Unicode Discussion

[Stuart Thiessen]

Well, part of my reason for using polar coordinates was because it limits one set of numbers to at least 360 and possibly even less than that depending on how specific the angle must be.  The distance certainly might be bigger perhaps, but even then if it is 48 units like the SW box, then that is still managable.

It seems from your email that you are thinking like we will do the rendering from within the font?  The project I will work on be using a rendering engine called Graphite.  It will take care of all the rendering provided we supply it with sufficient information.  Some things may be predictable and those we can build into the script that Graphite uses to render SW.  Those that are not predictable, we will need to have "invisible" symbols that provide the appropriate hints to the rendering engine.

In general, (x,y) coordinates do tend to be more difficult to scale simply because you do have to take into consideration the x/y ratio of your output device.  If I scale something on my device, unless your x/y ratio is the same as mine, your output may look different. However, if we encode it using polar coordinates and allow each output device to convert the polar coordinates to the appropriate (x,y) coordinates for its x/y ratio, then we will have cleaner output and less dependence on a particular x/y ratio.

Now, the scaling issue may be more important for a graphical approach than for a unicode approach, but since we are only at the beginning stages of our work on this, I haven't had a chance to test anything directly yet, but this is my starting approach. Certainly SWML and other developments will be a strong influence to how I will consider things.

Thanks,

Stuart

> -----Original Message-----
> From: SignWriting List [mailto:SW-L at ADMIN.HUMBERC.ON.CA] On
> Behalf Of Valerie Sutton
> Sent: Friday, July 11, 2003 12:37
> To: SW-L at ADMIN.HUMBERC.ON.CA
> Subject: Daniel Noelpp: Re: Unicode Discussion
>
>
> Subject: Fwd: Re: Unicode Discussion
>
> Hello Valerie, hello friends
>
> Today the SignWriting mailing list suddenly stopped accepting
> e-mails from me. I am trying again and if it didn't work
> out, Valerie will forward them for me.
>
> I am sorry!
> Daniel
>
> --- Weitergeleitete Nachricht / Forwarded Message ---
> Date: Fri, 11 Jul 2003 13:37:27 +0200 (MEST)
> From: Daniel Noelpp <d.noelpp at gmx.ch>
> To: SignWriting List <SW-L at ADMIN.HUMBERC.ON.CA>
> Subject: Re: Unicode Discussion
>
> > Hello Stuart, hello Ingvild
> >
> > IMHO polar coordinates is not so good an idea. Scaling is not
> > complicated at all using x and y coordinates, and I just had
> > a simple idea:
> >
> > "Position symbols" are special space symbols. They push the
> > following symbol to the right and vertically. For overlapping
> > symbols a "return to the origin" symbol (or something similar)
> > is neccessary, of course.
> >
> > This is quite a simple idea and scaling works automatically
> > this way because the spaces scale as every symbol does.
> >
> > So, why use polar coordinates, if x and y coordinates using
> > space symbols scale without problems? Space symbols using
> > polar coordinates don't work very well.
> >
> > Daniel
> >