From: Chris Lilley <chris@w3.org>

Date: Fri, 02 Mar 2001 21:28:50 +0100

Message-ID: <3AA00282.BF413C4A@w3.org>

To: Amit Jain <ajain@ptc.com>

CC: W3 SVG Mailing list <www-svg@w3.org>, W3 MathML Mailing List <www-math@w3.org>

Date: Fri, 02 Mar 2001 21:28:50 +0100

Message-ID: <3AA00282.BF413C4A@w3.org>

To: Amit Jain <ajain@ptc.com>

CC: W3 SVG Mailing list <www-svg@w3.org>, W3 MathML Mailing List <www-math@w3.org>

> Amit Jain wrote: > Now the conundrums: > I am thinking that since it is the same data that is presented in two > different ways, I can modify my program to output an XML file (using a > private schema) and I can convert this with XSLT to either HTML, MathML > or SVG. But I haven't ever coded XML before, so I want to know if all > this is feasible and whether the technologies are mature enough to be > able to cope with all this. The answer is yes and no. Yes you can output your calculations in XML, yes you can have the numerical results in XML and yes you can display the drawing in XML (using SVG). The trick is whether you can convert whatever your calculations outpiut into an engineering drawing; there the limitations are not XML but the processing technology that you are using for the conversion and how automatically thje geometry follows from the calculations. > And what would the best platforms to > parse/display/print the XML be. I have some misgivings about whether SVG > provides enough functionality to cope with non-trivial drawings that > involve scaling/rotation/translation of entities SVG can easily deal with translation, rotation and scaling of graphical elements. This is extremely basic functionality. > (I read an SVG tutorial > at IBM's website, and they said that they can't use SVG to generate SVG does not generate things; it is what something generates. > even > a simple pie chart from 3 data sets as it doesn't support mathematical > functions). This is bogus; SVG can clearly be used to draw a pie chart! SVG with script can be used to draw a pie chart on the fly, and there is an example in the SVG Test Suite which does exactly that. The issue is not with SVG itself but with whatever technology you are using to convert from your calculations to SVG. For example, your calculations ight give the wall thickness at certain discrete points and you would then need to calculate a smooth curve that passes through those points while obeying whatever criteria (maximum rate of change, minimum total weight, whatever) that you need. Having descibed that curve, for example as a series of continuous bezier segments, it can then easily be output to SVG. In the case of a full programming language (Java, C, ECMAScript) any DOM manipulation can be done and it is just a case of whether you know how (or have libraries that know how) to calculate your desired geometry; once you have the geometry then SVG can express it. In the case of XSL-T for example, you will have a harder time (but help is available, take a look at what SchemaSoft has done in the area of math libraries for XSL-T). > Also, I don't know a priori what size of paper I would be > printing on. You don't need to. You just give the aspect ratio and the file prints as big as it can. > Is it possible to query the display size from whichever > application is going to display SVG/HTML/MathML and format my output > accordingly ? Oh, and finally, does MathML support mixed mode output, > like: > > Solution to this quadratic equation = [formula here] = 5.4 That is a mix of MathML and some other grammar that can store text, which can be pretty much anything. -- ChrisReceived on Friday, 2 March 2001 15:29:00 GMT

*
This archive was generated by hypermail 2.2.0+W3C-0.50
: Saturday, 20 February 2010 06:12:50 GMT
*