SVG and Unicode

Stuart Thiessen sw at PASSITONSERVICES.ORG
Thu May 3 17:27:16 UTC 2007

The Korean alphabet does not use XY coordinates for its construction  
of words. My understanding is that there are specific rules for how  
the characters combine and so those rules are built into the  
rendering system.  Sometimes, like for Chinese, there are a limited  
set of helper characters that can be included in the Unicode stream  
that indicates how the various characters combine. The point being  
that spoken language writing systems are very restricted in the way  
their letters combine. Even Chinese has very limited sets of how they  
can combine their characters into a new character.

SignWriting, on the other hand, is much more free in how various  
symbols combine. Part of my research will involve studying what  
restrictions are present in the system (for example, can a contact  
symbol connect with an arrow? Can a face and an arm symbol interact?   
etc.) Once we know what restrictions are present in how symbols  
interact with each other and where symbols do tend to connect and  
where symbols do not connect, we can begin to develop rules for  
SignWriting with Unicode. Remember, for Unicode, we have to consider  
how to take a 1D text string and represent it as a 2D SignWriting  
word. So we have to consider what needs to be present to make that  
transition successful so that the renderer knows how to select the  
appropriate symbols and display them properly.

There are alternatives to XY coordinates and that will be what I will  
research to see if it will work.



On May 3, 2007, at 11:56, Charles Butler wrote:

> I am curious as to why XY coordinates would not be well received  
> with the Korean alphabet and its construction of words included in  
> the corpus.  That writing system requires every letter having its  
> forms and space locatives available.
> Arabian script has endings that have to be attached or moved  
> depending on the appearance of the word.
> When one works with Chinese, if one were working with the 32  
> possible brush strokes to construct a new symbol, one needs XY  
> coordinates.
> Same thing for Cree, apparently based on Gregg shorthand.
> Charles Butler
> Stuart Thiessen <sw at> wrote:
> See my comments below ...
> Stuart
> On May 1, 2007, at 9:10, Steve Slevinski wrote:
> > Unicode
> > -----------------------
> > MovementWriting in general needs a Unicode platter of 56 thousand
> > symbol places. Each symbol should get it's own space, or be
> > logically combined some how. It's going to be political to get a
> > new platter opened, but the existing 256 spaces in a nearly full
> > platter are not enough.
> My contacts who work with non-Roman scripts assure me that there is
> no problem with getting space for SignWriting. That is not the main
> issuel. The biggest issue is deciding what should appear in the slots
> and what should be handled by the renderer. In addition to that, we
> will have to convince them of our method of rendering SignWriting
> (whatever that ends up being). I strongly believe we will have to do
> some unique things to make SignWriting work with Unicode compared to
> the way most spoken languages are rendered in Unicode today. So that
> will be a greater issue than the number of symbols. My conversations
> with my contacts indicate that x/y coordinates may not be well
> received. So I am exploring other alternatives to x/y coordinates for
> symbol placement. Also, whether it is on existing planes or a new
> plane, we will have to see what is available when our proposal is
> ready. My initial efforts will be focused on SignWriting proper. But
> I agree that a strong case exists for bringing in all of the Movement
> writing symbols.
> >
> > Unicode and SVG
> > -----------------------
> > Once each symbol has a unique number, we can create font files for
> > SignWriting. We may be able to have a single font file, or we may
> > need multiple font files. Since we'll already have the SVGs
> > completed, we can use the SVG style fonts. This whole process will
> > be repackaging. There should be automated tools that can help.
> > Maybe even font experts..
> I believe that it should be no problem to contain all the symbols in
> one font file. After all, they have font files with nearly every
> writing system included that are several megabytes in size. It may be
> that our font file has very minimal support for spoken languages and
> only contains movement writing symbols.
> Thanks,
> Stuart

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the Sw-l mailing list