This LA Times article is missing critical information. SMS was not designed so deliberately by one man. It was mostly an accident of the technical specs of a GSM cellular network.

The specs defined the size of packets that were to be sent from the network, over the SS7 signaling network, and then wirelessly over the control channel from a tower to the phones connected to it. The tower needs to communicate with the phones to get their ID, check their credentials, tell them when to switch to another tower, tell them if they have a voicemail waiting, tell them to ring, accept their dialing call requests, etc. These are all “control” events.

Voice calls take place on TWO separate radio channels, one for voice in and one for voice out, and the radio is only activated to do this when needed. This saves on bandwidth and on battery. The control channel, on the other hand, is more or less always on. Phones use this channel to set up the calls and activate the voice channel. THIS is the original purpose of this channel in the GSM standard, and the size and length of control instruction packets was defined around these tasks.

So later on, some people decided to “tack on” the ability to send a message inside this same channel. It WAS smart, and it does make sense. They simply looked at the packet size they had available, and made the best of it. This made it into the GSM standard, and since GSM is a rigid standard, all the terminal vendors built handsets that complied.

It was a great “invention”, but it was invention by committee, not one guy (and more than one guy claims credit). That’s how the GSM Association works. Also, it was not as “free form” as the LA Times would have you believe – some guy typing messages out to see how long they were, and deciding on 160 characters. Maybe it’s true that he did that as well, but if so, he did it just to test whether 160 would be adequate. The SMS specs were already constrained to 140 bytes by the real priority technical specs, and the SMS function was just a “nice to have” add in.

You see, the network control signals were sent between tower and handset in a packet that included header information plus about 140 bytes of payload. If you wanted to piggyback on this existing channel, you had 140 bytes to work with – no matter what your typewriter experiment concluded.

Hillebrand, or someone, then took the 140 byte limit (1 byte = 8 bits = 1 character), and figured an easy way to use a reduced character set of 7 bits each to squeeze 160 characters out of the constraint (140 * 8 / 7 = 160)

Therefore, an SMS message isn’t limited to 160 characters because of some dude’s typewriter experiment. It is limited because the signaling channel’s standard packet size payload was 140 bytes. It’s a less glamorous reality.