Re: VMs: Image Source, Accuracy of Transcriptions

[illumin8@xxxxxxxxxxxxxx]

Larry Roux wrote:

> Now, while I agree that many "ii"s in EVA are a single character, Courier has its' own
> problems.
> 
> For instance, EVA "ckh" in Courier is simply "x".  That is losing information.  and "il" is
> just G?  iil is H.  WAY too much data is lost.

Not really. If you had the rule that "qu" ALWAYS went together and just
showed a "q" or a "u" you could extract the data for the accompanying
paired letter. Same with EVA/iin/ always equals CUR[M].


> While EVA breaks characters up too much at
> times it is FAR better than Courier.

Far better for a transcription that actually LOOKS like the VMs - No. I
think that honor goes to Frogguy. Far better for a transcription that
describes linkages - Yes. It depends on what kind of test you were
running or the transliteration you were proposing. If I transcribed the
English "d" as "c+l" (counterclockwise curve + straight ascender) I
would only be wrong if I were trying to pronounce it versus trying to
describe the shape. 
All we can do with the VMs is try to describe the shapes and hope that
someday we can find out how the letters relate to each other and maybe
pronounce them if they are language and how to manipulate them if they
are cipher.


> But we could argue this till the cows come home.  Until someone decodes the manuscript noone
> can guess exactly what the character set really is.  

Watson, I believe you've got it.


> Noone should feel bad about WHATEVER character set they use.

Here here. As long as they realize that what kind of test they perform
MAY be affected by their transcription scheme.


illumin8

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list