On Thu, Nov 1, 2012 at 3:03 PM, Carlos H. Cantu <lis...@warmboot.com.br> wrote: > DY> The proper solution is far not trivial to implement. > > Ok, but I understand that even not being trivial, it can be > implemented :) > > DY> An easier one will break existing applications as they will start > DY> calculating different numbers for the same queries. > > I think his is a relative point of view... I mean, people using > numeric/decimal with dialect 3 is already using "workarounds" due to > the currently existing problems/limitations (ie: using intermediate > casts, etc). So, if the change would cause different results but with > greater precision, I doubt anyone would get mad. > > Does any study was already made to check how different the results > would become? A result of 1.000004 becoming a more precise one (ie: > 1.0000049) can be acceptable, while 1.000004 becoming 1.000010 may not. > > Of course, any change of this type would need to be very well > documented and announced.
Maybe is time for Dialect 4 with all the Dialect 3+1 fixes ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_sfd2d_oct Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
