Re: [dpub-pwp] Pull request for rewrite of intro to PWP white paper

> On Dec 14, 2016, at 17:54, Ivan Herman <ivan@w3.org> wrote:
> 
> Florian,
> 
> I understand what you say, but…
> 
> (comments in the text)
> 
>> On 13 Dec 2016, at 04:47, Florian Rivoal <florian@rivoal.net> wrote:
>> 
>>> 
>>> On Dec 13, 2016, at 11:45, Dave Cramer <dauwhe@gmail.com> wrote:
>>> 
>>> On Mon, Dec 12, 2016 at 9:07 PM, Florian Rivoal <florian@rivoal.net> wrote:
>>> 
>>>>> Furthermore, a web publication is an ordered collection of resources. The publisher must provide a default sequence through the primary constituent resources
>>>> 
>>>> Really? An encyclopedia seems like a valid candidate for a web publication, but it is not fundamentally ordered. I agree that a web publication needs an entrypoint and needs to have all parts of the publication be reachable via internal navigation, but a linear ordering of the content seems to be only one way to achieve that. If you want to print the whole thing, you're going to have to find a way to linearize it, but it seems to me that the claim that ordering is a defining characteristic of a web publication is a tad too strong, even though it is not completely baseless.
>>> 
>>> The primary mode of organization may not be sequential, but I think
>>> it's valuable to have as a fallback. Even encyclopedias traditionally
>>> defaulted to alphabetical order. Being able to print something is a
>>> valid use case. I'd hate to hit "print" and then be faced with a
>>> drag-and-drop interface for sequencing all the components of a
>>> publication! Defaults do come in handy. Magazines and newspapers
>>> convey meaning via the spatial organization of their content, even if
>>> the constituent components mostly stand by themselves.
>>> 
>>> One lovely thing about books, magazines, newspapers, letters, etc. is
>>> that you can get to all the content without doing anything special.
>>> Just keep turning the page and you'll get everywhere. I can't do that
>>> with Wikipedia, or with the NYTimes website. Maybe there's something
>>> fundamental in a publication being  bounded and ordered, or at least
>>> potentially ordered.
>> 
>> Right. I think there is value in a Web Publication being ordered, or at least potentially ordered. I am not sure this is strong enough to disqualify from being a (P)WP things that are not.
>> 
>> In other words, it would be bad if the concrete format(s) we eventually settle on to describe PWPs did not have a means to express a linear order, and suggesting an approach for giving an order to works where it is not obvious is probably useful, but I don't think it would be good either if the format(s) were incapable of describing a non ordered work.
>> 
>> For instance, while it is probably convenient (and DRY) for the data-structure that is used to list all the resources in a publication to be able to express an order for these resources, if that ordering is mandatory and can be used for some (P)WP as the sole navigation mechanism, the (P)WP User Agents will have to present that ordering to users, making this unusable for non-ordered content.
>> 
>> Besides, even if there is an overall linear order for the document, not all resources are part of that order. Some, instead of being navigated to/from other resources, are used in other resources (images, fonts, HTML documents in iframes...), and are therefore neither before or after anything.
> 
> I do not think that follows from what Dave was saying.
> 
> * I believe we were only talking about a "default" ordering. That does not preclude (a) that a number of alternate ordering would be defined and (b) that does not mean that _all_ resources _must_ be part of that default order. It is perfectly conceivable that (to use your example below) the 'unordered.html' file is either part of an alternate order, or it is only reached via a link from a different file.
> 
> * The default order may mean a single file.

Then we are in agreement over how things should be. However, the way it is phrased and the place where it is phrased (the introduction) seems to imply that it is a defining characteristic of a publication that it be a ordered set of resources. If a publication can be composed of many unordered resources with a single one being in the default order (which you say is fine and I agree), then being ordered is not a defining feature of publications, and that sentence should probably be reworded a bit.

—Florian

Received on Thursday, 15 December 2016 11:28:35 UTC