What were the major differences between first generation languages and second generation languages?

Answer 1

The main difference is that first generation languages (1GL) are low-level while second generation languages (2GL) are high-level. Before there were any programming languages everything had to be coded manually using machine code, the native language of the machine (0GL if you like). However, it was quickly realised that the computer itself could be programmed to assemble the machine code from a symbolic language which became known as assembly language, the first 1GL language. Each machine required its own assembler because the symbolic assembly language, like the machine code it produced, was machine-specific. That is, machine code could only operate upon machines of the same type.

1GL languages made programming easier, however every program still had to be designed and coded in minute detail. Thus more complex languages began to appear, languages which incorporated a greater degree of abstraction between the source code and the machine itself, allowing the production of many assembly instructions in just a few lines of "high-level" source code. These became the first of the 2GL languages. The programs that performed the translation from source code to assembly code became known as compilers (which were originally written in assembly language, of course). The resultant assembly code could then be optimised by hand before being fully assembled. However, as these new languages advanced, they began to incorporate their own optimisers and could emit the machine code directly. More importantly, as each machine had its own compiler, the same high-level source code could be "ported" to different machines. These became the 3GL languages, with C being the most prevalent until the introduction of C++ in 1983.

These days the terms 4GL, 5GL (and even 6GL!) are used to describe some new languages, but they are meaningless terms because once you have portability between machines (which was achieved with 3GL), there's really nothing new you can add to any language that can surpass portability. All new languages are just better 3GL languages! The "generation" labels are typically used by marketing types to make one language seem superior to another even when the perceived "enhancement" is insignificant or, in some case, completely non-existent! There are perfectly adequate 3GL languages that, through simple improvements and continued development, are fully capable of achieving the same ends as any so-called 4GL+ language. Languages are best categorised by their primary features: whether they are functional, empirical, procedural, structured, object-oriented, and so on, or by the fields in which they specialise, whether it be general purpose programming, business programming or Artificial Intelligence programming. These tell you far more about a specific language than any meaningless "generation" label ever can.