Hi, The question is rather silly but raised since earlier in discussion Jurgen mentioned what the workspaces is the outdate way to work with APL.
So what is the current (modern) ways supported by GNU APL would be? I could imagine the organization of the program the following: 1) set of UTF-8 text files containing functions (possibly 1 function per file) 2) one main script file loading these functions, setting the data variables and performing the execution. In the old ways it could be the main workspace which copies necessary functions from other workspaces as a part of commands from ⎕LX variable and performing the execution. In addition Dyalog APL provides something called SALT - Simple APL Library Toolkit - a way to organize APL objects as a separate files and manipulate them. From the documentation: "One of SALT's function, Snap , also enables the construction of a directory structure corresponding to the namespace structure of a workspace, where each file in the structure contains the script of an APL object in the workspace." -- Br, /Alexey
