RE: agenda topic for June 4 call: formality of EPUB 3.2 approval steps

Hi Bill,

I think there may have been a misunderstanding. My take away from the meeting was that Dave was all for having a formal approval period for EPUB 3.2 and that the BG should definitely be a part of that approval. I don't think anyone is in disagreement about that. There is not a lot of precedent in the W3C for this sort of structure, so I believe that Dave was saying that we need to make up the process to some extent.

I think a good agenda for the SC is to propose a process for BG approval. I believe that is exactly what Dave was suggesting. Dave, will of course correct me if I am misunderstanding.

Tzviya



Tzviya Siegman
Information Standards Lead
Wiley
201-748-6884
tsiegman@wiley.com<mailto:tsiegman@wiley.com>

From: Bill McCoy [mailto:bmccoy@w3.org]
Sent: Thursday, April 26, 2018 10:26 AM
To: 'W3C Publishing Steering Committee' <public-publishing-sc@w3.org>
Subject: agenda topic for June 4 call: formality of EPUB 3.2 approval steps

Hi,

Dave Cramer in last BG call said that he thought we should not have any particular formality around the interim approval steps for EPUB 3.2 (such as IDPF's Proposed Specification) to avoid folks getting confused between W3C CG work product and WG work product.

I would like to revisit this and ask the SC to agree where the responsibility should be for deciding this and related questions about the formality of EPUB 3.2, so I would request we add this to our agenda for June 4 (unless it's clear from email that there is an obvious consensus and no discussion is needed).

To me, this question belongs primarily to the BG not the CG since the BG has the responsibility to address business-level ecosystem needs as well as oversight and approval responsibility for the work product of the CG. Of course the BG shouldn't be allowed to burden the CG excessively but it seems to me the BG should have the main voice in deciding how we need to promote and market EPUB. And approval of specifications is a key part of that promotion and marketing. Dave you may in fact have intended to provide a BG perspective representing Hachette Book Group, since it was during the BG meeting, but this wasn't clear (at least to me) given your co-chair-ship of the CG. But since it wasn't an explicit agenda topic I didn't want to chime in on it in our limited time.

If we agree that this is a BG decision then I would propose to add a topic for approval steps for EPUB 3.2 for the next BG call and perhaps have some email about that en route.

Because, I believe Dave's plan to avoid any formality is contrary to the business needs for a stable and reliable EPUB specification, both in general and based on the current need to crisply move past the EPUB 3.1 snafu. I agree with Dave and others that EPUB 3.2 is inherently not that important in the sense that what we need to be promoting is EPUB 3, but the fact that the latest extant version of EPUB 3 is problematic and that the organization that developed it no longer exists or supports it is a weak point. That work has started on what may be an EPUB 4 further confuses things. I'm told there were people at last Frankfurt saying "EPUB 3 is dead" and that's troubling. So I think it's critical that EPUB 3.2 be seen as a real thing not an ad hoc product of an ad hoc group. And the arrangement between IDPF and W3C was explicitly designed to facilitate this. The BG oversight and approval of the CG work on EPUB 3 revisions was intended to function in a directly analogous manner to the AC approval of W3C specifications produced by WGs, and the membership approval of IDPF specifications. Both approval processes include approval of interim versions not just a blessing of the final work product. So I think that we should do likewise and have at least one interim approval gate by the BG en route to finalizing EPUB 3.2, and do some PR around the hopefully positive result.

Yes we should avoid confusion about the CG work product being the same as a W3C Recommendation but this seems like a minor concern especially since we will be using the CG document template a la [1], something we've already agreed on and that I don't propose to revisit.

Dave's overall approach seems to be based on an argument that since CG's don't have any particular process requirements that therefore we should not impose any. But this is both contrary to the IDPF-W3C understanding as well as a logical fallacy. That some (maybe most) CGs are very ad hoc about their processes doesn't imply that we must be also.  That there are no particular process requirements for CGs doesn't imply that we can't have any, it only means that we are free to do whatever makes sense to meet the overall goals.

And while I'm not suggesting that we invent a new report template I do think we can and should get creative in maximizing EPUB 3.2 being seen as a "real specification" equally if not more vetted and valid than anything previously produced by IDPF. If we don't do this, we risk further muddying the waters which again with 3.1 being problematic would be extra bad. And as I've said, we need to show that EPUB 3 is alive and well and being well maintained within W3C.

If we quickly do an ISO standard based on 3.2 and/or a W3C Recommendation a la JSON-LD, then that will of course help. But in either of these paths we will not want to create forking. We don't want a 3.2 from the CG that is merely viewed as fodder for an ISO standard or W3C Recommendation, we would want these outputs to be completely compatible. So that still argues for focus on EPUB 3.2 being dependable, not the result of an ad hoc process. So to me the more "window dressing" we can put around this to maximize the perception of dependability, the better. Of course I don't propose that we overly burden the CG, but I don't think that for example bringing at least one interim draft to the BG for approval would do so (and again doing so is arguably an implication of having the approval responsibility per the IDPF-W3C deal as embodied in the respective charters).

Re: naming, I don't necessarily have a great answer but I also think this deserves more discussion. We are in agreement to use the CG Report template which makes it very clear in the preface that the results are not a W3C Recommendation but JSON LD 1.1 [1] very clearly calls itself a "Recommendation": "This document is one of three JSON-LD 1.1 Recommendations produced by the JSON for Linking Data W3C Community Group<https://www.w3.org/community/json-ld/>:"
So I think it's fair game for us to do likewise, adopt IDPF naming, or whatever the BG thinks will best serve the goal. So if we called the final result a "Recommendation", then we could call the interim milestone something accordingly. We should perhaps investigate what JSON-LD 1.1 did.

Again while I have a personal perspective here on what's best for the ecosystem and W3C, my main goal in sending this email is to clarify whether this group feels that this should be a BG decision and if so to get it in the stream to be an explicit agenda topic there which I think deserves a bit more than just Dave saying in passing what he wants to do as CG co-chair.

Thanks,

--Bill


[1] https://json-ld.org/spec/ED/json-ld/20180215/