[Development] Accelerating deprecation of MSVC 2017
alexander.blasche at qt.io
Wed Aug 21 14:31:55 CEST 2019
> -----Original Message-----
> From: Development <development-bounces at qt-project.org> On Behalf Of
> Edward Welbourne
> Sent: Wednesday, 21 August 2019 11:15
> To: Konstantin Tokarev <annulen at yandex.ru>
> Cc: Thiago Macieira <thiago.macieira at intel.com>; development at qt-
> Subject: Re: [Development] Accelerating deprecation of MSVC 2017
> 21.08.2019, 00:45, "Thiago Macieira" <thiago.macieira at intel.com>:
> >>> I'd like to propose we announce now that it will not be supported
> >>> for Qt 5.15.
> Certainly if we can persuade ourselves now that this will be prudent then, we
> should announce our intent as soon as we make up our minds.
> >>> Even if 2019 still has code generation problems.
> ... today, when we're making the decision. If MS somehow manage not to fix
> them before we release 5.15, we can always change our minds.
I think this is a bit too simplistic. If we decide now, reality tends to have a way of ensuring that we cannot change our minds. Developers tend to quickly jump at opportunities offered by the fact that compilers are dropped and utilize further features of the newer compiler. Therefore, turning back could be more significant work.
What is the hurry for this decision? Generally, we look at download figures as base line and decide based on them. Thiago, are we currently facing issues where we cannot further fix/work around problems or is this request closer to the nice-to-have camp? We'd need a reason why we potentially want to annoy half the people downloading Qt on Windows. Note that I don't know what the current download figures and just want to highlight the Qt user's perspective.
More information about the Development