Re: [pwg] Meta comment about complexity

Ric,

thank you for your third item:-)

Staying on the meta level: I am actually thrilled that the discussion was so intense; believe me, it is way better than having a working group that takes 3-4 weeks to have any meaningful discussion happening! 

However… I think what we have to be very careful in future discussions is to avoid bifurcations in a thread, let that be on github or in the mailing list (I am not sure why some discussions migrated to the mailing list, b.t.w.). I have seen many examples of this: for example, the thread on the information content on manifests[1] got a long discussion on whether the EPUB3 nav document is useful or not. While that is a possibly important issue for our work as well because it may influence how we would serialize the various items in the manifest in general, and the reading order/TOC in particular, this made the reading of the original issue very hard. 

Github issues are cheap resources! We should open a new issue when something like that happen, have the discussion there, and focus on one topic per issue. Ie, create a new issue, make a backward/forward reference in the discussion thread.

Ivan

[1] https://github.com/w3c/wpub/issues/6 <https://github.com/w3c/wpub/issues/6>
[2] https://github.com/w3c/wpub/labels <https://github.com/w3c/wpub/labels>



> On 28 Jul 2017, at 21:02, Ric Wright <rkwright@geofx.com <mailto:rkwright@geofx.com>> wrote:
> 
> Folks,
> 
> I have gotten way behind in my reading lately due to Readium management, preparing to move and some health issues (note to self: you can’t bench-press 200 pounds anymore so why did you think you could manhandle a 200-pound, fully saturated hot-tub cover down a set of stairs?  :-)
> 
> So I have spent most of today reading or re-reading all the issues and the last few weeks of traffic on the mailing list.  Two items strike me, one procedural and one substantiative (at least IMO).
> 
> First, there is a LOT of duplication and rehashing in the issues and emails.  This can’t be completely avoided, but I think we need to work a lot harder at keeping the issues and emails more granular and focused.  I know this is not a new comment, but reading all of it more or less at once certainly emphasizes this problem.
> 
> Second, as many of you know, I have had a fair amount of experience on both the reading-system side (SVG Viewer, Tahoe, RMSDK/ACS4/Digital Editions, Readium 1/2) and the authoring side (SVG/InDesign/Illustrator/LiveMotion, EPUB/InDesign/epubcrude/custom tool-chains).  
> 
> One of the reasons, IMO, for the lack of full-featured compelling EPUB3 documents is that authoring complex, full-featured EPUB3 is HARD. Rendering is actually easier, especially as a lot of the work is delegated to the browser. But most EPUB3 documents are really just EPUB2 with a navdoc.
> 
> Reading through all these emails and issues, I am kind of boggling at how someone could actually author something that could work as a plain web-page, as a WP, be packaged as a PWP and/or EPUB3/4.  It’s doable of course – anything is doable (old adage: “Sure, we can do that – how long do you want to wait and how much do you want to pay?”).  But if the bar to create portable content is too high, we will end up in the same place as EPUB3, with almost everyone creating content to the lowest common denominator, both because then it will work most places and it isn’t wickedly difficult to create.  If that happens, you could argue we have failed.  
> 
> I don’t have any magic wand or bullet to solve this, I just thought I would point it out (again) so we keep it in mind when we start contemplating complex scenarios.
> 
> Finally, bear in mind that reading systems, in general, can’t just support WP/PWP/EPUB4 – they have to support all the old stuff (EPUB2/3) as well.  In some cases, reading systems have simply crammed multiple engines in there (e.g. Adobe, BN and Kobo come to mind), but it is not an elegant solution and incurs a significant cost in testing if nothing else.
> 
> Anyway, just my two cents after a long, long slog through all the stuff.
> 
> BTW, there is a third item – Ivan is going to have fun when he gets back… :-)
> 
> Ric
> 
> 


----
Ivan Herman, W3C 
Publishing@W3C Technical Lead
Home: http://www.w3.org/People/Ivan/ <http://www.w3.org/People/Ivan/>
mobile: +31-641044153
ORCID ID: http://orcid.org/0000-0003-0782-2704 <http://orcid.org/0000-0003-0782-2704>