RE: VMs: Image Source, Accuracy of Transcriptions

["Larry Roux" <LRoux@xxxxxxx>]

"Oh, and the guy's name was "Currier" (ie, as in neither the font nor the messenger, so don't shoot the latter)."  
Oops!

"Easy there, tiger - dishing out flame-bait like that will leave you with a thumping headache in the morning. Be nice, now" 
Agreed.  The reason I joined this list was to share ideas/gain 'insight' not to 'incite'.  Truce called.

"We also have to consider why, if the VMs glyphs are so clear, glyph-level statistics don't appear to be moving us forward towards any kind of consensus on even the most basic issues."  
This makes me wonder if we need to come up with a <gasp> new font.  One where most of us can make concensus as to what logical glyphs are.  Of course that could cause more trouble than it is worth too.

I believe there is a LOT of good work going on in this list.  Whether I agree with any of your ideas or not I respect them all.  After all, anyone who takes the time and effort to try and find something new is contributing to the overall effort.  If the idea leads to a blind alley, then so be it.  




******************************
Larry Roux
Syracuse University
lroux@xxxxxxx
*******************************

>>> incoming@xxxxxxxxxxxxxxxxx 08/30/03 10:43AM >>>
Hi Larry,

At 10:15 30/08/2003 -0400, Larry Roux wrote:
>I am talking about statistical analysis here.  If you took a sample of my 
>handwriting you would find a glyph "this" which appears as one item but is 
>actually 4.
>
>This is not a printed work, so yes, we have to guess as to what 
>constitutes a single glyph.

We also have to consider why, if the VMs glyphs are so clear, glyph-level 
statistics don't appear to be moving us forward towards any kind of 
consensus on even the most basic issues.

>I would guess that you believe all word breaks are spaces too.  Have you 
>actually SEEN any old documents?  Look at latin.  Characters and words run 
>together.  "t" runs into letters.  "cr" can look a lot like eva "ch"  So 
>can "ti"  and both "ni" and "ui" can look like eva "iii" but in Courier it 
>is one glyph.

*sigh* Look at Quattrocento humanist documents and you'll see the general 
family the VMs comes from (OK, "or was hoaxed to look as if it came from" 
*sigh*). Lumping together 1000 years of Latin texts under one stylistic 
generalisation isn't going to help us here. :-(

>For statistics Courier is absolutely NO good.  EVA aint great either.  But 
>at least it breaks apart some glyphs into their OBVIOUS components where 
>Courier is just plain stupid in that regard.

EVA isn't for statistics, it's for regexp'ing into your transcription of 
choice (and what a choice it is). And, unless you know differently, we 
haven't got any obvious (semantic) components yet. Oh, and the guy's name 
was "Currier" (ie, as in neither the font nor the messenger, so don't shoot 
the latter).

>But, hey, you have all the answers, don't you?  Oh, wait, no you don't - 
>you just have criticism.

Easy there, tiger - dishing out flame-bait like that will leave you with a 
thumping headache in the morning. Be nice, now. :-o

Cheers, ....Nick Pelling..... 


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

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