search

w3c / rch-wg-charter

Charter proposal for an “RDF Dataset Canonicalization and Hash Working Group”
https://w3c.github.io/rch-wg-charter/
Other
12 stars 7 forks source link

how the canonicalization algorithm is chosen #105

Open pfps opened 2 years ago

pfps commented 2 years ago

There appears to be two different canonicalization algorithms that are referenced in the draft charter - one from Aidan Hogan and one from Digital Bazaar. It is unclear which one came first. It is also unclear which one performs better.
If one of these algorithms is chosen the WG should produce reasons for preferring that one that are only based on the characteristics of the algorithms.

pchampin commented 2 years ago

Section 6 of the charter reads:

Technical discussions for this Working Group are conducted in public: the meeting minutes from teleconference and face-to-face meetings will be archived for public review, and technical discussions and issue tracking will be conducted in a manner that can be both read and written to by the general public.

I believe it covers the "should produce reasons" part of your issue.

As for being "only based on the characteristics of the algorithms", I would leave it to the WG to decide which "characteristics" are most relevant to make this decision...

Therefore, I suggest to mark this issue as "postpone_to_WG".

dlongley commented 2 years ago

Do we need the charter to say that we might detail how to implement more than one canonicalization algorithm -- but recommend one for legacy applications and another for new applications? My view is that the WG already had the space to do that.

peacekeeper commented 2 years ago

we might detail how to implement more than one canonicalization algorithm -- but recommend one for legacy applications and another for new applications

The deliverables section says this: "RDF Dataset Canonicalization (RDC) - This specification defines an algorithm that implements an RDF Dataset Canonicalization function."

My understanding is that this will be only one algortihm, not more than one.

The deliverables section also says this: "Other non-normative documents may be created"

So I guess the WG could deliver one normative specification with one algorithm, an optionally additional non-normative documents with additional algorithms, if that is desired.

Personally I think this is fine and would also suggest to mark this "postpone_to_WG".