Date: Mon, 26 Jun 2023 16:40:01 -0300
They are using shared libraries, but they do *not* know about symbol
visibility. They write C++ and things just work. Symbol visibility is
useful knowledge for people dealing with the physical aspect of the
software (like me). And I know for sure there is a large part of developers
I personally deal with that don't know and never needed this information.
> therefore knowing about symbol visibility is knowledge that they should
possess.
When I wrote my first C++ code using boost I definitely had no idea about
symbol visibility. I added -lboost like the internet told me to, and my
software worked just fine. I don't understand why you draw that conclusion.
I'm not sure I agree there is a reason to standardise the definition of
libraries, because I don't think it will help in making more portable code.
The devil is in the details, and the details will be left as
"implementation defined" which will force the developer to code for a
specific implementation anyway.
On Mon, Jun 26, 2023 at 4:23 PM Thiago Macieira via Std-Proposals <
std-proposals_at_[hidden]> wrote:
> On Monday, 26 June 2023 10:28:00 PDT Breno Guimarães via Std-Proposals
> wrote:
> > > EVERY C++ programmer must know about them
> >
> > I disagree with this statement. I know many developers working with C++
> for
> > many many years who are working on the infrastructure that has been
> created
> > and never had to deal with library/application/symbol visibility and so
> on.
>
> I stand by my statement. I'd say there's over 90% chance that they are
> using
> shared libraries or DLLs of some sort in that infrastructure and therefore
> knowing about symbol visibility is knowledge that they should possess.
>
> Either way, that doesn't change the conclusion: libraries exist and
> there's
> reason to standardise existing practice, without detriment of the < 0.01%
> of
> users who don't use libraries at all.
>
> --
> Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
> Software Architect - Intel DCAI Cloud Engineering
>
>
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
visibility. They write C++ and things just work. Symbol visibility is
useful knowledge for people dealing with the physical aspect of the
software (like me). And I know for sure there is a large part of developers
I personally deal with that don't know and never needed this information.
> therefore knowing about symbol visibility is knowledge that they should
possess.
When I wrote my first C++ code using boost I definitely had no idea about
symbol visibility. I added -lboost like the internet told me to, and my
software worked just fine. I don't understand why you draw that conclusion.
I'm not sure I agree there is a reason to standardise the definition of
libraries, because I don't think it will help in making more portable code.
The devil is in the details, and the details will be left as
"implementation defined" which will force the developer to code for a
specific implementation anyway.
On Mon, Jun 26, 2023 at 4:23 PM Thiago Macieira via Std-Proposals <
std-proposals_at_[hidden]> wrote:
> On Monday, 26 June 2023 10:28:00 PDT Breno Guimarães via Std-Proposals
> wrote:
> > > EVERY C++ programmer must know about them
> >
> > I disagree with this statement. I know many developers working with C++
> for
> > many many years who are working on the infrastructure that has been
> created
> > and never had to deal with library/application/symbol visibility and so
> on.
>
> I stand by my statement. I'd say there's over 90% chance that they are
> using
> shared libraries or DLLs of some sort in that infrastructure and therefore
> knowing about symbol visibility is knowledge that they should possess.
>
> Either way, that doesn't change the conclusion: libraries exist and
> there's
> reason to standardise existing practice, without detriment of the < 0.01%
> of
> users who don't use libraries at all.
>
> --
> Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
> Software Architect - Intel DCAI Cloud Engineering
>
>
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
Received on 2023-06-26 19:40:14