A contribution to an improvement of PASCAL

The purpose of this paper is twofold° First of all we would like to present some of our proposals, concerning the desirable corrections in the Revised Report on Pascal and possible slight extensions of the langusgeo Secondly we want to argue with some of the critical remarks on Pascal as formulated several months ago by Conradi [1]~ We have experience with Pascal both as teachers and implementors~ We like the language, find it very useful for teaching programming an@ also as a tool for systems implementation. That is why we would like to improve it by indicating some of the weaker points in the language and its description, together with our proposals for changes, corrections,and possible extensions-without, however, disturbing the essence of the language. During our discussion we shall quite frequently refer to the Pascal User Manual ([2], part I, later referred to as UM) and the Report ([2], part 2}~ It happens that the Report is incomplete on several points~ some of them are explained in the UM, which unfortunately must be considered as a deficiency of the basic document defining the language. The Report is our basic source where Pascal semantics is concerned~ where it is neoessar~ we try to clarify possible doubts with the help of UM (and when ao anner can be found there, Pascal 6000 compiler can be used as an ultimate oracle, although we era not happy wlth such a solution}. 2e Some vroposals for changes aria_ Imvrovemen~s in First of all, we would like to suggest some corrections in the text of the Report. Our idea is to improve certain-obviously inaccurate-definitions! presented solutions reflect our understanding of intentions behind these particular constructs, 2.1. The Report is quite imprecise when defining the circumstances in which-having a variable and an expression with~dentical types-an assignment statement is correct (the same applies to the formal-actual parameter correspondence rules in the ease of cell by value).