RCH WG weekly – 26 April 2023

Sorry to miss today’s meeting, it sounds like there are some editor’s actions due, that hopefully Dan got the gist of. It would probably be worth raising individual issues on some of the finer points.

Regarding additional meetings, these are, of course, easy in my timezone, but it would be great if we could eventually find something more accomodating to Dan, Kazue, and Ivan.

Regarding the output of the canonicalization algorithm, I think it’s fine to say that this is the N-Quads sserialization rather than a normalized dataset (something that should be renamed, anyway). I don’t think we should get rid of this concept, and it may be useful for other implementations to tap into. Framing it as a Dataset and a Map from Blank Nodes to stable identifiers may be a good call. Regarding sorting the N-Quads, this is already described in the Serialization section, where N-Quads are already described as being sorted into code point orrder.

Perhaps the serialization algorithm needs to be more clear that when serializing individual quads, the blank node identifiers used are taken from the map described as being part of the normalized dataset. Potentially, and alternative map could be passed to the serialization part of the algorithm to allow for different labeling; it should be an ordered map so that the order of canonical identifiers is maintained.

Regarding TPAC, I intend to attend.

> https://www.w3.org/2023/04/26-rch-minutes.html


Gregg Kellogg
gregg@greggkellogg.net

Received on Wednesday, 26 April 2023 20:11:49 UTC