History--Eight Digit US telephone numbers? [telecom]

Well, it still exists:

What you are describing, to a tee, is the final selector or 'connector' in a Strowger (step) office.

The next to last digit, the 'tens' digit, would cause the switch to ratchet upward, the number of levels being equal to the digit on the dial, and the last digit, the 'units' digit, would cause it to rotate into the bank to the line being called.

#include ;-)

The first and second (incoming) selectors would step upward along with the dial pulses and then rotate to a vacant connection to the selector for the next digit.

A distinct audible 'clunk' could always be heard right after dialing the thousands and hundreds digits, as the selector rotated and seized the next level selector, but not after the tens digit. Similar clunks could be heard after each digit when dialing the office code, but the clunk after the first digit was often times masked by the extinction of the dial tone.

Columbia University used to have something like this in the early 1970s. However, I always thought of it as more of a step-based PBX with direct inward dialing than a true 'Centrex' with all of the switch-based bells and whistles such as call transfer, three-way, camp-on and such.

The only 'feature' I remember on that particular installation was that a hook flash would signal the human operator who could then do a call transfer or (do I really remember this?) three-way. IIRC, the operator was on a cord board. I also seem to recall that this was all driven by a crossbar tandem.

I can visualize that it would be very difficult to do any kind of 'Centrex' type features on a panel office with the totally inflexible architecture of the incoming and final frames. I think that hunt groups were the most fancy of the features that panel could do.

As I think back I can't think of any 'Centrex' or even any DID-PBX installations hosted on a #1 crossbar. I do recall some native #5 crossbar Centrex (or was it really DID-PBX, I guess it's a matter of semantics) installations and I do remember that #5 crossbar often times hosted the 101 ESS Centrex-CU installations, but I don't recall any #1 crossbar switches driving the 101 either.

I believe the first version of Centrex was essentially a PBX with direct inward dialing. Centrex II, which I think required ESS, had more of the "bells and whistles" with automatic call transfer and three-way, etc.

The older Centrex did transfers by hookswitch flash. For directly dialed calls and calls handled by a cordless board, it showed up on the operator's console. For calls handled by a cord board, it flashed the cord circuit supervisory (hookswitch) lights.

I don't think #1 crossbar did much of anything beyond switch calls. Indeed, the Bell Labs history, while discussing it, seems to downplay its role, focusing all the accolades on the #4 (toll) and #5 crossbar. My impression is that the #1 xbar was merely a more cost- efficient and trunk-efficient machine over panel, but didn't add anything in the way of customer features. Of course being trunk efficient was a major cost saving and one of the weaknesses of step in high traffic situations.

The history says they tried call waiting on the #5 but it wasn't cost- efficient to implement as a service on that machine.

I am aware the SBC actually deployed some calling features on a No. 5 XBAR on the north side of Kansas City, MO. A friend of mine subscribed to them circa late 1960s. I don't recall whether he was part of a test group or whether it was a public offering for that particular switch.

If there is a rotary group customer on the final selector it will search for a vacant line to that customer in the group just like in a earlier seclector in the switch train. The last line is busied out in the hunt group. If there is more than one hunt group, it works the same when when dialing the first number in each group.

When Phillips 66 headquarters in Bartlesville, Okla., adopted centrex, the engineers determine it was chaper to make it a centrex-CO and serve it out of the existing #5XB at the CO a block away.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

That's the one! Same geeks. :-)

This then confirms some rumors and speculation among 'enthusiasts' that Ma Bell had developed some kind of a 'stored program engine' for the #5 Crossbar that would replace the relay-logic markers and senders and such, and give the #5 Crossbar the ability to do the fancy Custom Calling Features just like the 1E.

This is the only first-person account I've heard of anything like this actually happening. I assume that it was never really rolled out simply due to the ESS deployment plans.

This answers for sure one thing I've kinda been curious about, but it reminded me of one thing I have never understood.

I always assumed that there would be something like that leaf switch to catch that all lines busy condition and return line busy to the caller, kind of like that 11th. step contact set on the other selectors. This explanation does make perfect sense.

One thing I'm still puzzled about ...

Is it possible on a Step office to have a hunt group with more than 10 (9) lines?

Was there maybe some circuitry to sense when all lines on a given level were busy and cause it to step upward one more level?

I can see how this might be done with a set of vertical contacts, similar to that on the linefinder, but I never recall hearing about anything like this in service.

Do you (anybody?) know?

It could have continued to work with SPC. A standard feature in the DMS-100 was 5 digit community dialing; i.e. dial the 5 digets plus the "#" sign. So far as I know, no LECs opted to implement this feature.

It only saved one digit. But, with the advent of mandatory 10 digit dialing it would be more desirable today.

I'm amazed at the number of trucks I still see on the road with their

Even more amazing is the prevalence of 7 digit numbers in "welcome to mumble mumble" guidebooks...

I would presume that ESS could easily handle 5 digit dialing. But I suspect it went away because there were too many other exchanges being created and the 5 digits weren't enough to make the local exchanges unique.

For instance, in border areas, residents could call across a state line to the opposite town using only seven digits. That made the exchange-codes in the border town had to be unique in _both_ area codes. That was fine in the old days, but as they began to run out of exchanges that was no longer workable. So, the border people had to dial 10 or 11(+1) digits.

As an aside, the Bell history mentions that long ago (1960s) Bell recognized they were running out of exchange-codes, and built/ programmed switches so that area codes could be an exchange code and vice versa. I believe this meant that the 1+ prefix would be necessary in places where it wasn't used before so as to differentiate the call.

Regarding the previous posts regarding the term Centrex vs. "inward dialing", it appears from the history that the function was developed first and given a name later. In the early 1960s there were a few different combinations of machine types that would support Centrex. Some users had a 701 dial step-by-step PBX which was modified to accept digits from the central office; others were served by a step-by- step central office that was modified; still others were served by No.

As time went on new features were added, such as automatic transfers and loop access consoles instead of direct trunk. ESS added more features.

Note that back then the switchgear was located about half at the central office and half at the user's site. The tradeoff for central office location--which meant all loops had to come from the CO--was that repairmen didn't have to go out to the user's site for service.

Why? In many placess with only one area code 10-digit dialing of a local number generates and error message to the caller.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

I checked the Bell Labs History 1925-75 and it does not appear that they hda replaced the relays with electronics in the common control unit. They considered it:

"In 1970 a committtee studied the possibilities of modernizing the No.

It should be noted that about the same time Western Electric developed a new electro-mechanical (crossbar) PBX, the 770A, that was more cost- competitive than certain low-end ESS designs by Bell Labs. "Even though electronics was clearly the way to the future, the 770A enjoyed a number of years when it was the economic choice on a first-cost basis over comparable designs."

(In 1970 I don't recall any electronic stand-alone cash registers, and I don't believe they became commonplace--more cost effective--until at least 10-15 years later. The cheap electronics we take for granted today was still in the laboratory in 1970. Anyone know when NCR ceased making stand-alone electro-mechanical and all-mechanical cash registers?)

They used electronics and mini-computers to aid in translation, AMA recording, and in signaling and timing circuits.

One innovation for No. 5 was the use of "wire spring" relays, replacing the U and Y type flat spring relays in 1953-54.

They did experiment with custom calling features on No. 5 crossbar in Columbus Ohio 1963-65, Iowa and Mass 1966, and Ohio 1971. They found that those services were too expensive to be done in crossbar, so rollout had to wait until ESS. They used piggyback twister memory in some cases.

The book is a bit vague on whether these tests were by an electronic or relay common control. I think the basic "logic" was still relay circuits, especially because it was done in the early 1960s when ESS itself was still under development. Further, the other comments above suggest electronics were still too new. I suspect that if they had a true electronic front end the cost of crossbar implementation would not have been excessively expensive.

Last, all-new installations of No. 5 crossbar ended in 1976.

But let's note that when it comes to machine "logic", a relay performs the same function as a transistor or vacuum tube, it's only bigger and slower. In the 1940s, IBM built sophisticated _programmable_ digital computers--the Harvard Mark I and its own SSEC using mostly relay logic, and Bell Labs had a relay computer, too. Also in the

Possibly because trucks sometimes venture outside of their home area codes?

Neal McLain

The first applicatins of DID in Southwestern Bell territory were at military bases--first Fort Sam Houston, Texas, and then at Fort Sill, Oklahoma.

Both of these were technically government-owned and operated PBXs. I don't know about Fort Sam, but Fort Sill was a step-by-step switch, I believe AE although I'm not sure. A prefix was assigned and off-base phones could then dial directly to what was formerly the extension number. (Before that you dialed the main number and a soldier/operator answered and dialed the extension you wanted.)

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

Why? Because where I see them is in the greater Los Angeles area, or Seattle, etc.

Where I live, that takes 20 minutes at 75 on the freeway. :-)

Then, there are overlays.

"Due to the limited access inherent in each level of the sxs switch, much effort was expended to obtain efficient gradings. In addition the concept of providing access greater than 10 was adopted. This was done by introducing an additional stage of 22 terminal switches, known as rotary out-trunk selectors to reduce the number of succeeding switches and trunks."--Ref. Trunk hunting Switches, Record, Dec 1928; Improved Graded Multiple for sxs offices, Record Sept 1944.

"At the terminating end of the switch train, new level-hunting connectors were developed that enabled the system to serve PBXs with more than ten trunks without the grading of the connector multiple, thereby obtaining better call completion." --Ref. "Level Hunting Connectors" Bell Labs Record March 1929.

See also Bell Labs Record Sept 1927.