On Fri, 26 Nov 2004 , Ken Ray wrote:
On 11/26/04 11:25 AM, "Wilhelm Sanke" <[EMAIL PROTECTED]> wrote:
> The problem - on our side - is located in the "refresh" handler of the > card script of the Control Browser. Presumeably the engine now treats > the number of layers in groups somewhat differently since version 2.6.1 > (Metacard reading).
Can you ellucidate, Wilhelm? What exactly changed?
Just curious...
Ken Ray Sons of Thunder Software
Ken,
what really changed escapes me, but it must be connected to the manner of reading the numbers of layers in a group.
You can easily see that for yourself. Add two lines in the part of the script determining the width of the indents, like in the following script
if word 1 of the name of control i is "group" then
if pad is empty then
put the number of layers of control i into groupcontrols
# ===========new line for testing================= put CR&groupcontrols after last line of fld "test groupcontrols1"
#===================================
else
subtract the number of layers of control i from the last item of groupcontrols
put comma & the number of layers of control i after groupcontrols
#====second line for testing==========================
put CR&groupcontrols after last line of fld "test groupcontrols2"
#===============================================
end if
put " " after pad
end if
if pad is not empty then
repeat while pad is not empty and last item of groupcontrols <= 0
delete last item of groupcontrols
delete char 1 to 2 of pad
end repeat
if pad is not empty then subtract 1 from last item of groupcontrols
end if
and you will see that the results in both fields "test groupcontrolsX" differ with engines before and after version 2.6.1.
E.g., I got a card with 3 groups - and a couple more controls in between - the first two groups contain 2 controls each, the third group is a nested one with 4 controls altogether.
The result of the testlines is:
engine < 2.6.1
2 2 4 for "test groupcontrols1"
and
1,2
for "test groupcontrols2"
whereas for engines =< 2.6.1 the respective results are
2 4
and
2,2 4,2
If you test the other "groupcontrols" values farther down you get similar deviations.-
Maybe this is a so far unnoticed engine bug introduced with engine version 2.6.1 and affecting somehow - I really cannot elucidate (this could be answered probably by Tuviah?) - how layers in groups ar accessed.
Regards,
Wilhelm Sanke <www.sanke.org>
_______________________________________________ metacard mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/metacard
