Hello!

On Mon, Mar 05, 2001 at 08:11:41PM -0000, keth@... wrote:
> [...]

> But I note the page numbers are missing from the GIF's.
> For example, in my copy, the last page "ÖXL ÖXUL-TRÉ"
> has the PAGE NUMBER "[532]" right in the middle above
> the column divide line.

> Page numbers are nice if you want to use the dictionary
> as reference. Also, the fact that the GIF's are black
> will cause ugly printouts, if you want to take out a page
> or two on a printer. (e.g.for ease of proofreading)
> But maybe printers can be told to print reverse images?
> At any rate, I think the default ought to be b on w.

> The images are also a bit wide for my screen. My impression
> is that a size reduction would not hurt. But probably the
> present size is an advantage for the OCR process (?)

> In any case, I think you already have a workable dictionary
> on-line, which deserves every ones compliments!

I did some conversion here and now, using a Unix box with some
free utilities.

From that came this:

first: make png's from the gifs:

for i in *.gif; do gif2png -n -s -O $i; done

second: make EPS (encapsulated PostScript) from that:

for i in *.png; do j=`basename $i .png`; pngtopnm < $i |
pnminvert | ppmtopgm |
pnmtops -noturn -dpi 600 -width 7 -height 11 -center | bzip2 > $j.eps.bz2
done

(the "bzip2" compresses the output to save disk space,
pnminvert makes black-on-white of the originals)

That step takes quite some time (half a minute per page on my box).

What I haven't done (yet?) is a small script to add page numbers.
Perhaps I could do something with TeX...

Í friði,

Hannah.