Re: [Geopriv] [geopriv] #5: Client guidance about the datum field

> no, it actually isn't. We assume it's WGS84 - as it's the most common
> CRS and the one everyone is moving towards, that's the safest way
> forward considering what I say  next.

Unfortunately 3825 doesn't say that either.

If that *is* the way forward, that would mitigate the effect of the version flags, although one would hope that a trivial "datum = value & 0x07" will make that unnecessary on future client implementations.

> yeah Ray, but he isn't writing a bis that adds the concept of
> versioning the second time around.

Agreed.
 
> If this were (time-machine alert) 2004, then this should have been
> written into 3825. I agree. But it's not 2004 -- and we shouldn't
> knowingly reduce the number of clients that get location they will
> keep without having the ability to inform anyone/thing there's a
> problem. Not when it affects emergency calling, which we all know
> your stance will affect it adversely.

My stance is based on an (implicit) understanding that clients would already suffer this problem if a new datum were introduced in future revisions of 3825.

> wow -- that's quality Monday morning quarterbacking 5 years later.

I apologise, that was poorly worded.

> Perhaps you shouldn't make you attacks so personal until you've
> written your own RFCs for others to thrash.

The first one is in the queue and due for publication September.

Ray