Francois LE COAT wrote:
> Hi Tim,
> 
> Tim Cook wrote :
>> Be glad they're releasing anything at all, for starters.  If they were 
>> being paid, they might be obligated to do certain things.  But they're 
>> not.  Got a problem with it?  Write your own raytracer.
> 
> I'm not speaking about my personal interest, but about general policy.
> I'm supporting POV for more than 10 years now. Specially on ATARI
> platforms. Do you know that Windows is *the* latest non Unix OS ?
> Why should it be promoted rather than any other OS ?
> 
> Regards,
> 

> Author of Eureka 2.12 (2D Graph Describer, 3D Modeller)
> http://eureka.atari.org

It's not being promoted.  The POV-Team is focusing on fixing bugs 
related to the new feature (multithreading).  To minimize distractions 
from this, they have decided to focus on a single platform for the 
development.

They could do source releases and focus on implementing the changes only 
in the generic code versions; however, due to past experiences the 
POV-Team has decided that time-limited binaries better suit their needs 
(and the community's).

Since they must do a binary release, on a single platform, they simply 
chose the platform with the greatest number of users.

...Chambers

Post a reply to this message

"Francois LE COAT" <lec### [at] atariorg> wrote in message
news:43e62a8f$1@news.povray.org...
> Hi,
>

> > POV-Ray 3.7.beta.11c is available from http://www.povray.org/beta/.
> >
> > This beta is currently for the Windows platform only and includes standard,
> > SSE2, and 64-bit binaries.
>
> That's a shame that POVRay beta is only available for Windows. My policy
> as a developer is to support Linux and MacOSX platforms ... I do regret
> that the POVTeam has the exact opposed policy :( Who benefit from it ?

POV-Ray is well-known for its multiple platform support.  The overhead involved
in producing a compile for each platform that POV-Ray supports for each beta
release would be extreme, and is unnecessary for the testing process.

By restricting testing to just one platform it benefits the POV Team, as less
time is wasted producing various compiles every time a new beta is released,
which allows the development to progress more quickly, which ultimately benefits
all POV-Ray users.  So in answer to your question regarding who benefits:  all
POV-Ray users benefit in the end.

Lance.

thezone - thezone.firewave.com.au

Post a reply to this message

Lance Birch <-> wrote:
> POV-Ray is well-known for its multiple platform support.  The overhead involved
> in producing a compile for each platform that POV-Ray supports for each beta
> release would be extreme, and is unnecessary for the testing process.

  It's not just a question of compilation: Frontends need to be created
for each platform. Not a trivial task.

-- 
                                                          - Warp

Post a reply to this message

"Warp" <war### [at] tagpovrayorg> wrote in message news:43e8be8f@news.povray.org...
> Lance Birch <-> wrote:
> > POV-Ray is well-known for its multiple platform support.  The overhead
involved
> > in producing a compile for each platform that POV-Ray supports for each beta
> > release would be extreme, and is unnecessary for the testing process.
>
>   It's not just a question of compilation: Frontends need to be created
> for each platform. Not a trivial task.

Definitely.  I'm simplifying things greatly - there is so much to do to create
each platform's release.  I was just noting the time overhead involved for each
revision of the beta if you were to create one for each platform... compiling
for each platform (as it is, there are three compiles; 32-bit/32-bit
SSE2/64-bit), compressing the various platform compiles with their documentation
into separate archives, updating the website with the multiple platform compiles
and no doubt writing notes for each, and so on.  It would probably take 5 times
as long just to organise the compiles for the revisions and get them on the
website, let alone the actual development involved.  One release is complex
enough, and for the purposes of testing and debugging the new internals one
platform is all that is required (and choosing the platform with the largest
beta tester user base is common sense).

People seem to have forgotten this is a beta.

Lance.

thezone - thezone.firewave.com.au

Post a reply to this message

Hi,

Lance Birch wrote :
> "Warp" wrote :
>> Lance Birch wrote:
>>> POV-Ray is well-known for its multiple platform support.  The overhea
d involved
>>> in producing a compile for each platform that POV-Ray supports for ea
ch beta
>>> release would be extreme, and is unnecessary for the testing process.

>>   It's not just a question of compilation: Frontends need to be create
d
>> for each platform. Not a trivial task.
> 
> Definitely.  I'm simplifying things greatly - there is so much to do to
 create
> each platform's release.  I was just noting the time overhead involved 
for each
> revision of the beta if you were to create one for each platform... com
piling
> for each platform (as it is, there are three compiles; 32-bit/32-bit
> SSE2/64-bit), compressing the various platform compiles with their docu
mentation
> into separate archives, updating the website with the multiple platform
 compiles
> and no doubt writing notes for each, and so on.  It would probably take
 5 times
> as long just to organise the compiles for the revisions and get them on
 the
> website, let alone the actual development involved.  One release is com
plex
> enough, and for the purposes of testing and debugging the new internals
 one
> platform is all that is required (and choosing the platform with the la
rgest
> beta tester user base is common sense).
> 
> People seem to have forgotten this is a beta.

I understand that a platform must be chosen to propose a beta version.
Otherwise it may be rather difficult to manage releases ... But why
is Windows chosen, knowing that it is the only OS that is non Unix ?

I can understand that there's much more beta testers under Windows,
because this system has a commercial dominant position. But this system
is also non compatible with every others !

If another system had been chosen, it would have been the best way to
port to others, and to have a representative implementation. You know
that POSIX threads are common to really a great amount of systems, if
it was the point ... Computing farms are running Unix systems ...

Windows is the worst choice to experiment a beta version. I always
personally ported POV starting from Linux sources anyway.

That's why I'm speaking of the way to "promote" Windows, which
sounds me to be the worst choice.

Regards,


Author of Eureka 2.12 (2D Graph Describer, 3D Modeller)
http://eureka.atari.org

Post a reply to this message

Francois LE COAT wrote:
> Hi,
> 
> Lance Birch wrote :
>> People seem to have forgotten this is a beta.
> 
> I understand that a platform must be chosen to propose a beta version.
> Otherwise it may be rather difficult to manage releases ... But why
> is Windows chosen, knowing that it is the only OS that is non Unix ?

Why do you worship Unix?

> I can understand that there's much more beta testers under Windows,

You just answered your above question.

> because this system has a commercial dominant position. But this system
> is also non compatible with every others !

Excuse me, but can you take a binary compiled for Mac OS X and run it 
under Linux?  Can you even run it under Yellow Dog Linux (compiled 
specifically for the G5)?  Can you run a Linux x86 binary and run it 
under Sparc Solaris?

All of these systems are incompatible with all the others.

> If another system had been chosen, it would have been the best way to
> port to others, and to have a representative implementation.

Nobody is porting the betas to other systems, so this is a moot point. 
Once the code base has stabilized, then the release of various ports 
will begin.

Besides, all that is needed for a "representative implementation", would 
be a representative sample of all POV users... the majority of which run 
Windows...

> You know
> that POSIX threads are common to really a great amount of systems, if
> it was the point ...

Irrelevant...


> Computing farms are running Unix systems ...

... also irrelevant...


> Windows is the worst choice to experiment a beta version.

Why?  The most users are on Windows.

> I always
> personally ported POV starting from Linux sources anyway.

What you do or do not do has no effect on the POV-Team.

> That's why I'm speaking of the way to "promote" Windows, which
> sounds me to be the worst choice.

They aren't "promoting" windows.  If they were to say, "POV-Ray only 
runs on Windows because it's the best OS out there!" then that would be 
promoting Windows.  Instead, they're saying, "We are releasing a buggy 
and incomplete version of POV-Ray for testing purposes only on the 
system which most POV-Ray users operate on."  Hardly promotional 
material, that.

> 
> Regards,

Salutations,

...Chambers

Post a reply to this message

Hi,

You're against Unices ... But it is the present evolution of OSes !

Are you experienced ?

Regards,


Author of Eureka 2.12 (2D Graph Describer, 3D Modeller)
http://eureka.atari.org

Chambers wrote :
> Francois LE COAT wrote:
>> Lance Birch wrote :
>>> People seem to have forgotten this is a beta.
>>
>> I understand that a platform must be chosen to propose a beta version.

>> Otherwise it may be rather difficult to manage releases ... But why
>> is Windows chosen, knowing that it is the only OS that is non Unix ?
> 
> Why do you worship Unix?
> 
>> I can understand that there's much more beta testers under Windows,
> 
> You just answered your above question.
> 
>> because this system has a commercial dominant position. But this syste
m
>> is also non compatible with every others !
> 
> Excuse me, but can you take a binary compiled for Mac OS X and run it 
> under Linux?  Can you even run it under Yellow Dog Linux (compiled 
> specifically for the G5)?  Can you run a Linux x86 binary and run it 
> under Sparc Solaris?
> 
> All of these systems are incompatible with all the others.
> 
>> If another system had been chosen, it would have been the best way to
>> port to others, and to have a representative implementation.
> 
> Nobody is porting the betas to other systems, so this is a moot point. 

> Once the code base has stabilized, then the release of various ports 
> will begin.
> 
> Besides, all that is needed for a "representative implementation", woul
d 
> be a representative sample of all POV users... the majority of which ru
n 
> Windows...
> 
>> You know
>> that POSIX threads are common to really a great amount of systems, if
>> it was the point ...
> 
> Irrelevant...
> 
>> Computing farms are running Unix systems ...
> 
> ... also irrelevant...
> 
>> Windows is the worst choice to experiment a beta version.
> 
> Why?  The most users are on Windows.
> 
>> I always
>> personally ported POV starting from Linux sources anyway.
> 
> What you do or do not do has no effect on the POV-Team.
> 
>> That's why I'm speaking of the way to "promote" Windows, which
>> sounds me to be the worst choice.
> 
> They aren't "promoting" windows.  If they were to say, "POV-Ray only 
> runs on Windows because it's the best OS out there!" then that would be
 
> promoting Windows.  Instead, they're saying, "We are releasing a buggy 

> and incomplete version of POV-Ray for testing purposes only on the 
> system which most POV-Ray users operate on."  Hardly promotional 
> material, that.

Post a reply to this message

Francois LE COAT wrote:
> Hi,
> 
> You're against Unices ...

Show me one single quote where I am against Unices.

> But it is the present evolution of OSes !

Feel free to send me an email (at bdc### [at] yahoocom) explaining why 
you believe this.  However, this discussion is rapidly becoming 
inappropriate for this newsgroup.

Again, the POV-Team has released a buggy and incomplete version of 
POV-Ray, for testing purposes only, on the platform which most POV-Ray 
users run.  When the new code has stabilized and is ready for general 
use, then it will be released on a variety of platforms.

> Are you experienced ?

Only with my wife, I saved myself for her.

...Chambers

Post a reply to this message

> "Francois LE COAT" <lec### [at] atariorg> wrote in message
news:43e8ea5d$1@news.povray.org...
> Hi,
>
> Lance Birch wrote :
> > "Warp" wrote :
> >> Lance Birch wrote:
> >>> POV-Ray is well-known for its multiple platform support.  The overhead
involved
> >>> in producing a compile for each platform that POV-Ray supports for each
beta
> >>> release would be extreme, and is unnecessary for the testing process.
> >>   It's not just a question of compilation: Frontends need to be created
> >> for each platform. Not a trivial task.
> >
> > Definitely.  I'm simplifying things greatly - there is so much to do to
create
> > each platform's release.  I was just noting the time overhead involved for
each
> > revision of the beta if you were to create one for each platform...
compiling
> > for each platform (as it is, there are three compiles; 32-bit/32-bit
> > SSE2/64-bit), compressing the various platform compiles with their
documentation
> > into separate archives, updating the website with the multiple platform
compiles
> > and no doubt writing notes for each, and so on.  It would probably take 5
times
> > as long just to organise the compiles for the revisions and get them on the
> > website, let alone the actual development involved.  One release is complex
> > enough, and for the purposes of testing and debugging the new internals one
> > platform is all that is required (and choosing the platform with the largest
> > beta tester user base is common sense).
> >
> > People seem to have forgotten this is a beta.
>
> I understand that a platform must be chosen to propose a beta version.
> Otherwise it may be rather difficult to manage releases ... But why
> is Windows chosen, knowing that it is the only OS that is non Unix ?

I have already answered this, but as Chambers has pointed out you have also
answered it:

> I can understand that there's much more beta testers under Windows,

Exactly.  There are more beta testers on Windows, therefore it makes sense to
release the beta compile for Windows.  That is the end of the story.

> because this system has a commercial dominant position. But this system
> is also non compatible with every others !
>
> If another system had been chosen, it would have been the best way to
> port to others, and to have a representative implementation. You know

There is no source released for the beta, so it can't be ported to anything.

> that POSIX threads are common to really a great amount of systems, if
> it was the point ... Computing farms are running Unix systems ...

What does that have to do with this test release?

> Windows is the worst choice to experiment a beta version. I always

That is incorrect.  As you said, "there's much more beta testers under Windows".

> personally ported POV starting from Linux sources anyway.

There is no source available for the beta, so this is irrelevant.

> That's why I'm speaking of the way to "promote" Windows, which
> sounds me to be the worst choice.

You are mistaken.  There is no "promotion" of Windows.  The simple fact is that
it has the largest tester user base and therefore is the platform of choice for
the beta test compiles.  There is no point releasing a test version on a
platform where there are fewer testers.  To choose to release a test compile for
a platform that doesn't have as many testers as another platform makes
absolutely no sense.

Lance.

thezone - thezone.firewave.com.au

Post a reply to this message

Francois LE COAT wrote:
> Otherwise it may be rather difficult to manage releases ... But why
> is Windows chosen

The answer is simple: apart from the other reasons quoted in this thread,
the simple fact is that most of the work in making POV smp-compatible was
done by two persons - myself and Thorsten. I develop mostly under Windows,
and Thorsten on the Macintosh. So it was a choice between those two as to
which platform got the beta (and not even that really, since even today
the Mac frontend isn't completely finished).

The way the POV-Ray backend communicates with the frontend has undergone
significant changes and it was decided that it was best to work out the
problems on one platform rather than inflict the fairly fluid state of the
interface on our volunteer unix developers.

There is no 'conspiracy' against unix users, it is simply pragmatism,
nothing else. Once our unix developers are happy with the state of the
code I am sure they will release a unix build.

At this point I think this thread should end as it is not achieving any
positive purpose.

-- Chris

Post a reply to this message