Anthony wrote:
On 7/4/06, Robert Scott Horning robert_horning@netzero.net wrote:
If you want to "publish" a book you've written that has Wikimedia project content, you can organize it and then submit it to this board for review. If the book is accepted for publication, it is somehow added to the WMF store.
I'd like to see this process eventually involve much more collaboration than it currently does. Right now the process seems to be that someone takes the content offline, builds a pdf, and then uploads it. Can't we make this a wiki process, where the pdf source code can be edited online directly, and then the pdf generated on demand? Sure, before the print run starts a stable version has to be chosen, but before that it'd be nice to be able to edit things. It's also a GFDL requirement to provide the source code. Is the source to [[Image:Big_Cats.pdf]] available somewhere?
Sorry to break this up into two replies, although this is two different topics.
The issue right now for editing the PDF file is mainly dealing with the markup of whatever system you are using to generate the PDF file. The process I've used in the past is to import the HTML into Open Office and do some minor fix-up of the content where HTML doesn't seem to do a good job for publication. This involves downloading higher resolution images and clearing out fluff that comes from HTML that looks good on-line but doesn't seem to work with a print format.
Some efforts have gone into streamlining this process, including using transclusion to make an entire Wikibook available as a single web page, and adding the <includeonly> <noinclude> markup tags at various places to help determine what is going to be in the final publication. Still, there is unfortunately a bit more hand labor to get the process accomplished than I would normally like, and there is some personal taste and flavor that goes into formatting a book. An automated HTML to PDF process may be possible, but I do question if HTML is really up to the task.
Essentially, this is an issue where there are people with editing/processing skills that are not developers, and part of the gulf that seperates the Wikimedia contributor/editors from the MediaWiki developers. Certainly an "automated" process would be desireable, but at the moment we are doing as best as we can with the tools currently available. Making a Wiki-syntax to PDF converter would be a non-trivial task that could be a whole major software project unto itself. I'm not convinced that the results from such a converter would look good either, or if it would look very amaturish simply because the content isn't designed to be displayed as a PDF for most of the Wikimedia project pages. Certainly PDF to HTML converters (like found on Google) look forced and don't seem to work out very well either, although in that situation it is mainly to read the content rather than having something "pretty" to look at.
BTW, if you want the "source" to Big_Cats.pdf, the full downloadable version is at: http://en.wikibooks.org/wiki/Wikijunior_Big_Cats/Complete_Edition
It is from this version that the pdf was created, although admittedly with some minor tweaks. It is debateable if this conforms to the GFDL completely, but all of the original source material is there in machine readable form. It at least complies with the spirit of the GFDL and we are doing as best as we can. Perhaps the original *.swx files need to be uploaded as well, but any major update is going to need to come from the original source HTML anyway. This is an experiment just to test the waters and see what direction we really need to head, so certainly any criticisms of what is going on are welcome.
Criticisms of creating the PDF and the "source files" can also be made to [[b:en:Image:Wikijunior bigcats frontpag.jpg]], which is also published under the GFDL but the "source" layers are not available for manipulation. I see these two issues as essentially the same thing.