1995-10-03 - Re: Introducing Johnson-Grace

Header Data

From: mpd@netcom.com (Mike Duvos)
To: cypherpunks@toad.com
Message Hash: a2d3cee2d1efd24f9a0f12621bc0dc2f0fc5c439207d028d9174bbbf3f6c3e2a
Message ID: <199510030602.XAA18317@netcom17.netcom.com>
Reply To: <95Oct2.185011pdt.21953@mail.jgc.com>
UTC Datetime: 1995-10-03 06:02:32 UTC
Raw Date: Mon, 2 Oct 95 23:02:32 PDT

Raw message

From: mpd@netcom.com (Mike Duvos)
Date: Mon, 2 Oct 95 23:02:32 PDT
To: cypherpunks@toad.com
Subject: Re: Introducing Johnson-Grace
In-Reply-To: <95Oct2.185011pdt.21953@mail.jgc.com>
Message-ID: <199510030602.XAA18317@netcom17.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


In an unsolicited advertisement,
Publisher Program <TechSupport@jgc.com> writes:

 > You may have never heard of Johnson-Grace Company, but you
 > have probably seen the benefits of our image compression
 > technology in America Online and Apple's eWorld.

Two painfully slow online services.

 > In fact, ART formatted images transmit three times faster
 > over conventional telephone lines than old-style GIFs and
 > JPEGs.

This might very well be true for a restricted set of images, but
I seriously doubt that arbitrary color photographic images take
three times less space with this system than with JPEG given
comparable retention of detail.

Also, are we suggesting that images transmit exactly three times
faster as both GIF and JPEG.  This is somewhat odd, given that
the difference in size between the two latter formats is often a
factor of ten? If you can beat JPEG by a factor of three, you can
beat GIF by about a factor of 30.

If you really can do this, you should be working on MPEG-4, and
not spamming our nice little list with advertising material.  I
should point out that neither Fractals, Lapped transforms, or
Wavelets can beat JPEG by a factor of three, and neither, I
suspect, can you.

[Huge Self-Promotion and Wonders of the Product Elided]

 > For example, our proprietary Splash(TM) feature displays a
 > full-size image in roughly one second; the image then
 > becomes sharper as more detail is received and decompressed
 > in additional layers, producing a high-quality final image
 > within seconds.

JPEG can do this quite easily with either the progressive or
hierarchical modes of transmission.  Why reinvent the wheel?

The engineering graveyard is littered with the bodies of various
entities who announced spectacular image compression advances. In
each case, reality dawned shortly after the hype died down.

 > Additionally, Johnson-Grace is in discussions with all the
 > other major Web browser companies to include ART technology
 > in their products. We expect broad support by the end of the
 > year.

Translation: We've learned how to use the "associate" command in
WinDoze File Manager.

 > For example, later this year you'll see news of ART speech
 > compression technology that will enable publishers to author
 > interleaved sound and images for real-time playback at
 > 14,400 bits per second.

It's been done.  Do the words "low bandwidth videoconferencing"
ring a bell?  I'll be truely surprised if you can beat
PictureTel's complex proprietary algorithm for sound and image
compression.

[Silly Form Deleted]

--
     Mike Duvos         $    PGP 2.6 Public Key available     $
     mpd@netcom.com     $    via Finger.                      $






Thread