mirror of https://github.com/CGAL/cgal
57 lines
2.6 KiB
Plaintext
57 lines
2.6 KiB
Plaintext
Concerning the code itself:
|
|
|
|
- Fixed support : IA_Fixed.h
|
|
- SGI'CC bugs me:
|
|
"tst5.C", line 25: error(1282): more than one instance of overloaded function
|
|
"CGAL_compare_y_at_xC2" matches the argument list:
|
|
function template "CGAL_compare_y_at_xC2(const FT &, const FT &,
|
|
const FT &, const FT &, const FT &)"
|
|
function template "CGAL_compare_y_at_xC2(const CGAL_Filtering<CT,
|
|
ET> &, const CGAL_Filtering<CT, ET> &, const
|
|
CGAL_Filtering<CT, ET> &, const CGAL_Filtering<CT, ET>
|
|
&, const CGAL_Filtering<CT, ET> &)"
|
|
argument types are: (NT, NT, NT, NT, NT)
|
|
cout << "Result 1st test: " << (int)CGAL_compare_y_at_xC2(px, py, la, lb,
|
|
lc);
|
|
- Look at Stroustrup to learn what references are...
|
|
- Try to fix that warning:
|
|
/usr/ccs/bin/as: "/var/tmp/ccBHnU0T.s", line 861: warning: 3 instructions
|
|
required between ldfsr and next FBfcc; nops inserted
|
|
- overwrite overlap_action() so that it differs between IA and IA_advanced.
|
|
This implies declaring it as virtual... It first needs a bench on
|
|
predicates. It could be defined using template...
|
|
- Add preconditions on the rounding mode => CGAL_get_rounding...
|
|
? try to fix the assembly for Mips (is it Geert' egcs-1.0.3 problem ?)
|
|
? template by float/double ? (=> must provide CGAL_FPU... for floats...)
|
|
? implement operator{<,>,<=,>=}(double) -> the current operators will be
|
|
implemented using them, and the tests (IA) > 0 will be faster ?
|
|
Also, this would simplify the *() and /() code.
|
|
But be careful with exceptions... => 2 different semantics...
|
|
? shall I protect ostream operator<<() by CGAL_FPU... ?
|
|
it seems there's a pb with Linux & Irix (& Solaris ?) here.
|
|
Hum... it's not MY fault, is it ?
|
|
|
|
Concerning the script:
|
|
|
|
- Add a check for the return type (must be CGAL_sign... but not FT)
|
|
(it won't even compile in this case, I guess) ?
|
|
- Directory where to put the script, and add it to the "make package".
|
|
|
|
Concerning the doc:
|
|
|
|
- re-read the doc carefully.
|
|
- modified sqrt() behaviour when negative.
|
|
- document the change of behaviour in the comparison operators, and the effect
|
|
of CGAL_IA_DONT_THROW_EXCEPTION...
|
|
- describe the filtering scheme (new class CGAL_Filter -> new paragraph).
|
|
- Re-document CGAL_to_Interval_nt[_advanced] (they change regarding the
|
|
ronding mode).
|
|
? document the speed up with (double * IA).
|
|
? document is_same().
|
|
|
|
Concerning the test-suite:
|
|
|
|
- small benchs on a few predicates.
|
|
- test cast functions (are they dependant on rounding modes).
|
|
- test NaNs (and Infs?). Comparisons with these should throw the exception.
|