Re: [SOLVED + CAVEAT] [libreoffice-users] rounding problem with Base

2013-12-30 Thread Tom Davies
Hi :)
You can make a Feature Request usignt he bug-report system
https://wiki.documentfoundation.org/QA/BugReport
Just hunt around the drop-downs to find the one that has Feature
request at the bottom of it's list.  At least i think it's at the
bottom, they might have moved it up in the last couple of years.
Regards from
Tom :)

On 29 December 2013 19:48, Ivan Sergio Borgonovo m...@webthatworks.it wrote:
 On Fri, 27 Dec 2013 19:06:58 +0100
 Ivan Sergio Borgonovo m...@webthatworks.it wrote:

 To make the change in format permanent you've to build a view from the
 query and change the format there otherwise if you change the format in
 the query the format is lost next time you run the query.

 A bit weird.

 If I could give a suggestion I'd make the default format for numbers in
 queries follow the same schema as in postgresql (and most DB I know).

 NUMERIC(A, N) * NUMERIC(B, M) - NUMERIC(A + B, N + M)
 and FLOAT/DOUBLE all significant digits.

 On Fri, 27 Dec 2013 15:57:47 +
 Tom Davies tomc...@gmail.com wrote:

  Hi :)
  Are you saying you have solved this problem yourself without help
  from the list or has the problem just changed a bit?

 To summarize it for posterity:

 I wasn't able to appreciate if any change I was doing to the type used
 (NUMERIC(N, M) or FLOAT or DOUBLE) in SQL operations was effective
 since all I could get out from the DB was 2 decimals either in
 simple format (1234567898,31) or scientific format (6,23E+015).

 Not being able to see any change tricked me into thinking that no
 matter what type I was using there was no effect on precision since
 I'm used to something like:

 test=# select 1.01::numeric(4,2)*2.02::numeric(4,2);
  ?column?
 --
2.0402
 (1 row)

 The problem was just a problem of output format that can be set once
 you run the query, and in the result window right click on the column
 name and change the format.


 Thanks for the moral support anyway.




 --
 Ivan Sergio Borgonovo
 http://www.webthatworks.it


 --
 To unsubscribe e-mail to: users+unsubscr...@global.libreoffice.org
 Problems? 
 http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
 Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
 List archive: http://listarchives.libreoffice.org/global/users/
 All messages sent to this list will be publicly archived and cannot be deleted


-- 
To unsubscribe e-mail to: users+unsubscr...@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted



[SOLVED + CAVEAT] [libreoffice-users] rounding problem with Base

2013-12-29 Thread Ivan Sergio Borgonovo
On Fri, 27 Dec 2013 19:06:58 +0100
Ivan Sergio Borgonovo m...@webthatworks.it wrote:

To make the change in format permanent you've to build a view from the
query and change the format there otherwise if you change the format in
the query the format is lost next time you run the query.

A bit weird.

If I could give a suggestion I'd make the default format for numbers in
queries follow the same schema as in postgresql (and most DB I know).

NUMERIC(A, N) * NUMERIC(B, M) - NUMERIC(A + B, N + M)
and FLOAT/DOUBLE all significant digits.

 On Fri, 27 Dec 2013 15:57:47 +
 Tom Davies tomc...@gmail.com wrote:
 
  Hi :)
  Are you saying you have solved this problem yourself without help
  from the list or has the problem just changed a bit?
 
 To summarize it for posterity:
 
 I wasn't able to appreciate if any change I was doing to the type used
 (NUMERIC(N, M) or FLOAT or DOUBLE) in SQL operations was effective
 since all I could get out from the DB was 2 decimals either in
 simple format (1234567898,31) or scientific format (6,23E+015).
 
 Not being able to see any change tricked me into thinking that no
 matter what type I was using there was no effect on precision since
 I'm used to something like:
 
 test=# select 1.01::numeric(4,2)*2.02::numeric(4,2);
  ?column? 
 --
2.0402
 (1 row)
 
 The problem was just a problem of output format that can be set once
 you run the query, and in the result window right click on the column
 name and change the format.
 
 
 Thanks for the moral support anyway.
 



-- 
Ivan Sergio Borgonovo
http://www.webthatworks.it


-- 
To unsubscribe e-mail to: users+unsubscr...@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted