$INSERT row changes (AREV Specific)
At 01 JUL 2004 04:29:03PM Ralph Johler wrote:
We are having a rather elusive problem and suspect it is caused by our "$INSERT filename, rowname" statements.
Does anyone know what happens when the "rowname" contents are changed AFTER the parent source code is compiled?
It appears that if we change the "rowname" contents, we need to recompile all the parent programs in order to avoid this elusive problems.
Our elusive problem is seemingly random conversion of arev records part of the record into the DOS format, part still in arev format. All the data is present, just goofed up into the first two columns (fields).
We use lots of DIM'ed arrays, which use the '$INSERT filename, rowname' for the reference to the data columns position numbers:
INSERT rowname contains
EQU NAME$ TO 1
EQU ADDRESS$ TO 2
EQU CITY$ TO 3
etc etc
Ralph
At 01 JUL 2004 05:04PM [url=http://www.sprezzatura.com]The Sprezzatura Group[/url] wrote:
Ralph,
Does anyone know what happens when the "rowname" contents are changed AFTER the parent source code is compiled? It appears that if we change the "rowname" contents, we need to recompile all the parent programs in order to avoid this elusive problems.
That is correct. Your equated values are 'hardcoded' into the compiled object code. They are not resolved at runtime.
World leaders in all things RevSoft
At 01 JUL 2004 05:27PM [email protected] wrote:
If you make a change to an INSERT record only those routines which have been recompiled will know of the changes. This is a pre-process of the compiler.
So be sure to recompile ALL routines, events, dicts etc which use a modified Insert record
[email protected] onmouseover=window.status=the new revelation technology .. a refreshing change;return(true)"
David Tod Sigafoos ~ SigSolutions
At 02 JUL 2004 09:26AM Hippo wrote:
May be a routine which searches all source codes for a given string is needed. I expect each AREV programmer created one for himself.
This was the first tool I've required.
At 02 JUL 2004 10:39AM Ralph Johler wrote:
Thanks! Seems pretty clear.
Now what would happen in this case:
Parent code uses $Insert row A
Parent code uses CALL Subparent One
Subparent One code uses $INSERT row A
We change $INSERT row A
We recompile Subparent One code, but NOT Parent code.
Could this condition perhaps cause the 'partial' conversion to DOS format of a data row?
At 02 JUL 2004 10:49AM Ralph Johler wrote:
Thanks - right you are there are more ways to look for code, then I would care to count!
At 07 JUL 2004 04:58AM Hippo wrote:
This means the subcode will interpret the table other way than parent code. I do not know what the code does, therefore I cannot say anything about the "DOS" conversion.