I just barely learned that IMAP is a protocol that allows you to read and respond to emails without using the internet. Would it be possible to use this application for packet radio and if so, how? I am using the Ubuntu 14.10 operating system.


IMAP is a protocol for transferring email, nothing more, nothing less. It typically uses TCP/IP as the underlying protocol, but it could be routed over something else (like packet radio). If you wanted to, you could configure TCP/IP over AX.25 but I don't know if anyone else is doing it, so you'd probably need to run your own mailserver on the other end as well. (You could even configure that mailserver to bridge to the internet, but keep in mind that you'd be responsible for any incoming traffic that made it onto amateur frequencies!)

What you can't do is use Thunderbird to interact with any typical packet BBS or mailbox systems, they work with different protocols. So essentially, what you ask is possible, I'm just not sure if it's useful.

  • 1
    $\begingroup$ Just an addition. Most networking models follow the OSI (Open System Interconnection) model. It describes several interrelated layers. Both theoretically and in practice one can change the lower layers without affecting the upper layers. This is why you see your e-mail application working on your PC while using wired cable Internet, but it also works while you have a wireless WiFi (802.11) Internet conection.<br><br>"Thunderbird email application" will fit in the Application layer. TCP/IP will be in the Transport layer. AX.25 will be in the Network layer. $\endgroup$ Feb 26 '15 at 22:36
  • 1
    $\begingroup$ @sessyargc.jp Yeah, except that AX.25 doesn't conform to the OSI model. Neither does TCP/IP, strictly speaking. TCP/IP over AX.25 performs especially horribly for this reason: they aren't really designed to layer together. $\endgroup$ Feb 27 '15 at 11:58
  • 2
    $\begingroup$ @sessyargc.jp Actually, in the OSI model, layer 7 isn't the software application; that is outside of the scope of OSI. Layer 7 (the so-called "application layer") is where application protocols live; for example IMAP, HTTP, SNMP, FTP, SSH, etc. A usable software application is going to implement one or more of the L7 protocols and thus, if you want to shoehorn it into the OSI model, would be something more like layer 8. Wikipedia: both the OSI application layer and the user interact directly with the software application. en.wikipedia.org/wiki/OSI_model#Layer_7:_application_layer $\endgroup$
    – user
    Mar 2 '15 at 7:35
  • $\begingroup$ Note that John Wiseman's BPQ32 BBS software provides some notional connectivity with traditional email clients. cantab.net/users/john.wiseman/Documents/… $\endgroup$
    – JSH
    Jul 26 '16 at 15:04
  • $\begingroup$ @sessyargc.jp actually basically nothing has ever used the OSI stack, which included things like X.213 instead of TCP and X.400 instead of SMTP. Everyone just decided that its layer model would be sort of useful to describe their own stack (except when it doesn't :) $\endgroup$ Mar 20 '18 at 4:05

While I'm not very experienced with packet radio (it's on my list of projects), I can comment in general on using IMAP on low-bandwidth connections. I have extensive experience using it on dial-up TCP/IP connections (PPP), which are much slower than modern broadband connections and have much higher latency. These are aspects it has in common with packet radio connections, although packet radio is much slower still (I was getting about 40kbps down and 28kbps up on dialup, and packet is either 1200 or 9600 bps bidirectionally and single duplex).

IMAP is ... well ... nearly unbearable on low bandwidth and high latency connections. It takes very long periods of time to negotiate connections and download messages. Watching the traffic using a packet analyzer (tcpdump), the protocol is very chatty. With packet radio being single duplex, I think using IMAP over packet radio would verge on being unbearable unless the amount of email in the IMAP account is very small, indeed.

That having been said, by all means, experiment with it and post your findings. I may just try the same to satisfy my instincts about it.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.