BACKGROUND
Starting in the 1950's, telephone companies were faced with a new challenge. Subscribers wanted to move digital data over their networks at high speed. Companies needing data communications links bought teleprinter/teletype systems (Telex) or other services. These were all slow, expensive, data only connections and were not suited to transferring large ammounts of data.
The analog voice lines of the time could be very noisy and the noise could come and go throughout the duration of the call. At one time it really was necessary to shout into the phone for a long distance conversation because of the noise added by carrying the signal long distances. If the line became noisy and the callers could no longer hear each other, they would simply shout "What did you say?" and the other party would repeat themselves. Sometimes the lines were so bad that even this did not work. The solution was to hang up and call again and hope you got a better connection that was less noisy. This issue with noisy phone lines remained true for most long distance calls all the way up until the early 1980's. Local calls weren't much better.
MODEMS
Throughout the 60's a cheaper solution to digital communication was under development by Hayes. Hayes was working on a device that modulated and demodulated a signal over the voice telephone system. Modems were developed to carry data over the voice telephone network. Thus, purchasing an ordinary voice line was sufficient to get basic data communication established. Modems were expensive, but modems were sold as being a "one-time expense" and thus were far cheaper than the other digital subscription solutions such as Telex; however, modems used the analog voice lines and were vulnerable to all the noise problems callers had. Modems trying to establish communication over such networks faced difficulties communicating and often dropped the call, usually in the middle of a large data transfer. Data file transfer protocols of the time (X modem, Y modem) had no provisions for resuming an interrupted data transfers and they were still slow (300 - 2,400 bps). If you began a data transfer over a long distance connection it could take several hours to transmit a large file. If the line quality degraded and you lost your connection two hours into the call, you had to start all over again and wasted money on a two hour long distance call. This could still be very expensive. Worse, it was a well known fact that the larger the ammount of data you needed to transfer, the more likely you were to fail. Early in the development of modems, data transfers over the voice telephone systems failed more often than they succeeded. This made a lot of people very angry at the phone companies and modem manufactureres.
ENTER X.25
One thing that must become clear when looking at the history of the telephone companies is that they move very, very slowly to solve problems. It took more than two decades to develop and then deploy X.25. Before X.25 was fully deployed, new technologies had arrived making it obsolete. The primary reason for the development of X.25 was capacity, not the capability to transfer digital data. Digitizing voice calls was simply the easiest way they could conceive of using a single trunk wire for multiple phone calls. The phone companies were spending considerable sums of money to dig new trenches and drop new trunk lines between central offices. If they were to retain their profit margins, they needed a solution that would allow them to use the existing trunk lines and equipment and still scale up the capacity. The solution was to packetize the voice calls and switch the packets, not the calls. Thus X.25 was developed to increase capacity. Because customers were literally screaming for higher quality phone lines (and some customers were demanding data transfer services) the phone company built X.25 to support these features as well. So that they wouldn't have to spend money for another three or four decades, they tried to forsee any possible developments and make X.25 capable of handling more than just voice calls.
The X.25 protocol was developed in the 1970's by the telephone companies to increase efficiency and to add the capability to carry digital data. Prior to X.25, calls were analog and circuit-switched between callers over a complete end to end circuit. These open calls tied up trunk lines that were part of that end to end circuit and thus blocked other calls between central offices thereby limiting the total capacity of the phone system. The X.25 protocol was built from existing technologies already in use at the time. X.25 links are digital and perform packet-switching. Packet-switching allows X.25 to carry multiple phone calls over a single link by breaking up each phone call into packets. Subscriber lines arrived at the central office where their calls were converted to digital information and then switched over the X.25 network to their destination. Only the subscriber's local loop remained open for the duration of the call and all the trunk links in between could carry multiple calls.