The language was published in book form, with introductory sections followed by flowcharts and then the 8008 assembler code.
The book described ways to save more memory, turning off arrays for instance, and how the user could add their own new features to the language.
Arnold was friends with professors at the University of Wyoming (UW), and through them had arranged to have an account on their Sigma 7 mainframe computer.
Arnold entered UW in 1975 and contacted Nat Wadsworth, one of the founders of SCELBI, pitching the idea of a compiled version of BASIC for their new platform.
It took Wadsworth several months to finally arrange a contract, which included sending Arnold an 8B development system.
Arnold speculated that, lacking these delays, SCELBAL could have been released at about the same time as Altair BASIC in late 1975.
The ad did not specifically link the language to the SCELBI platform, instead, it simply offered itself in book form as a complete source listing to create a version of BASIC on any 8008 or 8080 system with the requisite 8 kB of RAM.
[3] The manual provides well-documented assembly code for the entire math package, including entry points and usage notes.
[8] Among the few other differences was that the NEW command found in MS, which clears out existing program code and data, is called SCR for "scratch",[9] and the USR function, which called a machine language routine, was UDF for "user defined function".
[10] UDF allowed a single floating-point parameter to be passed to the user-defined function, whose machine-language code must have been loaded into memory at a fixed location ahead of runtime.
SCELBAL also allowed omitting the semicolon, which specifies the characters from the starting point to the end of the string.
For instance, in Atari BASIC the similar-looking code:[16] would instead output "EL", as the instruction translates to "print all characters between positions 2 and 3".