On Mon, Oct 21, 2024 at 04:29:19PM +0200, Jean-Marc Lasgouttes wrote:
> My conclusion is that it would only be nice to use with C++20 where one can
> use constinit to force a function to be computed at compilation time. Then
> we would be able to have:
> docstring const var = "sôme ascii text"_asc;
>
> This would not compile. Forcing the use of constexpr is annoying IMO.
Your call ;)
> > tl;dr: the lower intersection for LyX 2.5 gives GCC 8.5, Qt 5.12, Python
> > 3.6.
>
> So shall we enforce these gcc/qt versions?
Yes, I think we can bump to GCC 8.5 and Qt 5.12.
If someone wants to go through the sources and cleanup ifdefs we can switch
anytime.
By doing this we should get access to C++17, so my favourite new feature
highlights:
- if & switch with initializers: if (status_code c = foo(); c != SUCCESS) {...
https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2016/p0305r1.html
- structured bindings to get direct access inside pairs & maps in loops: for
(const auto& [key,val] : myMap) {
https://www.cppstories.com/2022/structured-bindings/
> Isn't it Python 3.8?
My assesment for LyX 2.5 was P3.6, but the reality is that we already bumped to
P3.8 AFAIU.
On old system people will need to install P3.6 in parallel (at least that
should be
simpler than Qt/GCC).
Pavel
--
lyx-devel mailing list
lyx-devel@lists.lyx.org
https://lists.lyx.org/mailman/listinfo/lyx-devel
