E0 Test Reading System and Test Cases

[richardigp]

We have updated AZARDI 23 to reflect the strawman E-Book Zero specification that Dave put up 25 July. AZARDI 23 The idea of this is to address the ePub3 criticisms of some that there are no test implementations or test cases. That is now a non-issue. 

The test cases we put up earlier have all been reprocessed to reflect the new spec ideas, and the previous spine/toc properties and cover treatment have all been treated as we understand the strawman spec E0 Test Cases Updated. 

I guess this is our attempt to kick some life back into this really essential development project. 

We appreciate this may appear very early in the discussion to go real-reading-system. But we are practitioners rather than the specification discussion people, we are maintaining AZARDI on approximately a monthly cycle, so it is not particularly difficult to test E0 ideas. 

We have the capability, development framework and people to implement just about anything (sensible) that can be thrown at us; and the idea of E0 is of course sensible rather than theoretical complexity.  Our focus tends to be coloured by K-12 education and interactivity and we can see that with a few section properties this can take off like ePubX never has. Our motivation is we really need, for the real world, a book package that is something 2013 looking forward to deliver to the world. Not a lot of Schema, XML fallback "stuff". Simple as reading systems, web-apps, apps, static sites, shipping to LMS, one package liberating the world of valuable publisher content. 

We created an E0 Format processor so any ePub we have in our production system can be made available in E0 more or less with the click of a button. (See the comparison E0/EPub3 samples they haven't been proofed, but were assembled by two demo team guys in less than a week E0/Epub3 Comparison Books). Our Formats on Demand programmer Nilesh can usually make the module processing changes in hours so it is not a massive barrier (Python is awesome). 

If anyone has a specific book or content they would like to see processed please don't be shy in asking. We process. Therefore we are! 

I would like to see this very sensible format discussion continue and if the team can come to enough agreement on various "features" we can turn them around reasonably fast at nominated intervals of around a month or less (depending on the feature of course). The focus on the features of clean HTML5 is very appealing to us and we are ready to package and process E0 10 different ways. Instead of being a terminal format like ePub3 it has a dynamic future. 

We would appreciate feedback and criticism from the E0 brains-trust. We have only done index.html and nav interpretation so far. Even at this strawman level the books behave as well as ePub3 any reading system. We can implement property behaviour (did epub:type - nobody noticed), section presentation features and more.

Everything is up for change. So far the index.html page works. We need more info on metadata? Is the cover strategy good-enough? the index remains hidden, the additive TOC is more sensible (my favourite) than the subtractive TOC (first implementation) or assigned TOC properties (as now deployed)? We have to think what people will do rather than just logic. 90% of books just want TOC and Spine it is only exception cases (our textbooks and similar) that want to muck around with them so the simple nav ol seems to be the best default IMHO but it is just an O. 

Anyway we cast our bread upon the waters.... My understanding of E0 is 1. Easy in production 2) Easy to implement for a wide range of packaging options 3) Easy to make a reading system 4) future-ready. 5) HTML5 all the way.