Re: Testing of current CVS version, problems in svg in sketch and eps font problem, proposal for an RDF config file

[Lars Clausen <lrclause cs uiuc edu>]

On Mon, 15 Jul 2002, James Michael DuPont wrote:
> 
> --- Lars Clausen <lrclause@cs.uiuc.edu> wrote:
>> On Sat, 13 Jul 2002, James Michael DuPont wrote:
>> 
>> > The SVG output in sketch is microscopic,
>> > the font information was not recognized. The lines were lost.
>> 
>> I don't know what SVG does for font support.  We currently output the
>> font
>> name as Pango tells us, which unfortunately doesn't always match the
>> PostScript font name.
> Ok, I guess we will have to tweak the fonts via some fontmap table?
> What SVG viewers do you want to support? I can try sketch.
> It would be nice to start presenting a list of the interfaces to other
> systems, what programs have been tested and what bugs are there in
> them.

I don't want to support any SVG viewers in particular; I want to support
SVG.  That's what standards is all about.  If a viewer doesn't display it
correctly, I'll check if I follow the standards, and if I do, the viewer is
at fault.

>> > or draw a line from the actor to an object without using multiple
>> line
>> > segments.  
>> 
>> I don't understand this.
> Actor --line1-> usecase --lin2--> object
> two line segements

Well, looking at for instance http://www.rff.com/UML%20Software%20Design.gif
or http://digilander.libero.it/domenikov/grafica/Use%20case.gif,
it's not like use cases always go between two objects.  I think mixing the
lines and objects into one would be a bad idea.

>> > In fact there is not any use case line symbol in the uml
>> > toobox, but the normal lines work.
>> 
>> Indeed there isn't.  How would such a line look?
> I think you would draw a use case line between an actor and another
> thing. The use case would be in the middle automatically, and could be
> moved after the line is hooked up.

There is a use case symbol.  I don't see how use case lines differ from
standard lines.  Mixing lines and objects would only be confusing.

>> I don't quite get this, either, but then I'm not familiar with the
>> standards of use case diagrams.
> I have been using : http://www.informatik.fh-luebeck.de/~st/UML/UML1.1/
> http://www.informatik.fh-luebeck.de/~st/UML/UML1.1/notation6.html#figure33
> 
> But if you google image search on UML use case
> http://images.google.com/images?num=100&hl=en&lr=&ie=ISO-8859-1&newwindow=1&safe=off&q=use+case+UML&sa=N&tab=gi
> you get a bunch of hits.

I see them.  Perfectly doable with current shapes.

>> > On this note, I would like to also add in besides the URL, 
>> > a list of RDF like properties. These would be definable using a RDF
>> > schema and then converted into DIA properties. I think that either
>> > XMLLIB could be used to parse the XML and then the properties could
>> be
>> > created on the fly as needed. Otherwise I can start looking for a
>> GPLed
>> > RDF lib that we can put on top of the libxml doms if there is none.

>From the looks of it, I doubt libxml can parse RDF.  Too much special
syntax. 

>> -Lars is drowning in TLAs.
> Sorry I have be reading up on all this stuff: here is a set of links

I'll have to look some more into this.  

Notice that since RDF has no layout to it, we can't make Dia read it
without adding some kind of auto-layout.

-Lars

-- 
Lars Clausen (http://shasta.cs.uiuc.edu/~lrclause)| Hårdgrim of Numenor
"I do not agree with a word that you say, but I   |----------------------------
will defend to the death your right to say it."   | Where are we going, and
    --Evelyn Beatrice Hall paraphrasing Voltaire  | what's with the handbasket?