ellis@flairvax.UUCP (07/01/83)
A recent submission by Lauren about Bell answering machines -- << If you paid less a month, you'd get "versions" that had progressively less incoming or outgoing message time. Subscribers were told that this was only fair, since different length tapes had to be installed, and cost varying amounts. In reality, there was only ONE version of the model 700 Code-A-Phone. If you didn't pay for the "maximum version", the installer would set a pair of little cams in the unit which would artificially limit the incoming and outgoing message times! Talk about "creative" product design... >> ...reminds me of MANY instances I've seen in our own industry. For instance, one computer I worked with had two versions -- one with only one interrupt level, another with 4. For several thousand dollars the local rep would come in and install a new board -- all for show since the boards were identical except for a simple modification anyone in the know could easily perform. Another company I had experience with charged varying amounts of money for 'different versions' of a computer -- scientific, business, and personal, and each of these with various suboptions. But they were all the same -- the cheaper ones were simply "castrated" by disabling certain logic in PROM. I wonder how often this kind of thing happens. It strikes me as immoral and dishonest, a glimpse into the darker, self-destructive side of capitalism. Think of it, creating a good thing, and then partially breaking it to get more money for an unbroken one! Michael Ellis - Fairchild AI Lab - Palo Alto CA - (415) 858-4270
leichter@yale-com.UUCP (07/01/83)
It's not just the computer industry! - although the story about the expensive field-upgrade from some IBM 360 to another that had the CE come in and cut a jumper is famous. When the alternator on my old '73 Pontiac went a couple of years back, I discovered that GM used the same part in a wide variety of cars. However, the list priced depended on what car's part book you looked it up in! My "Pontiac" alternator was cheaper than a "Cadillac" alternator (more expensive than the Chevy? I don't know.) -- Jerry decvax!yale-comix!leichter leichter@yale
charlie@cca.UUCP (07/02/83)
I have two great tales: It is well known that many computer vendors will sell different processor models which are software compatable but run at different speeds which are actually the same hardware with the clock reset. Honeywell in the early 1970's carried this to a further extreme. Processor models 66/07, 66/17, and 66/27 were called gearshift machines. They ran "more competitively" when running time-sharing than when running batch. This is because in the batch world, Honeywell was competing with IBM while in timesharing they were competing with DEC. IBM did not offer as good a price/performance as DEC. This had to have been implemented by placing the speed of the CPU under *software* control. The practice was discontinued, either because IBM got more competitive or too many hackers figured out how to patch the OS. A commercial service bureau (reportedly GE, but I'm not certain) had two compatible FORTRAN compilers. A checkout and an optimizer. Compiling with the optimizing compiler took three times as long, but the resulting code ran 40% faster. Exactly three times as long and exactly 40% faster. Hmmm...
mat@hou5e.UUCP (07/03/83)
I have heard that the Itty Bitty Machine Company did this in an odd way some years back -- they charged more for maintanance if the slowdown jumper waws removed. Their justification was that running at the higher speed reduced tolerances and decreased the MTBF. Mark Terribile Duke of deNet
dave@utcsrgv.UUCP (Dave Sherman) (07/03/83)
Apparently, when Lear-Siegler first came out with the ADM3-A (ADM, by the way, stands for American Dream Machine(!)), it was so (relatively) cheap that they introduced it as an upper-case terminal as a way of explaining the cheap price. Only later did people discover that you can flip one of the buttons under the screw-on panel to get lower case. This is what a (non-UNIX, non-net) friend tells me. Can anyone confirm? Dave Sherman Toronto
dee@cca.UUCP (Donald Eastlake) (07/03/83)
I have heard some faintly rational arguments for artificially different products. Assume you design and sell the Wizz-Bang Frammis which is the greatest thing since sliced bread for the high ticket end of the frammis market. It is also so cleverly designed that you have a very wide profit margin and still beat out all the competition (effectively these high profits are your reward for cleverness, investing in research, etc.). The only problem is that your Wizz-Bang Frammis is too expensive for the low ticket end of the frammis market. What do you do? If you ignore the problem, your competitors keep selling to the low end market and may be able to hold on to customers when they migrate to larger frammises. It would seem that the rational thing was to design and manufacture the Fizzle-Bang Frammis which uses your clever design principles and would also enable you to dominate the low-ticket end of the market; however, when you consider the high costs of designing, manufacturing, and maintaining a completely different design, the net profits become questionable. By simply installing a few jumpers so that the Wizz-Bang Frammis skips two out of three cycles, you have a product which seems just right for the low-ticket market. You retain the advantages of volume production, having to stock parts for only one product, etc. And, due to your high profit margins at the top of the market, you probably still make money at the low end. So you announce the Fizzle-Bang Frammis which is just a disabled Wizz-Bang Frammis. You might even figure it is better to loss a bit of money on each low end machine you sell to try to dominate the entire market. Of course, this whole thing falls apart if the market is competitive enough and large enough as a produce designed for the low end market is bound to make more money eventually. Donald Eastlake dee@cca, decvax!cca!dee PS: The above explanation does not decrease my general feeling of disgust at artifically disabled products.
rosin@hogpc.UUCP (07/05/83)
IBM once marketed a "thrifty fifty" series of tab machines, including a 407 'tabulating printer' (used by most universities as a way to list a deck of punched cards without tying up a CPU). This equipment was actually factory reconditioned and rented for considerably less than (supposedly) new facilities. The 407, however, ran 2/3 the speed of the higher priced model becuase 3 relays were ADDED to the machine to slow it down - printing only two out of every three cycles, thus allowing for a reduced rental!! (I also recall that some adders were removed/disabled as well.) While on a lecture/visit to Tuskegee Institute in 1965 I constructed a jumper to circumvent these relays on their printer. It was probably the most valuable outcome of the trip from their point of view. Bob Rosin, Ameican Bell, houx_!hogpc!rosin, 201-576-3549
padpowell@wateng.UUCP (PAD Powell[Admin]) (07/06/83)
Well, I have had the interesting experience of having a customer request a "slower" product. On a single board system under developement, it was neccessary to build a lot of effort into the IO stuff. Basically the silly users wanted a FIFO on input and output. The chip level design was horrible. But it turns out that a 6809 with a 256 byte ROM and a couple of jelly-beans (strawberry of course) does it quite nicely. When the design was presented to the managment of the company who was paying for this little gem, they refused it on the grounds that they were paying for a "Single Processor System". I refuse to describe the comedy that then took place as the cost/benefits of the thing were explained. Of course, it is now a major selling point... A Multiple Processor Control Board!!! Patrick ("Never give a gift to an accounant, he will want to know if it is tax deductable") Powell
hoffman@pitt.UUCP (07/06/83)
While we're on the subject of artificial differences, does anyone remember the price schedule on the DEC DL11 series? There were actually two different boards, the M7800 and the M7800-YA. The -YA model didn't have the chips for EIA; it was strictly 20 mA loop and was used in the DL11-A and DL11-C versions. The DL11-B, -D, and -E were all the M7800 board and had EIA drivers as well as current loop. That's really not my point, though. The DL11-E was a more expensive board than the others because it had modem control -- DTR, RTS, CTS, DCD, etc. were brought out to registers. It was only a jumper change (well documented in the engineering drawings) to convert a -B or -D board to an -E though. ---Bob "What, I've voided my service contract?" Hoffman U. of Pitt CS; pitt!hoffman; hoffman.pitt@Udel-Relay
dbl@duke.UUCP (07/06/83)
My recollection was that the standard Adm-3a was a upper case only terminal. It did indeed have a switch to put it in lower case mode, but this switch had no effect unless you had purchased and installed the ROM which contained the lower case character set. David Leonard (duke!dbl)
alanj@tekecs.UUCP (Alan Jeddeloh) (07/08/83)
Ah, yes! the DL-11 ... Later on DEC caught on and started de-populating the boards. If you bought a DL-11A, you got a board minus the RS232 drivers, a DL-11C was minus the current loop drivers. You had to buy a DL-11E to get everything. Back when I was an applications engineer for a Tektronix marketing group I used to keep one of the "good" boards in my desk drawer for building up special systems. As for other manufacturers, CDC also had a few up its sleeve. I think difference between a CDC 3300 and a CDC 3170 was the crystal on the master clock card. (The 3170/3300 were 24-bit, single accumulator, 1's complement, and had virtual memory (before IBM). It did take them a while to get their business instruction unit (string edits/moves/compares &c) to be able to restart after a page fault, but they did it.) I also seem to recall that the difference between some of the original CDC 6000-series was that on the lower performance models the program counters on some of the Peripheral Processors were wired to zero. I think that had something to do with the "barrel" architecture of the PP's -- they couldn't delete any of the circuitry because it was timeshared between the PP's. Can anyone knowledgeable on the CDC 6000's confirm or deny?
faunt@hplabs.UUCP (07/09/83)
DL11's were later manufactured from the same PC board, but not fully loaded, so to go from current-loop to RS232 you had to add a couple of IC's, diodes and caps for the IC power.
smh@mit-eddie.UUCP (07/11/83)
The recently posted descriptions of product lines with artificial pricing for essentially identical hardware prompts me to add two more: - When PDP11's were young, the common tty interface board was the DL11 (approximately a single-line DZ) which came in five flavors differing in their modem-control capabilities, RS232 vs. current loop, etc. I had a machine with a current loop board which I wanted to `convert' to a RS232. My DEC salesman could sell me a `converter' to do the job at about $50. This seemed reasonable at the time, but I was in a rush, so I thought to check the board to see if perhaps positions had been left for adding RS232 converter chips (MC1488,9). I was astonished to find them already present. Reading of the charts while being careful to ignore the fictions of the accompanying user manual proved that the only difference between the DL11A,B,C,D,E versions were jumpers on its 40-pin tty connector socket! - In the days of unit record (card) tabulator equipment, there was an IBM beast called a 407 which could actually do fairly sophisticated stuff, but was generally used around real computers as a slow card listing device. The machine was a wondrous and curiously reliable maze of cams, levers, and strange gizmos. There were undoubtably zillions of options, but two printing speeds were available: 120 and 180 CPM (cards per minute!). I never worked with the 180 CPM version, but I still remember the pleasant 3 Hz cycle of the slower (and more common?) kind: PRINT-PRINT-swish-PRINT-PRINT-swish-PRINT-PRINT-swish Make of it what you will -- I never looked inside at the cams... Actually, in sympathy with these two largest and allegedly reputable computer manufacturers, I can imagine pressures for such artificialities arising from the developmental history of a product line. A faster, cheaper, or better version of a frob if priced `honestly' can cause hard feelings in the poor slob who just bought the old style. Then too, what does a manufacturer do who leases out most of his units and has a huge inventory? Steve Haflich genrad!mit-eddie!smh