Re: addressable identifier?

I would ask the related question:
Do we care (from a spec perspective) where the manifest lives??

It could be external or it could be embedded in the “start” (or only) HTML.  Both have use cases that we care about (as mentioned in this thread).

Like Matt, I recommend that we proceed by saying nothing about the location of the manifest and see how other aspects work out as we begin to figure out what goes in it *and* whether there is only a single such thing or we even need more places to hold stuff.  That may all play into our decisions.

Leonard

From: Matt Garrish <matt.garrish@gmail.com>
Date: Thursday, July 27, 2017 at 7:48 PM
To: 'Hadrien Gardeur' <hadrien.gardeur@feedbooks.com>
Cc: "'Cole, Timothy W'" <t-cole3@illinois.edu>, 'Romain' <rdeltour@gmail.com>, 'Laurent Le Meur' <laurent.lemeur@edrlab.org>, 'W3C Publishing Working Group' <public-publ-wg@w3.org>
Subject: RE: addressable identifier?
Resent-From: <public-publ-wg@w3.org>
Resent-Date: Thursday, July 27, 2017 at 7:48 PM

> I was simply pointing out areas where we have true consensus vs partial agreement from this group so far

Right, and I'm just trying to understand the benefits of embedding. I know what I listed wasn't Dave's case, but it was (once upon a time in epub-land) bandied about as a reason for embedding, and I'm not sure it holds water.

But what about not trying to solve it now and instead leaving it as an open issue for the FPWD that we seek active input on? I know this is a wimp out on my part, but this might be one area we don't need to rush on, and would allow us to move on to areas where we'll probably find more agreement, like the structure of the manifest.

Matt

From: Hadrien Gardeur [mailto:hadrien.gardeur@feedbooks.com]
Sent: July 27, 2017 7:16 PM
To: Matt Garrish <matt.garrish@gmail.com>
Cc: Cole, Timothy W <t-cole3@illinois.edu>; Romain <rdeltour@gmail.com>; Laurent Le Meur <laurent.lemeur@edrlab.org>; W3C Publishing Working Group <public-publ-wg@w3.org>
Subject: Re: addressable identifier?

It tends to make the most sense when there are multiple primary resources (just not loving that term yet). In the case of a single-document publication, finding an external manifest to discover the page you were at is the only page seems like a redundant exercise. But then, extracting the script data and converting it to json maybe offsets any potential optimization embedding affords. (It's tempting to think up over-optimizations.)

I'm clearly in favour of an external manifest myself, for all the reasons you've listed and many others.

I was simply pointing out areas where we have true consensus vs partial agreement from this group so far (based on what I've seen expressed on this mailing list and on Github).

Ideally, yes, but this is an inherent problem on the Web and I'm worried about trying to do this in the specs this group is creating.

If this is a decision that the author takes when creating a WP and its manifest, I don't think it's that big of an issue. All the examples that you're pointing out Tim are mostly problematic because you have to guess what would be the proper choice.