laser-lovers@uw-beaver (01/28/85)
From: Brian Reid <reid@Glacier> Judging from the number of questions I've received, I guess I was too obscure in my comparison of ImPress and PostScript, with respect to the statement that "neither language completely dominates the other." Both languages are complete, which is to say they can both describe any image that a given raster printer can print. At some level, both languages must fall back on the crude hack of storing rasters. There is a very wide range of page images for which ImPress must fall back and use rasters but for which PostScript can still use an analytical description. I don't need to preach to you all about the benefits of an analytical description. I am not aware of any page images for which PostScript is forced to use rasters but for which ImPress can get by with an analytical description. That is what I meant when I said it would be fairly easy to translate ImPress into PostScript. In fact, since PostScript is a programming language, all you would have to do is to define one PostSript operator that corresponded to each ImPress operator, and then turn the ImPress file from prefix form into postfix form (Impress has you say "MOVETO X Y" while PostScript has you say "X Y MOVETO")--no big deal. No further translation would be necessary. (these operator definitions would just be prepended to the front of the file). When I said that "one format does not totally dominate the other", what I meant is that the ImPress format is superior for applications in which speed is at a premium. In theory an ImPress file can always be printed faster than a PostScript description of the same image, just because the PostScript interpreter must do more computing. Our Imagen 12/240 printer, for example, can print 12 pages a minute with an 8MHz 68000 processor; the Apple LaserWriter can barely keep up with 8 pages a minute with a 10MHz 68000 processor. The principals of Imagen will remember, if they stretch their memories, that they and I spent quite some time discussing page description languages back in 1980 and 1981 at Stanford. They will also remember that at that time (long before Adobe was invented) I told them I thought ImPress was too simplistic and too limiting; they gave me various reasons why it had to be that way. Well, in once sense they have proved me wrong, because they have built a fine company with many happy customers by selling laser printers using this format that I never liked; they have good reliable products, good documentation, good customer support (at last!!), good fonts, and so forth. The Imagen 8/300 is an absolute marvel of a machine, and almost everybody that I know who has one thinks it is wonderful. I mean no harm or ill will to Imagen; I'm just glad that finally they have some serious competition, and I'm especially pleased that the competition's page description language works more or less the way that I tried to convince Les Earnest and Luis Trabb-Pardo to go with back in 1980 or 81 (whenever it was). A couple of people asked in private mail for me to compare the Apple LaserWriter with the HP LaserJet, given that they both use the same marking engine and that they look very similar. I don't think the HP LaserJet is even worth discussing as far as its page description capabilities. It's like comparing rockets and bicycles--why bother. Bicycles are cheaper than rockets; the HP LaserJet is cheaper than the Apple LaserWriter. I suppose you could throw your bicycle through the air and call it a missile; you can also engage in ultra-intense hackery and get your HP LaserJet to do something that you can call typesetting, but realistically the HP machine is just a clever variation on a daisy-wheel printer. -- Brian Reid decwrl!glacier!reid Stanford reid@SU-Glacier.ARPA