HI,

with the addition of the fortran-2.eclass, it is possible to remove the
USE=fortran from the default profiles. Any objections?

justin

On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
> HI,
>
> with the addition of the fortran-2.eclass, it is possible to remove the
> USE=fortran from the default profiles. Any objections?
>
> justin

Nope, I actually suggested this back in December, and no one bothered
to respond.

Sounds good to me,
Matt

On 21/06/11 16:18, Matt Turner wrote:
> On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
>> HI,
>>
>> with the addition of the fortran-2.eclass, it is possible to remove the
>> USE=fortran from the default profiles. Any objections?
>>
>> justin
>
> Nope, I actually suggested this back in December, and no one bothered
> to respond.
>
> Sounds good to me,
> Matt
>

Removing fortran and adding an dependency on virtual/fortran lets
portage select dev-lang/ifc as the fortran compiler of choice.

I reverted it back temporarily, so everything should be fine now again.

justin

On Wed, 22 Jun 2011 09:15:35 +0200
justin <jlec@gentoo.org> wrote:

> On 21/06/11 16:18, Matt Turner wrote:
> > On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
> >> HI,
> >>
> >> with the addition of the fortran-2.eclass, it is possible to
> >> remove the USE=fortran from the default profiles. Any objections?
> >>
> >> justin
> >
> > Nope, I actually suggested this back in December, and no one
> > bothered to respond.
> >
> > Sounds good to me,
> > Matt
> >
>
> Removing fortran and adding an dependency on virtual/fortran lets
> portage select dev-lang/ifc as the fortran compiler of choice.

That's because portage doesn't suggest USE changes for one-of deps, and
just falls back to first USE-clean option.

> I reverted it back temporarily, so everything should be fine now
> again.

As said on IRC, if you really intend to do a thing like that, you
should re-enable fortfran on the gcc package by default.

--
Best regards,
Michał Górny

On 22/06/11 09:55, Michał Górny wrote:
> On Wed, 22 Jun 2011 09:15:35 +0200
> justin <jlec@gentoo.org> wrote:
>
>> On 21/06/11 16:18, Matt Turner wrote:
>>> On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
>>>> HI,
>>>>
>>>> with the addition of the fortran-2.eclass, it is possible to
>>>> remove the USE=fortran from the default profiles. Any objections?
>>>>
>>>> justin
>>>
>>> Nope, I actually suggested this back in December, and no one
>>> bothered to respond.
>>>
>>> Sounds good to me,
>>> Matt
>>>
>>
>> Removing fortran and adding an dependency on virtual/fortran lets
>> portage select dev-lang/ifc as the fortran compiler of choice.
>
> That's because portage doesn't suggest USE changes for one-of deps, and
> just falls back to first USE-clean option.
>

In my tests it does. if there only gcc[-fortra] present, it asks me to
change the USE.

>> I reverted it back temporarily, so everything should be fine now
>> again.
>
> As said on IRC, if you really intend to do a thing like that, you
> should re-enable fortfran on the gcc package by default.
>

I reverted to where we came from. Every further proceeding will be
coordinated with the toolchain guys. Generally I would like to remove
USE=fortran from default, as nearly solely science packages are written
in fortran and I doubt that nearly half of the user base do science.

I will come up with a plan.

jusitn

On Wed, Jun 22, 2011 at 3:01 AM, justin <jlec@gentoo.org> wrote:
> On 22/06/11 09:55, Michał Górny wrote:
>> On Wed, 22 Jun 2011 09:15:35 +0200
>> justin <jlec@gentoo.org> wrote:
>>
>>> On 21/06/11 16:18, Matt Turner wrote:
>>>> On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
>>>>> HI,
>>>>>
>>>>> with the addition of the fortran-2.eclass, it is possible to
>>>>> remove the USE=fortran from the default profiles. Any objections?
>>>>>
>>>>> justin
>>>>
>>>> Nope, I actually suggested this back in December, and no one
>>>> bothered to respond.
>>>>
>>>> Sounds good to me,
>>>> Matt
>>>>
>>>
>>> Removing fortran and adding an dependency on virtual/fortran lets
>>> portage select dev-lang/ifc as the fortran compiler of choice.
>>
>> That's because portage doesn't suggest USE changes for one-of deps, and
>> just falls back to first USE-clean option.
>>
>
> In my tests it does. if there only gcc[-fortra] present, it asks me to
> change the USE.
>
>>> I reverted it back temporarily, so everything should be fine now
>>> again.
>>
>> As said on IRC, if you really intend to do a thing like that, you
>> should re-enable fortfran on the gcc package by default.
>>
>
> I reverted to where we came from. Every further proceeding will be
> coordinated with the toolchain guys. Generally I would like to remove
> USE=fortran from default, as nearly solely science packages are written
> in fortran and I doubt that nearly half of the user base do science.
>
> I will come up with a plan.
>
> jusitn
>
>

Hey Justin,

One thing to note is that a few various python modules, like numpy,
really benefit from fortran. While many people using numpy are
scientists, many are not and further there are various modules that
depend on numpy that non-science folks use. I, for one, care little
about where that flag is set, since I have manually set that USE for
years now in make.conf. I am simply hoping to make you aware of the
fact that there are potential cases that could be easily overlooked.

Thanks!
Matt
--
Matthew W. Summers
Gentoo Foundation

On Wed, Jun 22, 2011 at 3:01 AM, justin <jlec@gentoo.org> wrote:
> On 22/06/11 09:55, Michał Górny wrote:
>> On Wed, 22 Jun 2011 09:15:35 +0200
>> justin <jlec@gentoo.org> wrote:
>>
>>> On 21/06/11 16:18, Matt Turner wrote:
>>>> On Tue, Jun 21, 2011 at 7:17 AM, justin <jlec@gentoo.org> wrote:
>>>>> HI,
>>>>>
>>>>> with the addition of the fortran-2.eclass, it is possible to
>>>>> remove the USE=fortran from the default profiles. Any objections?
>>>>>
>>>>> justin
>>>>
>>>> Nope, I actually suggested this back in December, and no one
>>>> bothered to respond.
>>>>
>>>> Sounds good to me,
>>>> Matt
>>>>
>>>
>>> Removing fortran and adding an dependency on virtual/fortran lets
>>> portage select dev-lang/ifc as the fortran compiler of choice.
>>
>> That's because portage doesn't suggest USE changes for one-of deps, and
>> just falls back to first USE-clean option.
>>
>
> In my tests it does. if there only gcc[-fortra] present, it asks me to
> change the USE.
>
>>> I reverted it back temporarily, so everything should be fine now
>>> again.
>>
>> As said on IRC, if you really intend to do a thing like that, you
>> should re-enable fortfran on the gcc package by default.
>>
>
> I reverted to where we came from. Every further proceeding will be
> coordinated with the toolchain guys. Generally I would like to remove
> USE=fortran from default, as nearly solely science packages are written
> in fortran and I doubt that nearly half of the user base do science.
>
> I will come up with a plan.
>
> jusitn
>
>

Hey Justin,

One thing to note is that a few various python modules, like numpy,
really benefit from fortran. While many people using numpy are
scientists, many are not and further there are various modules that
depend on numpy that non-science folks use. I, for one, care little
about where that flag is set, since I have manually set that USE for
years now in make.conf. I am simply hoping to make you aware of the
fact that there are potential cases that could be easily overlooked.

Thanks!
Matt
--
Matthew W. Summers
Gentoo Foundation

Am Mittwoch 22 Juni 2011, 16:35:07 schrieb Matthew Summers:
>
> Hey Justin,
>
> One thing to note is that a few various python modules, like numpy,
> really benefit from fortran. While many people using numpy are
> scientists, many are not and further there are various modules that
> depend on numpy that non-science folks use. I, for one, care little
> about where that flag is set, since I have manually set that USE for
> years now in make.conf. I am simply hoping to make you aware of the
> fact that there are potential cases that could be easily overlooked.
>
> Thanks!
> Matt
>

... and there are also typical end user software packages like digikam and kipi-plugins for photo processing, which rely e.g on opencv for face recognition, red eye removal, ...

--
Andreas K. Huettel
Gentoo Linux developer - kde, sci, arm, tex
dilfridge@gentoo.org
http://www.akhuettel.de/

Am Mittwoch 22 Juni 2011, 16:35:07 schrieb Matthew Summers:
>
> Hey Justin,
>
> One thing to note is that a few various python modules, like numpy,
> really benefit from fortran. While many people using numpy are
> scientists, many are not and further there are various modules that
> depend on numpy that non-science folks use. I, for one, care little
> about where that flag is set, since I have manually set that USE for
> years now in make.conf. I am simply hoping to make you aware of the
> fact that there are potential cases that could be easily overlooked.
>
> Thanks!
> Matt
>

... and there are also typical end user software packages like digikam and kipi-plugins for photo processing, which rely e.g on opencv for face recognition, red eye removal, ...

--
Andreas K. Huettel
Gentoo Linux developer - kde, sci, arm, tex
dilfridge@gentoo.org
http://www.akhuettel.de/

On 6/22/11 4:47 PM, Andreas K. Huettel wrote:
> Am Mittwoch 22 Juni 2011, 16:35:07 schrieb Matthew Summers:
>>
>> Hey Justin,
>>
>> One thing to note is that a few various python modules, like numpy,
>> really benefit from fortran. While many people using numpy are
>> scientists, many are not and further there are various modules that
>> depend on numpy that non-science folks use. I, for one, care little
>> about where that flag is set, since I have manually set that USE for
>> years now in make.conf. I am simply hoping to make you aware of the
>> fact that there are potential cases that could be easily overlooked.
>>
>> Thanks!
>> Matt
>>
>
> ... and there are also typical end user software packages like digikam and kipi-plugins for photo processing, which rely e.g on opencv for face recognition, red eye removal, ...
>

So then,

lets make a clear decision how we should handle the fortran USE.

1. leave it as it is as global USE enabled.

2. enable it as default for gcc

3. let package maintainers fix their packages,
means depend on fortran by default where it is the best and/or needed

thanks justin