
This thread was kicked off by Stevan White saying that he had taken some PG-generated epubs apart and re-formatted them, and that he though they were improved by doing this. Having been through this process myself about 18 months ago, and got nowhere with persuading PG to do anything about its epubs (my original proposal was simply that PG should add hand-corrected epubs to its current repository, which may now seem naïve, but was well-intentioned), I contemplated the feasibility of starting a web site myself, persuading others to join me, etc., etc. (and also having an offer to host my books on his site from BB), but eventually concluded that since my basic interest is in books, not running a web site, the simplest thing to do was to add them to the MobileRead library at (http://www.mobileread.com/). Now my personal belief (though I can't prove it) is that: 1. What Jim Adcock calls the dirt simple books constitute the vast majority of downloads from PG. 2. Much, if not most, of the great literature of the English language that interests me is already in the PG repository (although sadly often in a somewhat unsatisfactory state) and furthermore is contained within the above-mentioned dirt simple subset. 2. Most of these books could be satisfactorily encoded by limiting HTML markup to the styles contained in a relatively small css stylesheet, and by obeying a small number of rather uncontroversial HTML coding standards (things like what heading levels to use for volumes, chapters, sections within chapters, etc.). (They are, after all, dirt simple.) 3. If these books were thus encoded, then it would be a relatively simple matter to generate epubs from the HTML (and maybe even simpler to do it the other way round). Probably also you could generate plaintext from the HTML, so that you could (for these books) have a single master version in HTML. Over the last eighteen months as I've reformatted various PG-generated epubs in order to read them, I have slowly refined a set of styles which seem to work across at least a certain number of the dirt simple books. I don't kid myself that what I am currently doing is the best or most flexible possible answer. I didn't know HTML particularly well when I started, and I have already had one major re-think earlier this year when I went through at least two design iterations for the style sheet before eventually recognising that I was re-inventing the 'Decorator' design pattern from the book 'Design Patterns' by Gamma, Helm, Johnson & Vlissides. I've also had some very useful help from several contributors to this forum. (Anyone who is interested can see a fairly up-to-date version of the stylesheet I am using contained in http://www.mobileread.com/forums/showthread.php?p=1371707#post1371707 which is an Anthony Trollope novel "The Belton Estate".) I'd like to know if Stevan's books can teach me anything, so I would like to request that Stevan puts his books somewhere (maybe MobileRead, maybe he has a better place) and appends their URLs to this forum so that those of us who are interested in this problem have the opportunity to learn something. Bob Gibbins
participants (1)
-
Robert Gibbins