[Development] FP calculations and stability in Qt

[André Pönitz]

On Mon, May 13, 2019 at 12:53:42PM +0300, Konstantin Shegunov wrote:
> On Mon, May 13, 2019 at 12:42 PM Konstantin Ritt <ritt.ks at gmail.com> wrote:
> 
> > Writing and proposing a patch would take less time than discussing pros
> > and cons here.
> 
> Which I had in a minute fraction done, as a pass-by in the comments. I'm
> not, however, willing to take huge technical debt on the issue by investing
> copious amounts of time writing a specialized decomposition for this
> particular case, in this particular class, for this particular set of
> values. I'm not going to do it better than already done, is one, and
> secondly it's a huge time sink. Combine both to get: "dubious result for a
> big investment".
>
> What I'd really, really want, ideally, is to have a well tested tool(set)
> that I can rely on to do the job, whence my initial question: "can we talk
> about how feasible it is to pull something (or parts of it) in Qt, even if
> only internally, to facilitate stable fp calculation"

The answer is likely "Nobody will stop an attempt to talk, but given that such
issues have been around at least since Qt 4.0 (that's when I personally ran into
them) and the tendency lately is rather to not include well-known versions of $X
anymore but rather depend on random version of 'system' or auto-'updated' $X
this is unlikely to happen."

Andre'