[2cfffe] | 1 | |
---|
| 2 | |
---|
| 3 | Use the compiler flag |
---|
| 4 | |
---|
| 5 | -DNTL_TRANSITION |
---|
| 6 | |
---|
| 7 | to help with the transition to NTL 3.0 from earlier versions. |
---|
| 8 | |
---|
| 9 | This has the effect of undeclaring certain functions |
---|
| 10 | whose semantics in 3.0 is different than in versions < 3.0. |
---|
| 11 | Thus, one can use the compiler to find the trouble spots. |
---|
| 12 | |
---|
| 13 | THE LIBRARY CAN NOT BE COMPILED WITH THIS FLAG! |
---|
| 14 | ONLY USE TO FIND TRANSITION PROBLEMS IN CLIENT CODE. |
---|
| 15 | |
---|
| 16 | Undeclares all shift operators for NTL arithmetic type; |
---|
| 17 | in versions < 3.0, << was a conversion operator; now it |
---|
| 18 | is a shift operator. |
---|
| 19 | |
---|
| 20 | Undeclares division functions in ZZX; in versions < 3.0, |
---|
| 21 | these were defined in terms of pseudo-division; now they are |
---|
| 22 | defined as ordinary division with an error being raised if the |
---|
| 23 | result is not integral. Explicit pseudo-division functions |
---|
| 24 | are now provided for the old semantics. |
---|
| 25 | |
---|
| 26 | Undeclares the UpdateMap function in for ZZ_pX and zz_pX; |
---|
| 27 | in versions < 3.0, the output always had length n; |
---|
| 28 | now high-order zeroes are stripped. |
---|
| 29 | |
---|
| 30 | Undeclares the conversion from GF2X to GF2EX functions; |
---|
| 31 | in versions < 3.0, this was defined as creating a constant |
---|
| 32 | polynomial by reduction modulo GF2E::modulus(); |
---|
| 33 | now, it is defined as a coefiicient-wise "lift". |
---|
| 34 | GF2X and GF2EX happened to be called BB and BB_pX in |
---|
| 35 | versions < 3.0. |
---|
| 36 | |
---|
| 37 | Declares assignment and copy for RR to be private. The |
---|
| 38 | semantics of these have changed from "copy and round to |
---|
| 39 | current precision" to "exact copy". |
---|
| 40 | |
---|