Editorial 10-07 - SNA: A lingering death
I got my first car in 1980 when I was Junior incollege. It was an orangey-yellow FordMaverick, a 1972 with about 70k miles on it. It was kind of ugly, but that kind of ugliness added to its beauty in anodd way. I wasn't afraid to take thingsapart, such as my busted old lawnmower. It was putting them back together again that messed me up. So, when I lifted the hood of my"banana-boat" Ford, the simple straight-six engine was distinct --pulleys, distributor, alternator, etc. -- I could really understand what I waslooking at, even if I wasn't totally sure how it all worked. Whenever the car was in for service, I wouldtry to peer over the shoulder to watch what the mechanics were doing. I didn't want to make him nervous looking overhis shoulder. I asked if he would showme the distributor, because I never really understood what it was and why itwas such a big deal. They used a strobelight to check the timing and I thought the whole thing was very cool. I felt more comfortable with the whole ideaof understanding the inner workings of the engine. It was the sort of aimless exploration anddiscovery that a layman could do in those years. It was in that spirit, that I also began todig into the fledgling microcomputers at the time. (Nowadays, I lift up the hood my car anddon't really know what's what, nor would I even thing of trying to fixanything, excepting that maybe that poking around and looking at stuff wouldfix it.)
I got my first microcomputer in 1982, an Atari 800 with a realkeyboard, a cassette tape backup (no hard drives yet), and a Basic coder. It attached to a black and white TV. After I got simple stuff going, I wrote areally stupid and clumsy address list report. >Problem was that without a printer, it did little good except suck downa lot of evening and weekend time. Ibought a used 150lb Centronics 101 printer with its new Centronics parallelport in addition to serial. Just gettingit out of the car and up the stairs was an exercise unto itself. Using the printer and the computer, I wassuccessfully able to actually "do" something. My mother?s high school class reunion was myfirst customer (for their mailing list). It helped that my mom and her friend were the reunion chairwomen. Anyway, that's where this began. That was when I could actually make mycomputer do stuff, and printing was my interest. I explored the sending and receiving of dataand manipulated the sequences of Hex code I sent to the printer. I monitored the serial port for its Hexresponses. I gained a deep understandingthat it takes to establish communications between CPU and printer..
Fast forward to today. The technology has heaped layer upon layer of function onto simple 8-bitbyte code which runs printers at their most fundamental level. It is the understanding of the fundamentalswhich is commonly not taught, and the programmers, engineers, and designers,giving short shrift to "boring" old impact printers, seem to assume higher level language andinterfaces in their interoperations.
It against this backdrop that I present the following extremelyimportant analysis. The next release ofoperating system for the iSeries (AS/400), OS/400, will NOT SUPPORT SNA. Very few people understand how vast theramifications will be for printing. IBMsurely knows that many of the workarounds for printing over TCP/IP rely, attheir most basic level, on an SNA link from the host. IBM's solution? Advanced Function Printing (AFP) usingIntelligent Printer Data Stream (IPDS) for FLOW CONTROL over TCP/IP. Unless you happen to have seen this coming,you may not have IPDS implemented, and replacing all those impact printers (ordealing with Host Print Transform and PCL bandwidth and processing overhead)with new devices is going cause quite a lot of consternation. The whole concept of the "Black Hole of Printing" will become familiar to those who have puzzled at the lost jobs, strangeoutputs and flaky printer performance and network traffic they create.
I have been and will continue to offer solutions to thesecomplexities. The effort and cost thatso many people will expend on understanding the implications and potential solutionsto this issue will be astounding. I'llcut through the variables for you and help you through the most vexingchallenges.
David T. Mendelson
Argecy Computer Corporation
27280 Haggerty Road C21
Farmington Hills, MI 48331