The messages are time stamped so could in theory be reassembled to indicate a message. The SPOT device though has been designed to minimise the amount of Satellite bandwidth for eath SPOT device with the company running the SPOT system paying others for the satellite time.

I doubt whether the GPS position and the messaging information sent over the Satellite network would be more than 20 bytes in length. The GPS positional information and time stamp data could be sent using about 16 bytes. The SPOT unit serial number may be only 28 bits (3 Bytes and a Nibble) in length and the messaging system 2 bits with a couple of error correction bits (the other Nibble to make up the last byte). Overall the SPOT system isn't designed to send text messages but to just send the GPS and message data in the most compressed reliable format. Sendind a bit by bit message contained within the 20 or so byte format container packet isn't very efficient. The SPOT device is only capable of around 1900 messages before the batteries run out.

If composing an SMS text message on a possible future SPOT device then length of the text message would be around 140 bytes in length compared to the 20 or so bytes for each GPS update message currently sent. So one SMS text message will use the data bandwidth of around 7 GPS update messages. The designers of SPOT probably thought long and hard whether to include an SMS facility; the deciding factor was probably the overriding cost savings i.e. the initial market price point and the business model for the administration of the income stream or even the possibility that to gain access to the satellite network they had to demonstrate that the SPOT device would not specifically offer competition to SMS messaging services already available on the satellite constellation. It wouldn't take much to add a simple 1 or 2 line 16 character LCD display and a simple keypad to enter the SMS message into a future SPOT in addition to the current features. The problem then becomes the pricing mechanism to pay for the SMS messages.

e-mail is time stamped yes, but its time stamped several times so unless you look at the full smtp headers the time it arrived at your mailbox or mail provider can still be different than when it was sent.
I run into that a lot because my e-mail is trhough a small company I found in AZ and I'm in OH so the time zones and time stamps are always crazy, the times that show on on the mail someone recives from me will show some crazy time.

The device probably has logic to work with an intermittent signal. Which means it probably has buffer logic, and also since you're dealing with a satellite system, compression. So, instead of you sending out:

HELP
OK
HELP
OK
HELP
HELP
HELP
OK

Your message goes out as

OK - 3
HELP - 5

Not very useful.

That button on the front is for human interface into the logic on the chip. That chip is a "black box" to you. You have no idea what it does.

I think that the time stamp is actually the GPS time derived from the SPOT device. I've just had a quick look on the SPOT website and the update rates are set at about 5 minute intervals of the 911 type message and 10 minutes for the Tracking messages, so to send a message as described would have a baud rate about 0.0033 Baud (bits per sec) or to put it another way compossing a message length of 120 letters (average 4 bits/character) would take around 40 hrs to send. Perhaps someone should tell the NSA.

http://www.amfearliathmor.pwp.blueyonder.co.uk/morse_message.mp3

That´s interesting info, guys. Thanks. It seems to be more complicated than I thought. But if the original timestamp could have been read I think transmitting at least one word would be successful. The only way to find out is when someone who owns SPOT decides to try it. But I agree it´s not practical to communicate like this. It would be rather experiment for fun.

First of all, you have to assume the messages are recieved in the correct order. I'm not completly familiar with the Spot, but as I understand it, the messages are sent as emails which is a store and forward system and is queued on various email servers along the way. They may or may not be delivered in order.

Assuming they did come in order, for your simple example, it could also be "tete", "nn", "ke", "tr", or the beginning of "!". Add more characters and it becomes even more complex. With the absence of proper spacing between characters it would be difficult to make out what text you're trying to send.

I see someone already beat me to it, without someone on the receiving end knowing your touch on the keypad it would just come out ..---..--....-...-...-.....-..-...-...-...-...-....-....-

It could be solved. I would wait say more than 30 seconds before transmitting next character. The software processing and decrypting the e-mails would then recognize it from the time stamp and would add the spaces.

While this might even work, since the application is a nearly life-or-death usage relying on observed behavior of the system in a test or two might be rewarded by the system changing its behavior later when you need it the most.

Using undocumented features is hazardous to your intentions. (I've seen some computer designs fail in the field because designers used the performance they found in the sample parts but weren't in the spec. sheets for the production parts.)

I agree. If it wasn't designed to do it. Don't try.

Think about this: Denial of Service (DoS) attacks are common these days. The ground station that processes your message could stop listening to you because it thinks your device is malfunctioning.