stevan,
> From: Stevan Harnad <harnad@ecs.soton.ac.uk>
> Date: Mon, 24 Jul 95 20:23:03 +0100
> To: ginsparg@qfwfq.lanl.gov
> Subject: Re: Ginsparg funding
>
> What did you think of the Entomology journal's different tack
> to going hybrid: I always criticised hybridizing on the subscriber
> model, but bifurcating into the author page-charge model for the
> electro-version (and, ingeniously, dubbing it an investment in
> reducing reprint mailing costs!) sounds like a nice twist. But
> I'm worried about instability down the road, as the paper version
> withers up...
well if i understand their arithmetic the whole thing works out to under
$50k / year gross receipts. already low budget, though presumably either
largely volunteer or otherwise subsidized. you'd have to find out exactly
what that's supporting to determine how they'd manage at 30% = $15k / year
(getting closer to psycoloquy).
> And I'm sure you'll have some words about the Adobe viewer (which
> I couldn't get to work for me).
well we've had that here for a couple years now (lanl was a beta test site).
though never any trouble getting it to work, have had various comments for them
all along. (for a summary, see http://xxx.lanl.gov/hypertex/pdfcomments.txt
from last fall)
the adobe people are well-intentioned, but like most computer people
uni-dimensional. in this case it meant that we couldn't impress upon them
the importance of network hypertext two years ago (just as we couldn't
impress same upon the tex people, nor the importance of mathematical
typesetting to html or www people; hence the little hypertex/hyperps project
designed to be downwardly compatible to pdf 2.0).
their other problem is that they never imagined contextual markup, instead
only page markup. specifically that means we lose contextual information
through the hyperps -> pdf distiller (i.e. a universal internal ref to
"eqn.2.1" or "ref.14" in our hyperps gets translated to xy coords on page n).
but the next version of netscape (no longer free) due out this fall is alleged
to have built-in pdf support (this to satisfy e.g. the advertising community
that has been crippled by html's primitive page markup capabilities).
adobe has made noises about being sgml compatible in a future version, but
that doesn't really mean anything in the absence of a specific dtd
(document type descriptor). finally the adobe people valiantly tried to
bring out pc/mac/unix versions simultaneously, but in practice that meant
all versions are dumbed down to the level of the windoze version with its
hopeless gui. so the hyperghostview we've gotten out of the gnu people
(by supplying the correct pieces of modular code) is a vast improvement,
but adobe/netscape should catch up eventually (though both groups
could still benefit by listening to input from network architects, since
they're seemingly oblivious to certain standards and protocols that could
ease the coming burden on certain international networks -- rob fought
with the netscape hacks for three months last fall to get the
gunzip/uncompress properly implemented in the unix version via
Content-encoding: in the mime header, and then had to fight the same battle
for the mac/pc versions this spring, but the latter will only be corrected
in the for-sale versions this fall...)
pg
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:08 GMT