Beats me. All I can say is it works. If you open the file again in
photoshop the transparency is all NN grey. I guess that means you are
right. Interesting.

It is something like what happened when I made that lost perfect print for
the PP of A National Exhibit in Chicago, back in the good old days of 1952.
I was apprenticing under a Master, who was also the President of the Camera
Craftsmen ... very high altitude, you know. Everybody listened when he
spoke with his soft South Carolina accent.

But he was one of those old timers who had come up the ranks when there
were no photography courses in the schools ... not even at Rochester
Institute, in Kodak's home town. I, on the other hand, had been attending
the finest (and at that time only) university full degree curriculum in the
subject , at the University of Houston, and was filled to the brim with
theory and technique he had never heard of.

So in addition to making my perfect print for the exhibit, he and I had a
contest going between ourselves to make a really fine "High Key" (white on
white for you non photographic minds) print. I had my theory, and he had
his, and his was all wrong.

Mine was based on a complex series of unsharp masks to bend the curve the
way I wanted it (yes, there were unsharp masks before Photoshop ... in
fact that's where Photoshop got the idea ... they were made out of
photographic film ... but that's another story) ... anyway, mine was made
out of a bunch of unsharp masks. My theory was modern and impeccable.

His approach, on the other hand, was to overexpose the negative (thus
blocking up the highlights and taking all the sparkle out of them ... and
soak it in a pure Metol developer for a couple of hours at about 50 F ...
normal developer temperature was 68 degrees F ... which was guaranteed to
produce muddy gradations and probably stain the negative. There were two
different photo developing chemical compounds in black and white
photography then in common use (there were more, but these were the main
ones). Metol and Hydroquinone. Now, Metol was the one that gave you
unusably soft gradations that ended up with a gray and muddy print with
neither clear highlights nor deep shadow. All gray. No black, no white.
Hydroquinone, on the other hand, gave you an image with such a steep curve
that it was mainly solid black and solid white with very few steps in
between. Mixed together in varying proportions produced just about any
contrast you might want to have.

But he was going to use nothing but Metol, and not only that, he was going
to chill it down to where it quit working right, and then he was going to
overdevelop it by hours. It was totally against all the theory.

The funny thing was, we both got excellent prints. My high tech approach
and his I don't know what to call it approach, and both of them hung at the
Stevens Hotel in Chicago that fall.

So, you are probably right about the Gif 89a allowing only one color to be
transparent. It doesn't give you multiple transparent colors ... but there
is a little troll living under that bridge who handles it for those of us
who don't know that.