xmlhack: Insatiable SVG

If you think of SVG as a toy technology to draw nice pictures, wake up! SVG is invading your cell phone and this "graphical XML" might wipe out "text XML" (such as XHTML and XSL-FO), just as graphical user interfaces have wiped out text-based user interfaces.

SVG has been one of the most-talked-about technologies during XML Europe 2002, with four presentations and a countless number of references in other presentations and during the experts panel.

Some key facts which have drawn my attention are:

From the presentation by Dean Jackson, the number of mobile phone manufacturers and software editors supporting the SVG 1.1 "basic" and "tiny" subset showing that SVG will soon invade our cell phones and PDAs, and the fact that SVG 1.2 will include text-wrapping functions making it more suitable for applications with many text objects than it is currently.
From the presentation by Benjamin Jung, the comparison between a raster image and a "badly designed HTML page where content, transformation and presentation are badly mixed" and the ability given by SVG to separate these three components.
From the first presentation by Antoine Quint that Flash is similar in this aspect to raster images and that SVG + SMIL enables you to cleanly separate content and presentation from the animation.
From the problems that Antoine Quint had matching the resolution of his laptop with the video projector, the benefit of using SVG (which scales so perfectly) as a rendering format for presentation.
From the second presentation by Antoine Quint that the SVG DOM appears to achieve a much better degree of compatibility than (X)HTML DOMs have ever achieved, that the SVG extensions give you access (in addition to all the XML objects) to SVG and CSS objects, and that you could implement XForms and even a full graphical user interface using SVG and its DOM.
From an answer by Leigh Dodds during the experts panel that SVG could be used -and was being used- to render pages for printing in lieu of XSL-FO.

Putting all this together, one may wonder if SVG is not the first real application of XML as "SGML on the Web".

Other stories:

The C++ project is the second stage of http://sourceforge.net/projects/xslfo-proc. Myself and Eric just happen not to be hosting actual development on Sourceforge anymore, but I will intermittently provide updates at the Sourceforge location. There will be more visibility in the near future - right now we are putting a lot of time in getting the foundations in place.

From January until early this month I was doing intensive Perl prototyping work; I concluded several weeks ago. We are now well underway with the C++ production formatter.

It'll be a fairly conventional XSL formatter; the prototype gave me some good guidance in this regard. I like the way things are moving right now and I think it'll turn out to be a good, reliable, fast formatter. For me another big thing is writing the SWIG interfaces for the final formatting library. We also have it in mind to produce an executable.

Currently we are building on both Win2K and Linux. The code should be very platform-agnostic in any case. We are using libxml and pdflib at the present moment.

Based on our current rate of progress I'd guess that early fall is not unreasonable as a target for a reasonably functional system.

I agree that with Arved that SVG lacks all the formatting that XSL-FO provides. However, I can see that SVG is sexier and easier to approach than XSL-FO. XSL-FO is far from widely understood at this point. Notwithstanding Arved's fine contributions on the FOP project at xml.apache.org...

I would see SVG and XSL-FO as being very complementary. Myself and Eric Bischoff are planning to use an SVG subset as the data interface between the formatter and the renderers/viewers in our C++ xslfo-proc XSL-FO project. It is a very nice fit.

I am not an expert on SVG but in reviewing the spec I fail to see a formatting model. How would one express the concept of footnotes, or side floats, or markers, or keeps and breaks, using SVG?

SVG "in lieu of" XSL-FO? Hardly. As one possible mechanism for describing the formatted result, sure.