80-Bus News

  

May–June 1983, Volume 2, Issue 3











Page 54 of 59











54

Doctor Dark’s Diary — Episode 16.

Two "Pascal" compilers slug it cut! Is it Wirth reading this?

Resular readers will know that I like Pascal, and that I tend to go om at lendth about the virtues I think I see in the compiler I usually use. I must have hinted really well that I would like free things to review a while ado because the postman brought me a nice shiny new compiler ta Play withsy and I am sure I didn’t ray for it! This was a very nice surerise, as you can imagines and I raridly set about attempting to use it. It seems like a soaod idea ta compare and centrast the two Pascal systems I now haver so thats if vou are thinking abeut chansing from BASICr vou can get some sort of idea which of the tue would be better for your reurreoses. Obviously»: T am unable ta comment oon ether systems I have not vet seeny and this does nat imply any criticism of these others... Serves them right for not sending me freebies though!

The one I alreacty owned is the Hiseft Pascal 4 comeilerr sold for £46. The new ene is the Poly-IlIata Comeas system, which costs £124. elus VAT. To save a little spacer, I intend to refer to them as either Hiseft or Compas throughout. Tf vou buy the Hiseftry vou det a floerey disc and about fifty five rages of manugi,y loose eases, that is. Compas also comes on a florry cliscry but has a very smart blue folder which contains, at & Suess,» a hundred eases. Both of them run under CP/M, sa don’t get either if vou are running an unexreandec

Nascom 1.

Hands up all these who wondered why TO rut aucetation marks reund the word "Pascal" in the heading. Why? Well neither of these compilers is a full implementation of standard Fascals and both of them have extras that are not standard. I don’t think there is a full micre Pascal about vetr but am willing to acccert ome free if I’m wrong! What vou get are subsets of the lansuader with additions that the compiler writer thinks you will want. Hiseft only allows FILEs of CHARe like CE/M’s ASCII files. It dees net cater for VARIANTS in RECORD tyres. Comeas doesn’t have SET» POINTER: or CHAR tyres. Neither of then will let vou pass a Procedure or aio function aS @ Faramneter to an mather precedure or functions but I have never particularly wanted to do this latter thing! In facts I have never met anyveme whe could tell me in nice shert words why anyene would want to complicate a nice simrle language by doing such a bizarre thing. I dare sav it is all the rage in university computing departments... The SET tyres are somethins auite usefuls which is net the impression given in many books abeut Pascal, where examples often feature lists of cheeses. POINTERS are very useful, and their absence from Comeas indicates that dynamic data structures are alse missing. So if you want to use Linkedci Llistsr and build them ure dynamically: wou have to go about it in a less easy way than would be the case with Hisoft. Io will describe the variaus extras facilities the two compilers erovide later on in the article.

Hiseft only provide a comeilery which resuires ASCII text files as its ineuty and it -is ur te vou what vou use to eit them. Any reasonable word Frocessar will de the Job» and only a masochist would try toa use ED.COM for the ioe Sse there ais a hidden cost. Compas is a more complex system, designed ta make program development faster. It has a built in editors and can stay in memory alend with the source pregram and the comeiled result Cassuming there is room.

So yeu can edit yvour program, compile ity and then run ity all without returning toa CP/M. Learning to use a new editer is alwavs a froblems ancl at first it always seems as though the designer has done aut of his way ta choose

really unlikely kevs to give the ecitor its commands. The Compas ecditer still SEEMS very Strande to me, as it has two modes of operations and I always seem to have ait in the wren one. The other erablem is that the list of ecliter commands in the manual is serarate from the List of which keys vou have toa press. This is presumably te make it easier for Poly-Data toa change the manual to suit machines with different kevboards. I think it would have been &@ more friendly idea to put the two sections together, anc write a serarate manual for each machine the program is to be usec on. To be fairy this would not he a Problem once one became used to using the editarr but at first it is infuriating.

SSE AINE oy

Birger –


This is an OCR’d version of the scanned page and likely contains recognition errors.











Page 54 of 59