Russel Winder <russel russel.org.uk> writes:

The game is known to be fundamentally flawed, but despite this, the
results have effect on the perception of programming languages.

http://shootout.alioth.debian.org/

D should compete with C, C++ and beat Go.  If anyone has any time
perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
repository with the various codes in it?

--=20
Russel.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D
Dr Russel Winder      t: +44 20 7585 2200   voip: sip:russel.winder ekiga.n=
et
41 Buckmaster Road    m: +44 7770 465 077   xmpp: russel russel.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

bearophile <bearophileHUGS lycos.com> writes:

Russel Winder:

 The game is known to be fundamentally flawed,

It has flaws, but if you try to create something better you will probably
create something far worse.


 If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

This is a starting point:
http://shootout.alioth.debian.org/debian/benchmark.php?test=all&lang=gdc&lang2=dlang
I also have some improved implementations.

Bye,
bearophile

dsimcha <dsimcha yahoo.com> writes:

== Quote from bearophile (bearophileHUGS lycos.com)'s article
 Russel Winder:
 The game is known to be fundamentally flawed,

 It has flaws, but if you try to create something better you will probably
create

something far worse.

Agreed.  People need to appreciate the difference between "noisy" and
"completely
worthless".  Maybe it's easier for me than for most people because I work with
statistics and noisy data on a daily basis.

The Language Benchmark Game is noisy.  This means that you can't tell at a
fine-grained resolution which language is fastest, and small differences are
almost meaningless.  It is not completely worthless, though.  If it shows a
difference of several fold, you can be pretty sure that's real.  I see no
problem
with it as long as everyone understands that it's noisy and that small
differences
are to be taken with a grain of salt.

Walter Bright <newshound2 digitalmars.com> writes:

Russel Winder wrote:
 The game is known to be fundamentally flawed, but despite this, the
 results have effect on the perception of programming languages.
 
 http://shootout.alioth.debian.org/
 
 D should compete with C, C++ and beat Go.  If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

The thing about that shootout is D used to be on there, until the maintainer of 
the site removed it. He refuses to include D on the benchmarks.

Eric Poggel <dnewsgroup2 yage3d.net> writes:

On 10/31/2010 5:13 PM, Walter Bright wrote:
 Russel Winder wrote:
 The game is known to be fundamentally flawed, but despite this, the
 results have effect on the perception of programming languages.

 http://shootout.alioth.debian.org/

 D should compete with C, C++ and beat Go. If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

 The thing about that shootout is D used to be on there, until the
 maintainer of the site removed it. He refuses to include D on the
 benchmarks.

Is it excluded for lack for 64-bit support.  What can we do to get it 
back on there?

bearophile <bearophileHUGS lycos.com> writes:

Eric Poggel:

 Is it excluded for lack for 64-bit support.

Before being sure of this you have to receive an answer from that person.
D was removed maybe for the lack of a 64 bit implementation, but also because
that site author has reduced the amount of work needed to manage the site,
reducing the number of benchmarks and languages. Many D programs were very
C-looking (because they were designed to be fast if compiled with DMD), and
that site owner has always desired to show a high diversity in the language
solutions, and remove the very similar languages where possible.

Bye,
bearophile

dsimcha <dsimcha yahoo.com> writes:

== Quote from Walter Bright (newshound2 digitalmars.com)'s article
 Russel Winder wrote:
 The game is known to be fundamentally flawed, but despite this, the
 results have effect on the perception of programming languages.

 http://shootout.alioth.debian.org/

 D should compete with C, C++ and beat Go.  If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

 The thing about that shootout is D used to be on there, until the maintainer of
 the site removed it. He refuses to include D on the benchmarks.

Has he given a reason?  IIRC it's because D isn't/wasn't included in the package
repos of the distributions he uses.  This will probably get fixed soon:

Fedora already includes LDC.

GDC is now up to date (or at most one version behind) for D1 and is rapidly
catching up for D2.

Unfortunately, the Linux world is dominated by FOSS zealots for whom DMD isn't
open enough.  In the ideal world the DMD backend would be under an OSI-approved
license so that it could be distributed by even the most zealously pro-FOSS
distributions.

retard <re tard.com.invalid> writes:

Sun, 31 Oct 2010 21:45:56 +0000, dsimcha wrote:

 Unfortunately, the Linux world is dominated by FOSS zealots for whom DMD
 isn't open enough.  In the ideal world the DMD backend would be under an
 OSI-approved license so that it could be distributed by even the most
 zealously pro-FOSS distributions.

The Linux world isn't dominated by a single homogenous group called the 
'FOSS zealots'. There are as many ideological views as you can imagine. 
First, there's an ideological mismatch between 'free' and 'open' (it's 
the ESR vs RMS issue). ESR believes in right wing politics, RMS is a left 
wing fundamentalist. Some distros include non-free packages if they have 
the distribution rights. Some distros even provide commercial packages 
(http://www.fluendo.com/shop/product/fluendo-mp3-decoder/, http://
www.transgaming.com/ etc.)

For example the language shootout author uses Debian derived 
distributions. They don't include non-OSI aproved software in the 
community repositories. It's simple as that. http://www.opensource.org/
docs/osd - provide an OSI aproved license + 32 & 64-bit backends for x86.

Lutger <lutger.blijdestijn gmail.com> writes:

dsimcha wrote:

 == Quote from Walter Bright (newshound2 digitalmars.com)'s article
 Russel Winder wrote:
 The game is known to be fundamentally flawed, but despite this, the
 results have effect on the perception of programming languages.

 http://shootout.alioth.debian.org/

 D should compete with C, C++ and beat Go.  If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

 The thing about that shootout is D used to be on there, until the
 maintainer of the site removed it. He refuses to include D on the
 benchmarks.

 
 Has he given a reason?  IIRC it's because D isn't/wasn't included in the
 package
 repos of the distributions he uses.  This will probably get fixed soon:
 
 Fedora already includes LDC.
 
 GDC is now up to date (or at most one version behind) for D1 and is
 rapidly catching up for D2.
 
 Unfortunately, the Linux world is dominated by FOSS zealots for whom DMD
 isn't
 open enough.  In the ideal world the DMD backend would be under an
 OSI-approved license so that it could be distributed by even the most
 zealously pro-FOSS distributions.

Of which there are very few. A linux distro or community repository cannot 
distribute dmd at all, it is prohibited by the license. This is primarily a 
practical issue, not an ideological one.

retard <re tard.com.invalid> writes:

Sun, 31 Oct 2010 23:10:02 +0100, Lutger wrote:

 dsimcha wrote:
 
 == Quote from Walter Bright (newshound2 digitalmars.com)'s article
 Russel Winder wrote:
 The game is known to be fundamentally flawed, but despite this, the
 results have effect on the perception of programming languages.

 http://shootout.alioth.debian.org/

 D should compete with C, C++ and beat Go.  If anyone has any time
 perhaps a "cabal" should form to create a Bazaar/Mercurial/Git
 repository with the various codes in it?

 The thing about that shootout is D used to be on there, until the
 maintainer of the site removed it. He refuses to include D on the
 benchmarks.

 
 Has he given a reason?  IIRC it's because D isn't/wasn't included in
 the package
 repos of the distributions he uses.  This will probably get fixed soon:
 
 Fedora already includes LDC.
 
 GDC is now up to date (or at most one version behind) for D1 and is
 rapidly catching up for D2.
 
 Unfortunately, the Linux world is dominated by FOSS zealots for whom
 DMD isn't
 open enough.  In the ideal world the DMD backend would be under an
 OSI-approved license so that it could be distributed by even the most
 zealously pro-FOSS distributions.

 
 Of which there are very few. A linux distro or community repository
 cannot distribute dmd at all, it is prohibited by the license. This is
 primarily a practical issue, not an ideological one.

Right.. I forgot the issue is also practical. If DigitalMars doesn't 
allow redistribution, they simply cannot include it even if they wanted.

E.g. Canonical provides non-free software via some kind of partner 
programme. It's possible to easily install non-free applications ( http://
archive.canonical.com/ubuntu/pool/partner/ ). My guess is, they don't 
come and kneel before the Great DigitalMars. You need to actively offer 
your software to them.

"Diego Cano Lagneaux" <d.cano.lagneaux gmail.com> writes:

 Of which there are very few. A linux distro or community repository
 cannot distribute dmd at all, it is prohibited by the license. This is
 primarily a practical issue, not an ideological one.

 Right.. I forgot the issue is also practical. If DigitalMars doesn't
 allow redistribution, they simply cannot include it even if they wanted.

ArchLinux includes dmd (version 1) in its community repo. It does not seem  
so impossible as you say.

Gour <gour atmarama.net> writes:

On Mon, 01 Nov 2010 11:19:08 +0100
 "Diego" =3D=3D "Diego Cano Lagneaux" wrote:






Diego> ArchLinux includes dmd (version 1) in its community repo. It
Diego> does not seem so impossible as you say.

Don't forget about dmd2 in AUR. ;)

(http://aur.archlinux.org/packages.php?O=3D0&K=3Ddmd2&do_Search=3DGo)



--=20

Gour  | Hlapicina, Croatia  | GPG key: CDBF17CA
----------------------------------------------------------------

retard <re tard.com.invalid> writes:

Mon, 01 Nov 2010 11:19:08 +0100, Diego Cano Lagneaux wrote:


 Of which there are very few. A linux distro or community repository
 cannot distribute dmd at all, it is prohibited by the license. This is
 primarily a practical issue, not an ideological one.

 Right.. I forgot the issue is also practical. If DigitalMars doesn't
 allow redistribution, they simply cannot include it even if they
 wanted.

 ArchLinux includes dmd (version 1) in its community repo. It does not
 seem so impossible as you say.

I didn't say it seems so impossible. The sentence starts with the word 
IF. That's exactly what the prohibition of redistribution means. It's a 
fact, it's not an opinion. You don't need belief.

The ArchLinux developers have either made a deal with DigitalMars or they 
only provide a download script that fetches the binaries from the 
DigitalMars site. However this is not something they would allow in the 
standard core packages. Otherwise they couldn't provide any installers or 
live CD/DVD distributions.

=?UTF-8?B?IkrDqXLDtG1lIE0uIEJlcmdlciI=?= <jeberger free.fr> writes:

Diego Cano Lagneaux wrote:
 Of which there are very few. A linux distro or community repository
 cannot distribute dmd at all, it is prohibited by the license. This i=



s
 primarily a practical issue, not an ideological one.

 Right.. I forgot the issue is also practical. If DigitalMars doesn't
 allow redistribution, they simply cannot include it even if they wante=


d.

 ArchLinux includes dmd (version 1) in its community repo. It does not
 seem so impossible as you say.

	Are you sure? Here, "yaourt -Ss dmd" only show results from AUR
(which means that they download from DigitalMars anyway)

		Jerome
--=20
mailto:jeberger free.fr
http://jeberger.free.fr
Jabber: jeberger jabber.fr

"Diego Cano Lagneaux" <d.cano.lagneaux gmail.com> writes:

En Wed, 03 Nov 2010 21:36:10 +0100, J=C3=A9r=C3=B4me M. Berger <jeberger=
 free.fr>  =

escribi=C3=B3:

 Diego Cano Lagneaux wrote:
 Of which there are very few. A linux distro or community repository=




 cannot distribute dmd at all, it is prohibited by the license. This=




 is
 primarily a practical issue, not an ideological one.

 Right.. I forgot the issue is also practical. If DigitalMars doesn't=



 allow redistribution, they simply cannot include it even if they  =



 wanted.

 ArchLinux includes dmd (version 1) in its community repo. It does not=


 seem so impossible as you say.

 	Are you sure? Here, "yaourt -Ss dmd" only show results from AUR
 (which means that they download from DigitalMars anyway)

 		Jerome

Using pacman to avoid AUR results:

~> pacman -Ss dmd
community/dmd 1.058-1
     The Digital Mars D compiler.

Using packer, to see the AUR:

~> packer -Ss dmd
community/dmd 1.058-1
     The Digital Mars D compiler.
aur/gdc 0.24-1
     GDC, Digital Mars D Programing Language (DMD) frontend for GCC
aur/gdc-svn 243-3
     GDC, Digital Mars D Programing Language (DMD) frontend for GCC
aur/gdc-bin 0.24-3
     GDC, Digital Mars D Programing Language (DMD) frontend for GCC
aur/dmd2 2.050-1
     The Digital Mars D compiler (D2 version)
aur/dmd2-complete 2.050-2
     The Digital Mars D compiler & Standard Library (D2 version)
aur/dmd2-bin 2.050-1
     The Digital Mars D compiler (D2 version)
aur/dmd2-complete-bin 2.049-1
     The Digital Mars D compiler & Standard Library (D2 version)
aur/gdc1-hg 304-1
     GDC, Digital Mars D Programing Language (DMD) frontend for GCC (D1 =
 =

version)
aur/gdc2-hg 304-1
     GDC, Digital Mars D Programing Language (DMD) frontend for GCC (D2 =
 =

version)

Iain Buclaw <ibuclaw ubuntu.com> writes:

== Quote from Diego Cano Lagneaux (d.cano.lagneaux gmail.com)'s article
 En Wed, 03 Nov 2010 21:36:10 +0100, J=C3=A9r=C3=B4me M. Berger <jeberger=
  free.fr>  =
 escribi=C3=B3:
 Diego Cano Lagneaux wrote:
 Of which there are very few. A linux distro or community repository=
 cannot distribute dmd at all, it is prohibited by the license. This=




  is
 primarily a practical issue, not an ideological one.

 Right.. I forgot the issue is also practical. If DigitalMars doesn't=
 allow redistribution, they simply cannot include it even if they  =
 wanted.

 ArchLinux includes dmd (version 1) in its community repo. It does not=
 seem so impossible as you say.

 	Are you sure? Here, "yaourt -Ss dmd" only show results from AUR
 (which means that they download from DigitalMars anyway)

 		Jerome

 Using pacman to avoid AUR results:
 ~> pacman -Ss dmd
 community/dmd 1.058-1
      The Digital Mars D compiler.

Flashplayer is distributed via installing a "download, unpack and install"
script
on systems. There's nothing stopping vendors from using the same technique for
DMD.

Iain

Walter Bright <newshound2 digitalmars.com> writes:

Lutger wrote:
 A linux distro or community repository cannot 
 distribute dmd at all, it is prohibited by the license. This is primarily a 
 practical issue, not an ideological one.

And any one who has bothered to ask me, I have given permission to include it.

Eric Poggel <dnewsgroup2 yage3d.net> writes:

On 10/31/2010 7:25 PM, Walter Bright wrote:
 Lutger wrote:
 A linux distro or community repository cannot distribute dmd at all,
 it is prohibited by the license. This is primarily a practical issue,
 not an ideological one.

 And any one who has bothered to ask me, I have given permission to
 include it.

I think a lot of people have the "If you have to ask, the answer is no" 
mentality.  Maybe if distribution rights were codified in the license 
then there would be less apprehension about it.

It's your code, and you have the right do do with it as you please, but 
would you be willing add distribution rights to the license?

Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:

On 11/1/10 11:47 AM, Eric Poggel wrote:
 On 10/31/2010 7:25 PM, Walter Bright wrote:
 Lutger wrote:
 A linux distro or community repository cannot distribute dmd at all,
 it is prohibited by the license. This is primarily a practical issue,
 not an ideological one.

 And any one who has bothered to ask me, I have given permission to
 include it.

 I think a lot of people have the "If you have to ask, the answer is no"
 mentality. Maybe if distribution rights were codified in the license
 then there would be less apprehension about it.

 It's your code, and you have the right do do with it as you please, but
 would you be willing add distribution rights to the license?

I agree. Unmentioned availability of per-case customized licensing terms 
is very rare in the industry.

Andrei

Lutger <lutger.blijdestijn gmail.com> writes:

Walter Bright wrote:

 Lutger wrote:
 A linux distro or community repository cannot
 distribute dmd at all, it is prohibited by the license. This is
 primarily a practical issue, not an ideological one.

 
 And any one who has bothered to ask me, I have given permission to
 include it.

That's great. I guess the community (me included) could take more advantage 
of that. For most distro's it is not an option though, and even if they do 
it, it's for something their users really want. I'm not taking a position 
here, just saying that this is an issue in adoption.