Hi mrmikel,
Quote:
Originally Posted by mrmikel
As for epub3 support....
That may be one of the issues the developer had. I suspect it might be almost easier to create a separate program for epub3 than try to shoehorn it into this existing epub2 program especially since it is always aiming for valid code. I think there is a lot in common, but not enough to just be able to bolt on epub3.
|
The reason Sigil shoots for valid code all of the time is that is uses Qt and QWebKit and probably libXML2 and therefore needs valid code to have it display in bookview properly and to be properly parsed the first time. It can actually be really hard to parse poor, improperly nested, mismatched, mixed case, junky html to create something understandable. That is why both Sigil (and Calibre I think) use a modified version of tidy to try and make something that at least can be parsed.
epub3 will not change that. In fact it is valid (and makes sense) that both a nav.xhtml and toc.ncx be present, that metadata in the old format be allowed and present (although ignored) as well and so on. It still has to be valid code to be understood and previewed.
I tend to think that epub3 simply is epub2 with strange new metadata formats with refines and ids, new xhtml navigation, etc. The scripting stuff and use of links to get internet based content can be ignored for now. And it should still parse just fine and can still be kept as "valid" code.
Quote:
|
Kindle is the biggest part of the marketplace and many people start with epub as a base. It is not a stationary target though and there is the possibility of endless support questions about why the created epub won't work once converted to Kindle.
|
I like epub much better and find it much more sane in so many ways than the compiled Kindle formats. As one of the developers of KindleUnpack, I see the internals of the Kindle format in all its glory. Amazon spends huge amounts of developer time shoe-horning in new "features" into an ancient, small-memory footprint compiled format (mobipocket) just to keep backwards compatibility.
None of what they have done is needed or useful at this point. What a colossal waste of developer time and money! Amazon truly does not understand the KISS rule of engineering nor do that understand how to effectively deploy a software development team to have maximum productivity at minimum cost.
But you are right, part of the problem of keeping Sigil alive would be all of the time it takes to support all of the user errors that people keep reporting that really should not exist. As far as I can tell, Kovid is a true master at rejecting these insane reports with a minimum of effort. All young developers should learn from his approach, otherwise they run the risk of never getting any serious development done.
Interesting talking to you.
Take care,
KevinH