Hello,
I wondered if someone (Walter) could clarify the licensing for the Phobos
library components of the D distribution.  It's clear how the generously donated
front-end to the compiler is licensed, and that the compiler components not
included are under a proprietary license, but I was not clear on Phobos - esp.
because it is being included in the gcc distribution.  I'm very interested in D
but it appears that the Phobos library is essential to any usable implementation
and I'm just not sure how it is being licensed.

Tbanks..

Craig_Welch2@yahoo.com wrote:

> Hello,
> I wondered if someone (Walter) could clarify the licensing for the Phobos
> library components of the D distribution.  It's clear how the generously
> donated front-end to the compiler is licensed, and that the compiler
> components not included are under a proprietary license, but I was not
> clear on Phobos - esp.
> because it is being included in the gcc distribution.  I'm very interested
> in D but it appears that the Phobos library is essential to any usable
> implementation and I'm just not sure how it is being licensed.
> 
> Tbanks..

Hi,

This is what Ant (developer of DUI) spoke about this problem:

-----------------------------------------------------------
"isn't phobos glt/artistic?

As a preventive measure I delete all Synesis source code from
my machine.
Luckally I never used any and my project (DUI and leds) still
compile with the new version of phobos.

My problem is that in the future some Synesis might see
some similarities on their code and mine and claim ownership
of something I wrote...
Ok here is the complete message I wrote last night:

*** original complete message ***

IANAL so I might be going overboard for nothing...

please enlighten me.

from phobos.loader:

 *              (ii) Any derived versions of this software (howsoever modified)
 *              remain the sole property of Synesis Software.


Can we live with this?
I can not.

I can't admit that in the future Synesis claims ownership of some thing I
wrote because it might have some similarities.
With the current IP and patent problems I can not afford to look at code
with that type of license at least until I discuss this with my lawyer.

As preventive mesuares:
I just deleted all source code from Synesis from my computer.
I deleted any files on the phobos src tree that contained the string
"ynesis" (except std/mmfile.d that doesn't contain the Synesis license)
and all my old (previous versions) dmd.zip files.
and I deleted the recent downloaded dmd.zip for dmd 0.86

Matthew, if there is another procedure to follow by someone that doesn't agree with the Synesis license let me know.

phobos still compiles (after changes to linux.mak) and DUI and leds still compile and run with the new phobos.

Walter, can you distribute phobos without the source code from Synesis?
Acording to some recente reading from growklaw If one has access to
some code and produces something similar it's assumed that it is a copy
even if the author claims never had look to the code
(for the US I guess).

"header" files should be OK unless the ridiculous claim SCO made on linux .h files is sanctioned by the courts.

Thanks,

Ant"

-------------------------------------------------------------

As you can see, he removed all bad licensed sources from phobos and it still compile his softwares (DUI and LEDS).

I think that if we want a free/open library, the best way is to remove the problematic sources or, preferably, the authors of that code change to a Free licence.

Bruno.

This is helpful, thank you, although for what it is worth I took a look at the Synesis license and it appears to me to be a pretty standard clause - derived has a particular meaning which is in many such licenses - consult an attorney if you are concerned and all that.  What I was concerned about were the areas under dmd/src/phobos/internal and dmd/src/phobos/std - there is a mix of things which are copyright digital mars all rights reserved, some things copyrighted by others and clearly licensed, and some with no notices whatsoever.  Under the dmd/src/dmd area things are clearly marked - but dmd/src/phobos is not within that folder and does not have clear indications regarding license in the way that the dmd/src/dmd area does.  The license at the higher dmd level is clearly restrictive.  I certainly respect Walter and other's right to distribute their work as they see fit - but I cannot see being able to make use of D at present without those areas of the tree, and technically I do not believe it is open sourced at this point, which is one of the requirements that I have.  What I am looking for is clarification on that, I would like to get into the language and make use of it, but do not see that as worthwhile unless the phobos lib is clearly open.  It may be that I'm being too picky - forgive me if so - but it also may be that this is the intent (to restrictively license portions of phobos) and I wish to be clear on that before investing time and effort into working with the language.

In article <c785qi$21b$1@digitaldaemon.com>, Bruno A. Costa says...
>
>Craig_Welch2@yahoo.com wrote:
>
>> Hello,
>> I wondered if someone (Walter) could clarify the licensing for the Phobos
>> library components of the D distribution.  It's clear how the generously
>> donated front-end to the compiler is licensed, and that the compiler
>> components not included are under a proprietary license, but I was not
>> clear on Phobos - esp.
>> because it is being included in the gcc distribution.  I'm very interested
>> in D but it appears that the Phobos library is essential to any usable
>> implementation and I'm just not sure how it is being licensed.
>> 
>> Tbanks..
>
>Hi,
>
>This is what Ant (developer of DUI) spoke about this problem:
>
>-----------------------------------------------------------
>"isn't phobos glt/artistic?
>
>As a preventive measure I delete all Synesis source code from
>my machine.
>Luckally I never used any and my project (DUI and leds) still
>compile with the new version of phobos.
>
>My problem is that in the future some Synesis might see
>some similarities on their code and mine and claim ownership
>of something I wrote...
>Ok here is the complete message I wrote last night:
>
>*** original complete message ***
>
>IANAL so I might be going overboard for nothing...
>
>please enlighten me.
>
>from phobos.loader:
>
> *              (ii) Any derived versions of this software (howsoever modified)
> *              remain the sole property of Synesis Software.
>
>
>Can we live with this?
>I can not.
>
>I can't admit that in the future Synesis claims ownership of some thing I
>wrote because it might have some similarities.
>With the current IP and patent problems I can not afford to look at code
>with that type of license at least until I discuss this with my lawyer.
>
>As preventive mesuares:
>I just deleted all source code from Synesis from my computer.
>I deleted any files on the phobos src tree that contained the string
>"ynesis" (except std/mmfile.d that doesn't contain the Synesis license)
>and all my old (previous versions) dmd.zip files.
>and I deleted the recent downloaded dmd.zip for dmd 0.86
>
>Matthew, if there is another procedure to follow by someone that doesn't agree with the Synesis license let me know.
>
>phobos still compiles (after changes to linux.mak) and DUI and leds still compile and run with the new phobos.
>
>Walter, can you distribute phobos without the source code from Synesis?
>Acording to some recente reading from growklaw If one has access to
>some code and produces something similar it's assumed that it is a copy
>even if the author claims never had look to the code
>(for the US I guess).
>
>"header" files should be OK unless the ridiculous claim SCO made on linux .h files is sanctioned by the courts.
>
>Thanks,
>
>Ant"
>
>-------------------------------------------------------------
>
>As you can see, he removed all bad licensed sources from phobos and it still compile his softwares (DUI and LEDS).
>
>I think that if we want a free/open library, the best way is to remove the problematic sources or, preferably, the authors of that code change to a Free licence.
>
>Bruno.

You should share your ideas in the digitalmars.D newsgroup. There are most of the users. And some (as me) think that the license is a real problem in the general acceptance of D. I am new in D, and it is a fantastic programming language, but without a free licence I don't think D becomes widely used so soon.

Bruno.


Craig_Welch2@yahoo.com wrote:

> This is helpful, thank you, although for what it is worth I took a look at
> the Synesis license and it appears to me to be a pretty standard clause -
> derived has a particular meaning which is in many such licenses - consult
> an attorney if
> you are concerned and all that.  What I was concerned about were the areas
> under dmd/src/phobos/internal and dmd/src/phobos/std - there is a mix of
> things which are copyright digital mars all rights reserved, some things
> copyrighted by
> others and clearly licensed, and some with no notices whatsoever.  Under
> the dmd/src/dmd area things are clearly marked - but dmd/src/phobos is not
> within that folder and does not have clear indications regarding license
> in the way
> that the dmd/src/dmd area does.  The license at the higher dmd level is
> clearly
> restrictive.  I certainly respect Walter and other's right to distribute
> their work as they see fit - but I cannot see being able to make use of D
> at present without those areas of the tree, and technically I do not
> believe it is open
> sourced at this point, which is one of the requirements that I have.  What
> I am looking for is clarification on that, I would like to get into the
> language and make use of it, but do not see that as worthwhile unless the
> phobos lib is
> clearly open.  It may be that I'm being too picky - forgive me if so - but
> it also may be that this is the intent (to restrictively license portions
> of phobos) and I wish to be clear on that before investing time and effort
> into working with the language.
> 
> In article <c785qi$21b$1@digitaldaemon.com>, Bruno A. Costa says...
>>
>>Craig_Welch2@yahoo.com wrote:
>>
>>> Hello,
>>> I wondered if someone (Walter) could clarify the licensing for the
>>> Phobos
>>> library components of the D distribution.  It's clear how the generously
>>> donated front-end to the compiler is licensed, and that the compiler
>>> components not included are under a proprietary license, but I was not
>>> clear on Phobos - esp.
>>> because it is being included in the gcc distribution.  I'm very
>>> interested in D but it appears that the Phobos library is essential to
>>> any usable implementation and I'm just not sure how it is being
>>> licensed.
>>> 
>>> Tbanks..
>>
>>Hi,
>>
>>This is what Ant (developer of DUI) spoke about this problem:
>>
>>-----------------------------------------------------------
>>"isn't phobos glt/artistic?
>>
>>As a preventive measure I delete all Synesis source code from
>>my machine.
>>Luckally I never used any and my project (DUI and leds) still
>>compile with the new version of phobos.
>>
>>My problem is that in the future some Synesis might see
>>some similarities on their code and mine and claim ownership
>>of something I wrote...
>>Ok here is the complete message I wrote last night:
>>
>>*** original complete message ***
>>
>>IANAL so I might be going overboard for nothing...
>>
>>please enlighten me.
>>
>>from phobos.loader:
>>
>>*              (ii) Any derived versions of this software (howsoever
>>modified) *              remain the sole property of Synesis Software.
>>
>>
>>Can we live with this?
>>I can not.
>>
>>I can't admit that in the future Synesis claims ownership of some thing I
>>wrote because it might have some similarities.
>>With the current IP and patent problems I can not afford to look at code
>>with that type of license at least until I discuss this with my lawyer.
>>
>>As preventive mesuares:
>>I just deleted all source code from Synesis from my computer.
>>I deleted any files on the phobos src tree that contained the string
>>"ynesis" (except std/mmfile.d that doesn't contain the Synesis license)
>>and all my old (previous versions) dmd.zip files.
>>and I deleted the recent downloaded dmd.zip for dmd 0.86
>>
>>Matthew, if there is another procedure to follow by someone that doesn't agree with the Synesis license let me know.
>>
>>phobos still compiles (after changes to linux.mak) and DUI and leds still compile and run with the new phobos.
>>
>>Walter, can you distribute phobos without the source code from Synesis?
>>Acording to some recente reading from growklaw If one has access to
>>some code and produces something similar it's assumed that it is a copy
>>even if the author claims never had look to the code
>>(for the US I guess).
>>
>>"header" files should be OK unless the ridiculous claim SCO made on linux .h files is sanctioned by the courts.
>>
>>Thanks,
>>
>>Ant"
>>
>>-------------------------------------------------------------
>>
>>As you can see, he removed all bad licensed sources from phobos and it still compile his softwares (DUI and LEDS).
>>
>>I think that if we want a free/open library, the best way is to remove the problematic sources or, preferably, the authors of that code change to a Free licence.
>>
>>Bruno.

I agree with you:  D is a fantastic
>programming language, but without a free
licence I don't think D becomes
>widely used so soon.

In the message
http://www.digitalmars.com/drn-bin/wwwnews?D.gnu/549

Walter told us
(answering a similar question from myself) that
"It should be under a dual
gpl/artistic license. I'll see about fixing it."

I think it would be very
nice if Walter could fix this soon, since I would
like to use D to develop
free software (and hopefully many people in the
free software comunity would
love to do the same), but if the license of
phobos is not clear, we cannot use
it.

(since for example, GPL'ed code cannot be linked legally with code
under
an incompatible license, for example propietary)

Best regards,
Pablo De Nápoli
pdenapo AT yahoo.com

In article
<c78hd9$k34$1@digitaldaemon.com>, Bruno A. Costa says...
> 
>You should share
your ideas in the digitalmars.D newsgroup. There are most
>of the users. And
some (as me) think that the license is a real problem in
>the general
acceptance of D. I am new in D, and it is a fantastic
>programming language,
but without a free licence I don't think D becomes
>widely used so soon.
>
>Bruno.
>