Re: K.I.S.S. Factor (Re: Numerics)
Dominique Colnet <[email protected]>
| Newsgroups | gmane.comp.lang.eiffel.smalleiffel |
|---|---|
| Message-ID | <[email protected]> |
Francois Colonna wrote: > Colnet Dominique a écrit : >> >>> As has been stated, this complicates libraries such as the OpenGL, >>> where everything either has to be 'cast' to a REAL_32 on the Eiffel >>> side, or left for OpenGL to cast. In the old scheme of things there >>> was a direct correspondence between a 32 bit 'C' float and REAL, and >>> it is disappointing that SE has complicated that. >> >> >> Sorry to have complicated your job for OpenGL. It is difficult to >> introduce new types (i.e. REAL_32, REAL_64, REAL_128, REAL_80 and >> REAL_EXTENDED) whithout changing our habits. Even for us, it was an >> important work to update our general purpose library. > > > > why not to have compilation options to allow this kind of compatibilty ? Because one library can be correct with one compiler switch while another library could be incorrect with the same compiler switch ! If you want to use those two libraries in the same executable, you can't (and don't ask us to switch compiler rules each 2 lines of code ;-). More seriously, the compiler switch solution is the worst one because it promotes libraries incompatibilities. Furthermore, we do not want to maintain several version of our SmartEiffel/lib library. Best regards as usual, -- -------------------------------------------------------------- [email protected] -- IUT (Nancy 2) -- INRIA Lorraine http://SmartEiffel.loria.fr -- The GNU Eiffel Compiler POST: Loria, B.P. 239,54506 Vandoeuvre les Nancy Cedex, FRANCE Voice:+33 0383913140 Mobile: +33 0665362381 Fax:+33 0383581701