RE: Subroutines vs Subprocedures was RE: Indicators

Lim,

what is your point? It was you who brought up the question of subprocedures, 
that may call subroutines in the main line or other subprocedure to initialize 
global variables.

I think, you have to redesign your application, if you want to do such a thing. 
Initialing global variables should be done at the beginning of a loop 
(iteration) in the main line, either in a subroutine or a subprocedure.

Because if you want to call a subroutine that is outside a subprocedure, the 
simplest solution is to transform that subroutine into a subprocedure. So it is 
accessable from any part in the programme.

And personally, I do not see a difference in a subroutine or a subprocedure. 
Both have their usefullness, depending on the structure of the code, the 
problem to solve. For me, a subroutine can be called from any part in the code 
as well (hence my dislike of opcode LEAVESR: where will I jump back to?); I do 
not need a subprocedure for that.

Just my 2 Euro-cents.

Regards,
Carel Teijgeler.

*********** REPLY SEPARATOR  ***********

On 24-8-04 at 8:52 Lim Hock-Chai wrote:

>Joe P and Bob C:
>It is true that using you save some typing time when using sub-routine. 
>However, the problem comes when sometime in the future you happen to have to 
>create a sub-proc and it really need to execute one >the of sub-routine that 
>was created because of you want to save some tying time.
>
>Joe L:  
>For some reasons you seems to think that it is reaaaally bad to access/update 
>global variables from sub-proc.  I don't know about >your shop, but if I apply 
>such a rule in my shop, I don't think I will be using sub-proc very often at 
>all.  All fields from F spec are >global.  In most case, I need to use them in 
>most of my sub-proc.  If would be quite a pain if I've to pass them in every 
>time I need >them. I'm not saying that it is good to access/update them in 
>sub-proc, but if you design it correctly, it should be fine.