Le mercredi 28 février 2018 à 21:11 -0500, Orcan Ogetbil a écrit :
> On 28 February 2018 at 10:03, Nicolas Mailhot wrote:
> > Le 2018-02-28 15:28, Orcan Ogetbil a écrit :
> > >
> > > On 28 February 2018 at 09:19, Nicolas Mailhot wrote:
> >
> >
> > > These are all _very_ edge use-cases.
> >
> >
On 28 February 2018 at 10:03, Nicolas Mailhot wrote:
> Le 2018-02-28 15:28, Orcan Ogetbil a écrit :
>>
>> On 28 February 2018 at 09:19, Nicolas Mailhot wrote:
>
>
>> These are all _very_ edge use-cases.
>
>
> Those are *not* edge-cases.
We have a few thousand build failures. If you cannot find as
Le 2018-02-28 16:03, Nicolas Mailhot a écrit :
Most system libraries are written in C/C++ so pretty much all the
language toolchains we ship (except for toy languages not intended to
produce complex apps) will read C/C++ header files from their
compilers to use those system libraries, and will B
Le 2018-02-28 15:28, Orcan Ogetbil a écrit :
On 28 February 2018 at 09:19, Nicolas Mailhot wrote:
These are all _very_ edge use-cases.
Those are *not* edge-cases.
Most system libraries are written in C/C++ so pretty much all the
language toolchains we ship (except for toy languages not int
On 02/28/2018 04:28 PM, Orcan Ogetbil wrote:
On 28 February 2018 at 09:19, Nicolas Mailhot wrote:
My point was *non-C-compilers* can read *C/C++* header files because they
need to read the ABI definitions to use it from their non C/C++ code.
That makes a C/C++ header file consumable by pretty m
On 02/28/2018 11:28 AM, Rafal Luzynski wrote:
28.02.2018 09:33 Nicolas Mailhot wrote:
Le mercredi 28 février 2018 à 00:11 -0500, Orcan Ogetbil a écrit :
Shouldn't we consider having -devel packages Require gcc or gcc-c++?
What good is a header package without a compiler anyway?
This would al
On 28 February 2018 at 09:19, Nicolas Mailhot wrote:
> My point was *non-C-compilers* can read *C/C++* header files because they
> need to read the ABI definitions to use it from their non C/C++ code.
>
> That makes a C/C++ header file consumable by pretty much any kind of
> compiler, so it's *comp
My point was *non-C-compilers* can read *C/C++* header files because
they need to read the ABI definitions to use it from their non C/C++
code.
That makes a C/C++ header file consumable by pretty much any kind of
compiler, so it's *completely useless* to try to make C/C++ devel
packages pull
On 28 February 2018 at 05:28, Rafal Luzynski wrote:
> 28.02.2018 09:33 Nicolas Mailhot wrote:
>>
>> Le mercredi 28 février 2018 à 00:11 -0500, Orcan Ogetbil a écrit :
>> >
>> > Shouldn't we consider having -devel packages Require gcc or gcc-c++?
>> > What good is a header package without a compile
28.02.2018 09:33 Nicolas Mailhot wrote:
>
> Le mercredi 28 février 2018 à 00:11 -0500, Orcan Ogetbil a écrit :
> >
> > Shouldn't we consider having -devel packages Require gcc or gcc-c++?
> > What good is a header package without a compiler anyway?
> > This would also (indirectly) pull in the comp
Le mercredi 28 février 2018 à 00:11 -0500, Orcan Ogetbil a écrit :
>
> Shouldn't we consider having -devel packages Require gcc or gcc-c++?
> What good is a header package without a compiler anyway?
> This would also (indirectly) pull in the compiler and fix most of
> these failed builds.
>
gcc
Le 2018-02-21 15:28, Stephen John Smoogen a écrit :
I guess the problem I am having is I have no idea what we are
"fixing". It all seems like needless form filling for no benefit. If
XYZ-a is always going to pull in ABC.. why are we are explicitely
saying we need ABC. Is the end goal to have eve
12 matches
Mail list logo
