james said:
>   I have created page images 600 pixels wide using Scan Tailor and
>   the original much larger page scans done by the Internet Archive.

you did great.

but i screwed up.

in two ways.

first, i forgot to tell you to trim the margins to a tiny slice.

second, i made a math error.  the size we're aiming at is
for the _width_ to fit into the screen(s) you will be using...
that's _after_ the margin-trim.  so these are too reduced.

i'd have to blow up your images now, which is not good.

it'd be much better for you to re-do what you did, since
you will be downsizing the original larger-sized images.

you'll see the reason for this when you look at the app
i coded that lets you step through and check the names.

>   http://zenmagiclove.com/bhaga/scrabble.py

it's not finished -- doesn't record your decisions yet --
but you will get the idea.  for each word, the app shows
the first 3 instances (with pagenumber and linenumber),
along with the scan for the first appearance, to review...

so if the diacritics are right, you click the "good" button.
if not, you click the "bad" button.  go ahead and try it...
(since it doesn't record your data, you can click freely.)

whether you will want to use this app or not is a choice
that you can make after we discuss the various issues...

you might decide to attack the tasks in a different way.
i just wanted to show you one way that you _can_ do it.

-bowerbird