-
Submitted by: askln (askln)
SFID: 796675#⁠
Submitted By: askln
Given
CREATE TABLE t \(id INTEGER\);
the following statements give corresponding errors:
1\) SELECT ? FROM t =\> Data type un…
-
Currently, `QuotientRing_generic` has no method `random_element()`, thus the one of the class `Ring` is used, and only integers coerced into the `QuotientRing` are returned.
I add such a method, m…
-
### Proposal
I think we should update our hash functions and at the same time make some changes to the Hashable protocol. Ideas for these updates come from recent developments in languages such as …
-
In our last meeting we started talking about how to run tests, like should we compile stuff in release mode or development mode before running the test? There are multiple effects from the optimizatio…
-
The hazards of unsigned arithmetic in C/C++ are well known, and modern style guides generally discourage their use for anything except bitwise manipulation and cases where the extra bit of positive ra…
-
I propose that we add a rational type to either the stdlib or the core data types, replacing floating-point numbers in many cases.
# Rationals can complete integers without requiring lossy floa…
-
Consider:
```
constexpr float inf = 1.0f / 0.0f;
constexpr float nan = 0.0f / 0.0f;
```
Clang currently rejects these, while GCC accepts: https://godbolt.org/z/oG4sP9TMc
I believe Clang is wro…
-
The following valid fortran fails to be parsed
```fortran
program main
integer*4 bar
call foo((bar)*2)
end program main
```
Giving:
```
ProgramFilefortran-src: test.f,…
-
#### Is your feature request related to a problem?
It is currently possible to perform arithmetic on an `Interval` iff it is an `Interval` of numbers. For `Interval` `iv`, `iv - c == Interval(iv.l…
-
Any fraction whose denominator is not a power of 2, cannot be accurately calculated using native-float arithmetic, neither can it be accurately stored in a native-float variable.
For example, in th…