VMs: RE: Folio 39r

["GC" <glenclaston@xxxxxxxxx>]

Nick wrote:

> To make it easier to proof as is, precede each line of
> text with its line
> number. Preceding each word with its word index within
> the line (in a faint
> colour and smaller font) would also help. Determining
> an error's position
> took just as much time as finding it in the first place. :-)

I agree with this, and have given it much consideration.  Line
numbers are relatively standard in all transcriptions.  Word
separations are relatively arbitrary, as a quick glance at the
interlinear suggests, so I have been hesitant to rely on "word" as
a unit for some time.

One thing that tends to create this disagreement are what have
been termed "half spaces", which I also call "half-pauses".  These
appear to be an artifact of compilation, where writing is set to
paper in sets of three to four glyphs most often.  This could be
caused by anything from pen/ink capacity or compilation of the
cipher itself.  I also confirmed this in readings of several
lines, one of which read "Nou embre sixteen".  The particular
location I can't recall right now, (but I did report it in the
archives somewhere).  I remember checking this passage against the
interlinear, and all transcribers were in agreement that the space
in "Nouembre" was a word break.  This indicated to me that word
"spaces" were not always where they should be in normal language,
and because of this any statistic based on word length would serve
only the most general of theories.

> One step up from there: PHP could generate (from your
> database) simple
> web-pages based on forms, where each word is preceded
> by a named checkbox,
> whose checkedness could be made to persist in an online
> database. This
> would allow multiple listees to flag changes independently.
>
> That kind of mechanism wouldn't be browser-specific at
> all, so just about
> everyone should be comfortable with it. :-)

Please tell me more about this, as PHP is a term I'm not familiar
with.  In private would probably be better, as this may be
considered off-topic for the list.

> For "security" (such as it is on the net), these pages
> could be placed in a
> password-protected directory... much more than that
> would probably be
> overkill. I doubt we'd get any script kiddies in this group. :-)

Passworded directories we certainly have!  A linked on-line
database file that accesses by glyph or "string" and is linked to
images of each line of text of the VMS can be kept to around 10mb,
depending on the number of information fields, but if images are
stored directly in the database and not linked, download would
increase to somewhere between 60 and 80mb, outside the range of
many of our friends.  The goal was to make it on-line accessible,
quick search, with rapid field display on a 48k modem.  There are
a multitude of alternatives, but most of them are outside my
technical expertise.

I will currently entertain any suggestions concerning an on-line
database, what we should expect of it, whether it should include
alternative transcriptions, etc.  I already have the interlinear
translated to dBase, a nearly universal platform.  (That's the old
interlinear, not the one that represents Currier, 1SSG, 2SSG,
D'Imperio, etc., in EVA instead of their original transcription
alphabets - they were all primarily glyph-based transcriptions.)

Any suggestions on what you'd like to see?  While I'm doing this
we may as well make it something that suits the majority of needs,
even those who prefer EVA.  I warn you that increasing the scope
of the project is a little more than I wanted to take on for
private use, but with a little help on aspects beyond my immediate
concern, this could be a very good tool for everyone.  The one
thing I'm sincerely lacking is a databased version - by glyph - of
an entire EVA transcription.  Any takers?

GC