json files I've found contain a "lastUpdate" attribute.

[SebastianZimmeck]

The change has been merged in #106.