Page 1 of 2
[Implemented] Flags EditorGadget and double float
Posted: Tue Oct 14, 2003 4:19 pm
by JoRo
I was surprised to see, that in version 3.8 there are no flags for the editorgadget.
Double Float
I thought, it is not necessary to point to, I am really not the first. But until today, it is not possible to develop business application. They need doubles, for financial calculation.
Johannes
Re: Flags EditorGadget and double float
Posted: Tue Oct 14, 2003 4:36 pm
by Berikco
JoRo wrote:I was surprised to see, that in version 3.8 there are no flags for the editorgadget.
Double Float
I thought, it is not necessary to point to, I am really not the first. But until today, it is not possible to develop business application. They need doubles, for financial calculation.
Johannes
The business applications i made never used floats or doubles in 19 years programming.
Always worked internaly with cents, and divided by 100 to display results.
This was always more accurate for my usage.
Could be different in your situation of course.
Posted: Tue Oct 14, 2003 5:53 pm
by freedimension
Doubles are more than needed, but not for business applications.
Scientific applications are in bad need of doubles, also some API Calls need or return doubles (Integers mostly).
Posted: Tue Oct 14, 2003 6:32 pm
by Berikco
freedimension wrote:Doubles are more than needed, but not for business applications.
Scientific applications are in bad need of doubles, also some API Calls need or return doubles (Integers mostly).
Exactly my point
Scientific applications, and some API calls of course, but i did never see a need to use in a financial app.
Posted: Tue Oct 14, 2003 7:29 pm
by Rings
and never for industrial appz.....
Posted: Tue Oct 14, 2003 7:35 pm
by LarsG
Berikco wrote:freedimension wrote:Doubles are more than needed, but not for business applications.
Scientific applications are in bad need of doubles, also some API Calls need or return doubles (Integers mostly).
Exactly my point
Scientific applications, and some API calls of course, but i did never see a need to use in a financial app.
Rings wrote:and never for industrial appz.....
"Never" is a really strong word here guys
It all depends on <i>which type</i> of financial or industrial application you're supposed to make.
-Lars
Posted: Tue Oct 14, 2003 10:47 pm
by freedimension
LarsG wrote:It all depends on <i>which type</i> of financial or industrial application you're supposed to make.
An exact and reliable one !?
Posted: Tue Oct 14, 2003 10:54 pm
by LarsG
freedimension wrote:LarsG wrote:It all depends on <i>which type</i> of financial or industrial application you're supposed to make.
An exact and reliable one !?
That could be it.. yes!
-Lars
Posted: Wed Oct 15, 2003 6:53 am
by JoRo
Not sure, what is the right in english, selling/building houses, facilcity management, financial calculations. I was teacher for this, and my pupils had to buy an pocket calculator with 12 numbers display or more for school, not for business.
Great numbers are reached soon, what are 10.000.000 for a calculation for an area with 50 one-family houses, or the calculation for a fond of an market centrum?
By the way, over the half of the software, which is in use in this market, is not able to calculate right. that means a lot of trouble for the users. The only one I know here in germany with total acceptanz from the tax office is "Datev", some others specialized like "SAP" too.
I really cannot understand, what sens this restriction makes in PB.
I also would like to continue my mathematical work on fractals, espacially the mandelbrotmenge, but not with this floats.
Johannes
Posted: Wed Oct 15, 2003 8:06 am
by Rings
My job is to build industrial software (since 1992) and i know all pro's and cons of using floats/doubles in software.if you can avoid the use of floats then do it.Dealing with Longs is the easiest way as Berikco describes.
btw, wishing doubles is also an old post from me in autumm 2001 when i began with PB....., and is already on my wishlist for Scientific appz ....
JoRo wrote:.....I was teacher for this.....
hehe, teacher sucks everytime ......
Posted: Wed Oct 15, 2003 10:09 am
by Psychophanta
I can see pro's of existence of double precision and extended double precision floats, or even 128bit precision numbers in PB.
But the only "con" i can see is the lower speed of calculations in front of single precision ones.
Posted: Thu Oct 16, 2003 11:08 am
by Road Runner
Psychophanta wrote:But the only "con" i can see is the lower speed of calculations in front of single precision ones.
The FPU does everything in extended precision anyway, there is no speed difference.
Double Precision Float
Posted: Mon Oct 20, 2003 10:08 pm
by mikew
It was just yesterday when a bootstrap algorithm diverged because of a precision problem of the floats in Purebasic. As a scientist, I would urgently need the double precision floats.
Posted: Mon Oct 20, 2003 11:11 pm
by Karbon
I don't think anyone is arguing that they're not needed - just that they're not needed for money calculations!
Yes, we do need several variables types not currently in PB!
Posted: Tue Oct 21, 2003 6:52 am
by JoRo
May be, you have convinced me, that in financial calculation the way with integers is the better one. Yes, there a less problems.
That does not affect the necessarity of doubles, fo rintergs and floats.
But no one interested in the editorgadget? I have just problems to get a second click recognoized. I have to use the Api GetFocus_ inside my WaitwindowEvent, and then filter out, all the events that can occur. I am not able to set some styles, which are not changeable after creating. The functions for the Editorgadget are stringorientated, but it is typical for an richedit to contain more then 64K text. It needs string independent functions, like those of the Api.
Johannes