|
|
Post by Stefan Pendl on Apr 24, 2022 2:42:03 GMT -5
right$() is not trimming the string, just concatenating in the code below.
'calculate pi.bas
'Author: Stefan Pendl
'Date: 21.04.22
n=-1
p=0
npi=0
pi = Asn(1) * 2
count=0
difference=1/10
row=2
startTime=time$("ms")
while difference>0.000001
do
n=n+2
if p=0 then
npi=npi+4/n
else
npi=npi-4/n
end if
locate 1,row
print npi
p=not(p)
scan
count=count+1
loop until abs(pi - npi) < difference
endTime=time$("ms")
elapsedTime=endTime-startTime
if elapsedTime > (60*60*1000) then
elapsedHours=int(elapsedTime/1000/60/60)
elapsedTime=elapsedTime-(elapsedHours*1000*60*60)
end if
if elapsedTime > (60*1000) then
elapsedMinutes=int(elapsedTime/1000/60)
elapsedTime=elapsedTime-(elapsedMinutes*1000*60)
end if
if elapsedTime > 1000 then
elapsedSeconds=int(elapsedTime/1000)
elapsedTime=elapsedTime-(elapsedSeconds*1000)
end if
print count; " calculations needed to calculate pi"
print " to a difference of "; difference; " from "; pi
print " calculation took ";elapsedHours;":";right$("0";elapsedMinutes,2);":";right$("0";elapsedSeconds,2);".";right$("00";elapsedTime,3)
difference=difference/10
row=row+5
wend
end
Thanks in advance for fixing.
|
|
|
|
Post by Carl Gundel on Apr 25, 2022 9:00:53 GMT -5
right$() is not trimming the string, just concatenating in the code below.
Help me out. This doesn't work as expected? start$ = "0123456"
for x = 1 to 6
print right$(start$, x)
next xAlso, I tried your example. Aside from how slow it runs because of the sluggish mainwindow, it seems to work fine. |
|
|
|
Post by Brandon Parker on Apr 25, 2022 21:18:48 GMT -5
Carl,
I think the issue that Stefan is trying to point out occurs when one attempts to concatenate within the first parameter to the Right$() function call. In Stefan's example, the milliseconds' value should be left-truncated to the rightmost three digits after concatenating 00 to the beginning of it. So, if the millisecond value is 17, we should be seeing 0017 left-truncated to 017.
The code below works as expected in LB 4.5.1, but it fails in LB 5 (output shown)
Here is my test code...
start$ = "0123456"
For i = 1 To 8
Print Right$(start$;"a", i)
Next i
Print
For i = 1 To 8
Print Right$("a";start$, i)
Next i
LB 4.5.1 Output
LB 5 Output
I hope this helps...!
{:0)
Brandon Parker
|
|
|
|
Post by Carl Gundel on Apr 26, 2022 8:19:17 GMT -5
Carl, I think the issue that Stefan is trying to point out occurs when one attempts to concatenate within the first parameter to the Right$() function call. In Stefan's example, the milliseconds' value should be left-truncated to the rightmost three digits after concatenating 00 to the beginning of it. So, if the millisecond value is 17, we should be seeing 0017 left-truncated to 017. The code below works as expected in LB 4.5.1, but it fails in LB 5 (output shown) Thanks Brandon. Seems to be a subtle bug in the way that semicolon concatenation is happening. The code actually works in Run BASIC, on which LB5 is based, so some sort of regression since that work was done. Using + instead of ; works as expected. It's interesting because I never use ; in these kinds of coding, only +. In my mind use of the semicolon is only for PRINT or for coding with #handles, because that is also derived from PRINT. I'd be tempted to remove support for ; in non PRINT situations, but that horse has already left the gate. start$ = "0123456"
For i = 1 To 8
Print Right$(start$+"a", i)
Next i
Print
For i = 1 To 8
Print Right$("a"+start$, i)
Next i |
|
|
|
Post by Carl Gundel on Apr 26, 2022 8:35:05 GMT -5
Thanks Brandon. Seems to be a subtle bug in the way that semicolon concatenation is happening. The code actually works in Run BASIC, on which LB5 is based, so some sort of regression since that work was done. I went back to the very first publish of LB5 in my source control system and bug is still there. LOL Interesting. More work for me. |
|
|
|
Post by Carl Gundel on Apr 26, 2022 23:04:43 GMT -5
Thanks Brandon. Seems to be a subtle bug in the way that semicolon concatenation is happening. The code actually works in Run BASIC, on which LB5 is based, so some sort of regression since that work was done. I went back to the very first publish of LB5 in my source control system and bug is still there. LOL Interesting. More work for me. Figured it out. It'll be a little tricky, but it's gonna be just fine.  |
|
|
|
Post by Brandon Parker on Apr 27, 2022 20:22:46 GMT -5
Yeah, hopefully, it will not be too difficult...
Just keep plugging away at it.
{:0)
Brandon Parker
|
|
|
|
Post by Carl Gundel on Apr 28, 2022 16:47:05 GMT -5
Yeah, hopefully, it will not be too difficult... Just keep plugging away at it. It's easy to see what's happening. The cleanest and more elegant solution, not so easy. |
|
|
|
Post by Stefan Pendl on Apr 30, 2022 13:30:10 GMT -5
Using + instead of ; works as expected. It's interesting because I never use ; in these kinds of coding, only +. In my mind use of the semicolon is only for PRINT or for coding with #handles, because that is also derived from PRINT. I'd be tempted to remove support for ; in non PRINT situations, but that horse has already left the gate. I use the semicolon as a shortcut to "+str$()" to concatenate strings and numbers. Sorry for not providing a simple example. Below find a simple example to demonstrate the difference between LB4 and LB5.
num = 12345
for i = 1 to 6
print right$("n"; num, i)
next
print
for i = 1 to 6
print right$("n" + str$(num), i)
next
end
|
|
|
|
Post by Carl Gundel on May 17, 2022 13:17:19 GMT -5
There is a fix for this in build 354.
|
|
