Symbol consistancy?

Valerie Sutton sutton at SIGNWRITING.ORG
Wed Aug 23 22:25:59 UTC 2006


SignWriting List
August 23, 2006

Charles Butler wrote:
> I guess I am confused.  I was under the impression that in  
> building, for example, a Sign Symbol Sequence you'd be picking an  
> exact handshape and palm facing with only one possible code per  
> handshape/palm facing, not a mirror image of a possible shape.  If  
> one sorts by handshape, how can one be certain one is getting the  
> handshape AND rotation one has requested if the system doesn't  
> contain the coding for a particular handshape, hand (right or left)  
> and rotation. I guess that it is the encoding system for the  
> computer program that I'm curious about.  If it is built by a  
> drawing system, then conceivably you'd only need one handshape per  
> handshape, the rest would be flips, mirrors, mathematical  
> rotations, and fill ins.

The encoding of the SSS works beautifully as we have it right now,  
even with this issue of 10 palm facings. First, using the Mirror Key  
does give you the proper SSS ID number for the flopped version of  
it...so using those keys still gives you ID numbers. Second, since we  
had no choice but to have 6 palm facings and not 10 palm facings, in  
the SSS for computers, I chose to include the most commonly used palm  
facings out of the 10...I went through and counted how many times we  
had used some of the 10 palm facings that are not there, and it was  
less than 1 per cent of the time...very very rare...so it would take  
a huge amount of signs using that one palm facing in the dictionary,  
before there would be any conflict and we are talking about so many  
other variables within the spelling of a sign that the chances of  
this hitting us are slight...maybe after we have sorted dictionaries  
with 10,000 entries by SSS...not by spoken languages...maybe by that  
time we will know more about this issue...if it is really necessary  
for us to expand to 10 palm facings, that will be a great detriment  
to SignWriting software development because there will be too many  
symbols...so there may be another technical way to tweak this issue  
when it really arises...

So far there are only two computer programs that can produce sorting  
dictionaries by SSS using the IMWA as it is designed right now...that  
is SignBank8 and SignPuddle 1.5 or later...right now we are all using  
SignPuddle 1.0...but when Steve comes out with 1.5 you will be able  
to play with some of this to get your own answers!!

Remember there is already an alpha test version of 1.5 on the  
web...and you already played with the spelling column a little,  
Charles...as you know, the editor of the dictionary can tweak each  
spelling manually to make it sort properly...so there is flexibility  
with it all...so if you find such a glitch, then you can rectify it  
yourself in the Spelling Column...when we have 1.5 released I will be  
teaching everyone about SignSpellings for sorting dictionaries...

So far I have only sorted by SSS with around 500 to 1000 signs, so I  
do not have enough experience to know about some of the glitches  
myself...it is a real adventure and I look forward to the challenge  
of sorting large dictionaries by SSS!!

No matter what, for the first time in history, there is software to  
create dictionaries sorted by SSS and that is an accomplishment...

Thanks for your input, Charles -

Val ;-)



More information about the Sw-l mailing list