In article <4B86596F.5000603 at softjar.se>,
Johnny Billquist <bqt at softjar.se> writes:
Well, the E&S PS/340 also was a raster display
system, as an add-on to
the PS/330.
See
http://www.bmsc.washington.edu/people/merritt/graphics/ps330/ps330.html
That's the first I've heard of that; it must have been an interim
product before or contemporary with the 390.
Very few early E&S systems survive and since they never sold in high
numbers, its also hard to find people who used them.
I think it might very well be that which is shown on
the video.
I still doubt that. As I mentioned earlier, the PS series was higher
performance than what we see in the video. Also, being a high-end
system the monitors and tablets would have been labelled with the E&S
brand and I don't see that in the video as I have seen on monitors,
keyboards and tablets used for those systems. Now that I think about
it, the lack of a keyboard is another giveaway that this is not an E&S
system. A keyboard for the serial terminal is shown, but not one for
the graphics system directly. The E&S PS series had integral
keyboards and were full-blown terminals and not just a memory mapped
frame buffer.
For the black enclosure look, see these items at the CHM:
<http://www.computerhistory.org/collections/accession/102633847>
<http://www.computerhistory.org/collections/accession/102633848>
<http://www.computerhistory.org/collections/accession/102633849>
PS 330:
<http://www.computerhistory.org/collections/accession/102691276>
For a Picture System brochure, look here:
<http://www.computerhistory.org/collections/accession/102646288>
CHM has a brochure for the PS 340 but its not online:
<http://www.computerhistory.org/collections/accession/102646825>
<http://www.computerhistory.org/collections/accession/102646821>
Ditto PS 320, PS 330:
<http://www.computerhistory.org/collections/accession/102646824>
<http://www.computerhistory.org/collections/accession/102646823>
<http://www.computerhistory.org/collections/accession/102646822>
Raster operations are another issue. The video is
showing a bitmapped
high resolution picture, and it obviously isn't too fast at loading
another high res picture.
The video also shows some painting operations, particularly what would
be called clone brush by most programs today. This is a raster op,
albeit the simplest of raster ops -- copy.
However, with a memory-mapped framebuffer, copy is easily implemented
on the CPU and can be reasonable fast, certainly fast enough to keep
up with the size of brush they were using in the video. Hardware
rasterops become more important when you start running a windowing
system and need to move around big chunks of screen as a common case
(moving windows in response to the mouse).
Bottom line for me is that this looks like a 3rd party memory-mapped
framebuffer and not much more. That makes sense, because it would be
the cheapest way to make graphics available at the time and if you
aren't doing molecular modeling or manipulating complex CAD models, it
isn't worth spending hundreds of thousands of dollars so you can
doodle.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>