return to main page

Special Features Installed
- just starting - not verified

also starting a list RFQs
(Request for Quotes) for special engineering
*not* installed on this machine - just for fun -

A comment from Carl Claunch < Carl.Claunch @ gartner . com > Wed, Jul 15, 2015
The optional features were a mix of causes.

In some cases, these come about because the sales force is finding some issue that blocks their sale of a 1401 system. An example would be Punch-Feed-Read, where a customer operating with the punched card mechanical predecessor equipment would have some part of their processing where they needed to punch information into an input card. It might be values that are calculated based on the input card and then stored there, but the key is that the customer had a workflow that involves a single card which is created at two different times, one part done based on some calculations or through merging information from other cards. To change the way they do things would be disruptive and require changing many stored cards from older runs. The potential buyer raises this as a reason that they will wait a bit before replacing their tab equipment with a shiny new 1401 system. Sales complains that they are losing business due to a product limitation. Someone develops PFR as an option. It may be enough to allay the customer concern, even if they don’t end up using it.

In some cases, competitors force creation of these options. Someone has an alternative system which runs faster. Which is cheaper to develop, a new faster card reader/punch, a new faster CPU, or a feature that gives you an effective speedup for certain kinds of jobs?

In other cases, an engineer has a flash of inspiration, suggests a way to achieve something like Read Punch Release or Overlap, and it matches market needs. Each option is a way of extracting some additional revenue from a customer, too.

In other cases, a customer who is fairly knowledgeable and big enough to have clout due to their annual spend with IBM would suggest something should be possible, offer to pay for it, and then engineering would figure out how to do it.

IBM entertained requests for custom features through an RPQ process, which could then be turned into standard offerings (but options not included on every machine).

The 1401 evidences many waves of modifications and enhancements that were designed well after the initial systems were shipped. Look at all the addressing modifications for larger storage, index registers, address modification instructions, just as one small area that reflects changes which would have been MUCH easier to have accommodated in the initial architecture and design. Addressing fields with three BCD characters would not be the choice you made if you planned from the start to have 16000 locations and indexing – or at least the way I imagine the architecture would be made. They seem like clever hacks added after the fact – “we can use the zone bits of the middle character”.

Back at that point in computing history, most programmers were highly focused on efficiency in design. Writing code to save a few character positions, a few machine cycles, or take advantage of special conditions to run faster – these were items of pride. Programmers would show off share their clever ideas widely. Not sure if there were contests a la IOCCC but lots of peer admiration when someone found a way to do things a bit faster/better/smaller.


Please note:

this list is just starting, July 3, 2006, has not been discussed with the various group leaders

Feature Name Comments
added memory total 16,000 characters
sterling (British)  
overlap and what a problem that can be
STL (Selective Tape Lister) July 3, 2006 - John Van Gardner (living in Georgia) noted pictures of lot of red wiring on the 01A4 gate. "When I looked at the card side photo the cards in F05, F06, F07 & F08 look like they each have two large power transistors. There is also a cable with all red wires coming from the C26 socket. The power transistors could be used to pick magnets. The only feature I could think of was the STL (Selective Tape Lister). It is described on page I-62 of the 1401 Special Features manual A24-3071-2 which you have. It has 8 magnets to space the individual tapes. A lot of banks had this feature on their printers.

I looked at the 01A4 plug chart you posted on the web and it is like mine. Then I looked at page sheet 4 of 6. It shows the 1403 summary connector #2 and the wires used in the STL feature. They come from the 02B8-C02 connector. If the red cable goes to the 02B8 gate thats probably the STL feature. That feature should include ALD pages and "

?Remote Recycle Control box? July 3, 2006 - John Van Gardner noted: " Also attached is a scan of the ALD page which shows the Remote Recycle Control box. Note the 75 position connector. This is why I was asking about the connector that Bill Flora identified as the 1407 connector. The RRC connector was in the same area of the machine and also had a terminator plug. It was a good tool for trouble shooting bugs where the machine would not loop. I wonder if they had these in Germany."

a list RFQs
(Request for Quotes) for special engineering
*not* installed on this machine - just for fun -

One of the interesting (challenging) things about computers (or I imagine other expensive machines) is the response by manufacturers to provide support for "features" not in the current design of the - machine - being provided.

These are usually called by the generic name "RFQ" or "Request for Quote", a request for the manufacturer to provide a price quote on some unusual feature. Even if the new feature is quoted and sold to the customer it is called an "RFQ". Sometimes the "RFQ" gets quite popular, and may enter the standard product line.

This is the start of a list of RFQs for the 1401 system.

Two printers on one 1401 - an exchange with Van Gardner

Two printers on one 1401
He wrote the 1401 peripheral support program that drove 2 printers and the 1402 at full speed. When I told him I had written about him he said he still had card decks and listings from those days. I'll have to start working on him about those.
Two printers on a 1401 - what an engineering challenge !?! I'm more used to systems that had external device controllers and you just cabled them up to the existing data buss. The command to the controller included the starting memory address and they just hauled their own data out of memory.

How was it done on the 1401?
The dual printer 1401 was an RPQ Attachment. It consisted of a second print buffer gate mounted in a frame like the 1406. The second printer was attached to this frame. It was called a 1944 or 1924. I am not sure of which. A print command with a D Modifier character selected the second printer and it only took 2 ms to send the data to the buffer and from then on that printer was on its own. I'll have to see if John Malone still has any information on the RPQ.

return to
main page