Piotr Kopczyn'ski wrote:
Now to the matter.
I haven't got a clear idea what to do, I'm not even sure about the guidelines for this task nor have a complete feeling of the goal. On the other hand I have some chances to make a piece of valuable work here.
The task was described as "likely XSLT". Fortunately I have the basic understanding of XML and XSLT (I've recently accomplished a subject at my university devoted to XML and XSLT, so I'm quite fresh with it). In addition, I got to like it and feel a need to deal more with such translations.
The fact is, I don't know the "XML Schemas" and practical aspects of SVG and KML. I will need some time to get through their documentations. Anyway, both specifications are worth knowing, so I will gladly learn them.
I guess the job may be considered as import and export respectively. Importing (KML to SVG) shall be relatively simple and with a small risk of an "out-of-time" defeat. The other direction (export) seems more hard and needing deeper considerations. Of course, the importing part might also need some "parameters" (options in import wizard?).
As I am freaky/scarry of not meeting the deadline, I like to find a safe-feeling solution first. An example of such is making a little website for converting KML to SVG using for example Firefox built-in XSL processor (I've used this solution in my last XSLT project, although it didn't involve any SVG rendering). I have then made a silly and tiny web framework (using PHP + .htaccess + mod_rewrite) for passing parameters to the XSLT file (parameters are dynamically inserted into original file as lines with an xsl:param definition each). Thus I could make a small portal driven only by XSLT :-) As this solution is quite comfortable and "recent" for me, I believe "I can make it", and in the meantime we can optionally consider if I will be able to deal with Inkscape internals (that is what I don't know at all, so this may be more risky).
That's it for a minimum. I'm waiting for your opinion and expectations. If you agree for solving this task via XSLT, I'm likely to success with it.
At this point I would say the most important thing is to get your application submitted to google via the SoC web application. It sounds to me that you know more about what you need to do than you give yourself credit for. To get a little extra confidence read through the previous years' proposals on the wiki to get an idea of the form and style. You'll want to include research in your timeline. I doubt you will need to work too closely with the inkscape internals if your project is primarily to be carried out in XSLT.
Good luck!
Aaron Spike