Re: Maybe there is an error on this?

It would seem the simplest solution is to use decimal(32,16) - or whatever
you need for the precision, and to use a FORMAT for the display ?

As I said - I've worked on many many commercial applications - and I've
never used a decimal(x) in any of them..


It might be easier to fix the 4gl compilers to detect decimal(x) and
replace it internally with a decimal (32,16) [ instead of the current
(x,2)] and generate a format for the form fields or something ?


Failing that - I think it would need some considerable development to add
in the proper functionality ...







On 5 December 2013 12:57, Alain Siverly <[email protected]> wrote:

>  ok, then you are a lucky guy if you do not need it :-)
>
> we have 470 columns on the database defined as "decimal(16)" and to change
> so much columns on customer systems takes too much time so we have a kind
> of major problem.
>
> we also have 735 variables defined as decimal(16) in the 4gl code and if
> we have to change it, then we must use one variable for display values on
> screen forms and another for calculation puposes, instead of only one
> variable for both.
>
> the reason of using decimal(16) is to show 2 decimal numbers on screen
> forms (which is common on commercial applications) but calculate internally
> with more precision.
>
> which kind of definition do you recommend in order to get the most
> precision, as an alternative to decimal(16)?
>
> cheers
> alain
>
>
> Am 05.12.13 13:04, schrieb Mike Aubury:
>
>   Yes - internally - a DECIMAL(x) defaults to DECIMAL(x,2) for Aubit4gl
>
> This is a known limitation
>
>
>        |       KW_DECIMAL OPEN_BRACKET INT_VALUE CLOSE_BRACKET {
>
> $$=new_simple_variable(NULL,DTYPE_DECIMAL,atoi($<str>3)*256+2,0);
>                         add_feature("DEFINE_DECIMAL");
>
>                 }
>
>  It would be quite difficult to fix this one - as it would be quite far
> reaching.
>
> Personally - I have never used a DECIMAL(x) - only ever a DECIMAL(x,y)
>
>
>
>
> On 5 December 2013 11:58, Alain Siverly <[email protected]> wrote:
>
>>  Hello,
>>
>> in the attachement we show an example where a DECIMAL(16) variable rounds
>> reduces the decimal numbers to 2 digits internally. 4js uses internally all
>> digits but shows only 2. Also Informix seems to act the same way as we
>> found in the I-4GL manual:
>>
>> "For DECIMAL(p,0) data types, any significant digits
>> (within the declared precision) in the fractional part of a data value
>> are stored
>> internally, but are not displayed. For example, if a data value of 0.25
>> is stored
>> in the DECIMAL(25,0) variable Q, the value of Q is displayed as zero, but
>> the
>> Boolean expression (Q > 0) evaluates as TRUE."
>>
>> If this is an error, could it be possible to get a correction on A4GL?
>>
>> Thanks in advance!
>>
>> Cheers
>> Alain
>> VENTAS AG
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Sponsored by Intel(R) XDK
>> Develop, test and display web and hybrid apps with a single code base.
>> Download it for free now!
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Aubit4gl-discuss mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/aubit4gl-discuss
>>
>>
>
>
>
> ------------------------------------------------------------------------------
> Sponsored by Intel(R) XDK
> Develop, test and display web and hybrid apps with a single code base.
> Download it for free now!
>
> http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk
> _______________________________________________
> Aubit4gl-discuss mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/aubit4gl-discuss
>
>

------------------------------------------------------------------------------
Sponsored by Intel(R) XDK 
Develop, test and display web and hybrid apps with a single code base.
Download it for free now!
http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk

_______________________________________________
Aubit4gl-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/aubit4gl-discuss