
I agree with Carlo: TEI indicates mostly the plain _structure_ of the text, without any formatting. You will have to add that in a later stage (and interpret the structure from the formatting when digitizing an existing book), and also, I do not use PG-TEI. I typically add rend-attributes to my TEI files, to help the conversion to HTML and ePub, and as a result get things like: http://www.gutenberg.org/files/38748/38748-h/38748-h.htm http://www.gutenberg.org/files/38571/38571-h/38571-h.htm (Just a few of hundreds of books I have produced using TEI. Don't look at the ePub's, they are not mine; I can provide them if needed from my tool-set, but the PG process doesn't allow them in.) That HTML is reasonable clean, and works on most browsers. By the way, my tooling allows me to adjust the formatting by use of CSS to almost anything, including making the book look like the original in a high degree. It is not TEI to blame, it is the misunderstanding of what it is and how to use it that causes trouble... Jeroen. On 2012-02-05 08:33, Carlo Traverso wrote:
I don't think that one might adopt TEI for DP formatting, (sometimes too verbose), and DP markup cannot be directly transformed in TEI (the DP markup is ambiguous, it has too many "This will draw the attention of the PPer"). But making DP markup a bit more precise (e.g. adopting different markup for center, flush right and flush left insetad of /*...*/) DP markup might be easily transformed into elementary TEI. There are plans to have guiprep generate TEI output (instead/in addition to HTML).
And TEI is simpler than HTML, since it is an intermediate step, that just marks the structure, and leaves the display to a next step. So you don't have to care to the bugs of IE 6 or firefox 4 beta. TEI is not meant for direct viewing in browsers, although some styles of TEI can be visualized with some standard XSLT. For example, some OTA TEI files display as unformatted text since there is no display engine. Just look at the source, either in a browser or in a text editor.
Of course the difficulty is only moved to the next step. And it is what Marcello's tools can do (as epubmaker does for RST). And PG-TEI is crippled in the same way as PG-RST is crippled (RST in itself is extensible and can produce much better results, but these extensions are not supported by epubmaker).
So people really wanting to do a service to PG should propose a different toolchain for PG TEI submissions, better configurable and giving better end results.
By the way, where are the source fot the TEI toolchain? I coundn't find a mention of them on PG, (searching the site for TEI returns nothing).
Carlo _______________________________________________ gutvol-d mailing list gutvol-d@lists.pglaf.org http://lists.pglaf.org/mailman/listinfo/gutvol-d