HN Theater @HNTheaterMonth

⬐ EVdotIO

The entire AT&T archive channel is really fascinating. There is something about phone switching systems that scratches that nerd itch for me.

⬐ walrus01

Researching the full extent of the old AT&T long lines network, the AUTOVON network, L3 carrier systems is a very deep rabbit hole to go down. It's a combination of all the major pre fiber Intercity telecom links in north America plus a lot of amazing military grade underground bunkers in some key locations.

⬐ t3rabytes

Are there any books on it?

⬐ walrus01

there's a guy who published a photo book:

https://99percentinvisible.org/article/vintage-skynet-atts-a...

but I think most of the info is online and from internet-based hobbyists and retired AT&T long lines people who staffed some of these sites.

⬐ jchulce

These old ESS switches are currently being retired en masse in favor of packet-based softswitches. For example, see the Verizon network disclosures page [1]. Every listing for "Planned Network Change" or "Switch Retirement" is another 5ESS or DMS100 being ripped out.

[1] https://www.verizon.com/about/terms-conditions/network-discl...

⬐ timthorn

Not quite - there are a bunch of CS2K retirements listed there, which was the soft switch evolution of the DMS100. I'm personally a little sad about those retirements as my first professional job was on the small team at Nortel developing the prototype of a DMS100 call server controlling a packet switch fabric. We started with ATM but moved to Ethernet when it became clear that was going to be the packet fabric of choice. Still, it's had a good innings and they were fun times.

⬐ larrywright

I hadn't heard the name Nortel in a long time - I used to work for a company that had a lot of Nortel phone switches. Wikipedia says that they're now defunct, which makes me a little sad.

⬐ runlevel1

The dot-com bubble, mismanagement, and accounting shenanigans took it from being ~1/3 of the entire value of the Toronto Stock Exchange and an employer of almost 100K employees to bankruptcy within the span of a decade.

⬐ sneak

Packet switched voice was such a mistake. Even under great conditions it is ROUTINELY worse than the worst days of circuit switched and TDM.

Kids these days will never know how great phones worked in the 80s and 90s. Fuck.

Give me that circuit switched reliability over high res audio with occasional/intermittent/unavoidable packet loss and jitter any day of the week.

⬐ heisenbit

Who cares about reliability, it is not as if lives depend on it. /s

If I look at the availability of phone service in my home (German metro area, cable) it is way below anyone would gave dared to provide in the TDM/ISDN times. I strongly believe only the redundancies people have with also having a mobile phone has allowed this to continue with lives being lost and regulators clamping down.

⬐ Symbiote

Are there statistics, or is this personal experience?

I've never had a landline phone fail, but I also haven't had one for about 12 years.

⬐ cafard

I have had a landline phone fail--a tree fell over on the next block down and took the wires along. Also, many years ago, I was in a building where we found the phones would work without power, but they wouldn't ring. While waiting to go home, we would resort to "polling"--pick up a line and see whether anyone was on it.

⬐ walrus01

Yeah, but also in an era of vastly less inter city capacity for long haul. Properly implemented Opus on a network with under 1% packet loss sounds great.

⬐ sneak

Yeah but how often are two people connected by networks that maintain <1% PL and low jitter for the entirety of a length of a call?

I'm thinking of just shipping 79xxs to my close friends and telling them to plug them directly into their routers, but even that will go to shit around 20:00 local as their residential networks overload.

⬐ walrus01

The weakest link is absolutely the residential class last mile contended access networks. I've regularly seen 0.00% packet loss for weeks at a time on business class connections between two points in the continental US, but when one end is slightly flaky, all bets are off.

⬐ dayofthedaleks

The holiday bombing in Nashville provided clear evidence of what we've lost in terms of PSTN resiliency. I don't think there would have been region-wide loss of service for greater than 24 hours in say 1994.

⬐ whalesalad

This is so cool, Glendale is my home town. I always wondered what the inside of those big huge windowless buildings looked like.

⬐ Tempest1981

This surprised me, at 2:08

"These cable cutters are sharpened in such a manner that they will cut a clean cut through these, and will not cause any shorts"

Seems like there would still be brief millisecond shorts? But maybe that's ok?

⬐ anotha1

Probably just trying to avoid continued shorts from wires that are torn apart instead of cleanly cut.

⬐ lalalalala777

No rollback plan here.

⬐ tibbydudeza

Give new meaning to the term "we are cutting over".

⬐ ehayes

Or is it in fact the original meaning? (Genuinely curious)

⬐ function_seven

Yeah, pretty sure this is the literal origin of the term. Or closer to it anyway.

⬐ runlevel1

So far, the only dictionary I see with any hint at the etymology (apart from saying it's cut + over) is Wiktionary.

From: https://en.wiktionary.org/wiki/cutover

3. The process of quickly replacing a telephone switchboard, in which the connections are duplicated to the new machine and the original connections are then suddenly disconnected.

1913: Bell Telephone News, vol. 2, no. 7, page 18:

> After the heat coils were pulled on the old main frame, the remaining step was merely to pull the wooden plugs on the new switchboard by the strings attached to them (which are bunched together) upon a signal given from the old terminal room, indicating the removal of the heat coils above mentioned. By this means the cut-over was accomplished almost momentarily, the process occupying not over two seconds’ time.

Surprisingly, many dictionaries don't have this sense of the word at all.

EDIT 1: Changed the dictionary from Wordsense to Wiktionary since that's where they were getting it from.

EDIT 2: Interestingly, articles in "Bell Telephone News" also use the term "cut into service":

> Two telephone exchanges which the Western Electric Company has been installing in Pekin, China, were cut into service on April 2

> In June, 1910, the second unit (Belmont) B board was cut into service

> Cut-Over at Watertown - The New York Telephone Company's new exchange at Watertown, NY, has been cut into service.

I wonder if "cut into" refers to splicing. Or, it could just be that "cut into" naturally emerged because "cut over" didn't adequately capture the case where no previous connection existed.

⬐ None

None

⬐ wsh

This Usenet thread (1991) explains more about the procedure:

https://yarchive.net/phone/cutover.html

⬐ op00to

Thanks, this is good content!

Ma Bell had some ill ill communication. Those were the days, when the real paperclip payphone trick still worked. I'm not even sure pay-phones still exist in the Western world unless they're Tardises.

Btw, here's a video from her vault about an actual cutover done in 45 seconds with a few dozen folks with big garden cutters: https://youtu.be/saRir95iIWk

PS: I'm so old, I remember the days of operators, telephone books and fingers doing the walking. :) Speaking of old meets new, I recently used an antique rotary phone that's built into a brass end-table to diagnose an Xfinity VOIP call issue. It still works.

There's an old AT&T video showing the careful coordination and huge amount of manpower needed to upgrade a single telco switch. They had rows of workers with cable cutters, in order to literally cut the entire service over to the new equipment. Total downtime was just 47 seconds.

(And even during the cutover, they're careful not to interrupt any emergency calls.)

https://www.youtube.com/watch?v=saRir95iIWk

⬐ setquk

That was absolutely amazing. Thanks for posting.

⬐ the_duke

So, what exactly happened there?

Where were the signals going after the cables have been cut?

⬐ Animats

The new switch was connected to each line in parallel with the old switch, but it was able to ignore incoming signals until told to take over. The old switch was hard wired enough that it could not be remotely commanded to have no effect on incoming signals. The line relay attached to each line had to be physically cut out of circuit.

⬐ None

None

⬐ camtarn

That is spectacular - thanks for the link!

⬐ DanBC

I love the "DON'T TOUCH THIS" post it at 3:55