You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@felix.apache.org by Nicolas Lalevée <ni...@hibnet.org> on 2010/11/29 10:13:36 UTC

Missing ASL header

Hi,

I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.

I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].

But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].

Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?

cheers,
Nicolas

[1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
[2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
[3] http://svn.apache.org/viewvc?view=revision&revision=983901

Re: Missing ASL header

Posted by Nicolas Lalevée <ni...@hibnet.org>.
Le 1 déc. 2010 à 17:58, David Savage a écrit :

> 2010/12/1 Nicolas Lalevée <ni...@hibnet.org>:
>> Hi David,
>> 
>> 
>>> Ok no response on this, but I've decided to update the file in any
>>> case. Richard H pointed me at the
>>> $felix-svn/framework/src/main/resources/default.properties file that
>>> includes basically the same information as jvm-packages.properties.
>>> This includes a version attribute which I think will be useful later,
>>> so I've scrapped the previous contents of jvm-packages.properties and
>>> gone with this new format.
>>> 
>>> You should find this file (with associated ASL header) here:
>>> 
>>> http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>> 
>> finally I used a modified version of jvm-packages.properties [1] because I don't have a "property expander". Or maybe there is in Ivy but I wanted to quickly fix the unit tests. I'll probably review this later to use the new format.
>> 
>>> 
>>> Out of interest I believe bushel and sigil are providing similar
>>> capabilities - sigil has a mapping of OSGi meta data into an ivy
>>> resolution - however the way we go about it looks quite different to
>>> what bushel (does/proposes?). We perform import package resolution
>>> using an OSGi resolver outside of ivy and then synthesis ivy modules
>>> at runtime, I believe bushel is intending on pushing the meta
>>> information into ivy.xml files and then use the standard ivy resolver.
>> 
>> I think you have clearly summarized the difference here :)
>> 
>>> Not saying either approach is right/wrong I think it'd be great if
>>> there is any useful collaborations we can build once you're setup on
>>> apache - let me know?
>> 
>> The code has been imported. I am writing some doc right now. I'll try to probably write something about the difference with sigil. Hopefully I'll write something fair ;)
>> I'll keep you posted if I write anything about Sigil.
> 
> Great it's always useful to get feedback :)
> 
>> 
>> I'm not sure how we can collaborate and share some code, but I would be happy to do so. I'll keep subscribed to felix-dev.
> 
> Yep understood just to say I'm not pushing an uber answer here - in
> the end it's about making a good set of tools to build code with -
> where ever that happens is fine by me. I like what we've done with
> Sigil, but there are definitely many ways to think about these
> problems, I think at this stage trying all sorts of possibilities is
> good. Should I follow the ant-dev list for bushel?

that's it.

Note that the name "bushel" is disappearing into the "OSGi capability of Ivy".

Nicolas

> 
> Regards,
> 
> Dave
> 
>> 
>> Nicolas
>> 
>> [1] http://svn.apache.org/repos/asf/ant/ivy/core/trunk/src/java/org/apache/ivy/osgi/core/jvm-packages.properties
>> 
>> 
>>> 
>>> Regards,
>>> 
>>> Dave
>>> 
>>> On Mon, Nov 29, 2010 at 4:19 PM, David Savage <da...@paremus.com> wrote:
>>>> Hi Nicolas,
>>>> 
>>>> The commit you found replaces the profile files from Eclipse with a
>>>> new file which I generated independently. My understanding is that
>>>> given this file contains information that is public knowledge to the
>>>> Java community, is of a different format and uses different file
>>>> access semantics this constitutes it's own IP. As such I'm happy to
>>>> apply the ASL header on this basis. If that is acceptable then I will
>>>> push an update with this clarification, PM's please advise?
>>>> 
>>>> Regards,
>>>> 
>>>> Dave
>>>> 
>>>> 2010/11/29 Nicolas Lalevée <ni...@hibnet.org>:
>>>>> Hi,
>>>>> 
>>>>> I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.
>>>>> 
>>>>> I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].
>>>>> 
>>>>> But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].
>>>>> 
>>>>> Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?
>>>>> 
>>>>> cheers,
>>>>> Nicolas
>>>>> 
>>>>> [1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
>>>>> [2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>>>>> [3] http://svn.apache.org/viewvc?view=revision&revision=983901
>>>>> 
>>>>> 
>>>> 
>> 
>>

Re: Missing ASL header

Posted by David Savage <da...@paremus.com>.
2010/12/1 Nicolas Lalevée <ni...@hibnet.org>:
> Hi David,
>
>
>> Ok no response on this, but I've decided to update the file in any
>> case. Richard H pointed me at the
>> $felix-svn/framework/src/main/resources/default.properties file that
>> includes basically the same information as jvm-packages.properties.
>> This includes a version attribute which I think will be useful later,
>> so I've scrapped the previous contents of jvm-packages.properties and
>> gone with this new format.
>>
>> You should find this file (with associated ASL header) here:
>>
>> http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>
> finally I used a modified version of jvm-packages.properties [1] because I don't have a "property expander". Or maybe there is in Ivy but I wanted to quickly fix the unit tests. I'll probably review this later to use the new format.
>
>>
>> Out of interest I believe bushel and sigil are providing similar
>> capabilities - sigil has a mapping of OSGi meta data into an ivy
>> resolution - however the way we go about it looks quite different to
>> what bushel (does/proposes?). We perform import package resolution
>> using an OSGi resolver outside of ivy and then synthesis ivy modules
>> at runtime, I believe bushel is intending on pushing the meta
>> information into ivy.xml files and then use the standard ivy resolver.
>
> I think you have clearly summarized the difference here :)
>
>> Not saying either approach is right/wrong I think it'd be great if
>> there is any useful collaborations we can build once you're setup on
>> apache - let me know?
>
> The code has been imported. I am writing some doc right now. I'll try to probably write something about the difference with sigil. Hopefully I'll write something fair ;)
> I'll keep you posted if I write anything about Sigil.

Great it's always useful to get feedback :)

>
> I'm not sure how we can collaborate and share some code, but I would be happy to do so. I'll keep subscribed to felix-dev.

Yep understood just to say I'm not pushing an uber answer here - in
the end it's about making a good set of tools to build code with -
where ever that happens is fine by me. I like what we've done with
Sigil, but there are definitely many ways to think about these
problems, I think at this stage trying all sorts of possibilities is
good. Should I follow the ant-dev list for bushel?

Regards,

Dave

>
> Nicolas
>
> [1] http://svn.apache.org/repos/asf/ant/ivy/core/trunk/src/java/org/apache/ivy/osgi/core/jvm-packages.properties
>
>
>>
>> Regards,
>>
>> Dave
>>
>> On Mon, Nov 29, 2010 at 4:19 PM, David Savage <da...@paremus.com> wrote:
>>> Hi Nicolas,
>>>
>>> The commit you found replaces the profile files from Eclipse with a
>>> new file which I generated independently. My understanding is that
>>> given this file contains information that is public knowledge to the
>>> Java community, is of a different format and uses different file
>>> access semantics this constitutes it's own IP. As such I'm happy to
>>> apply the ASL header on this basis. If that is acceptable then I will
>>> push an update with this clarification, PM's please advise?
>>>
>>> Regards,
>>>
>>> Dave
>>>
>>> 2010/11/29 Nicolas Lalevée <ni...@hibnet.org>:
>>>> Hi,
>>>>
>>>> I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.
>>>>
>>>> I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].
>>>>
>>>> But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].
>>>>
>>>> Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?
>>>>
>>>> cheers,
>>>> Nicolas
>>>>
>>>> [1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
>>>> [2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>>>> [3] http://svn.apache.org/viewvc?view=revision&revision=983901
>>>>
>>>>
>>>
>
>

Re: Missing ASL header

Posted by Nicolas Lalevée <ni...@hibnet.org>.
Hi David,


> Ok no response on this, but I've decided to update the file in any
> case. Richard H pointed me at the
> $felix-svn/framework/src/main/resources/default.properties file that
> includes basically the same information as jvm-packages.properties.
> This includes a version attribute which I think will be useful later,
> so I've scrapped the previous contents of jvm-packages.properties and
> gone with this new format.
> 
> You should find this file (with associated ASL header) here:
> 
> http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties

finally I used a modified version of jvm-packages.properties [1] because I don't have a "property expander". Or maybe there is in Ivy but I wanted to quickly fix the unit tests. I'll probably review this later to use the new format.

> 
> Out of interest I believe bushel and sigil are providing similar
> capabilities - sigil has a mapping of OSGi meta data into an ivy
> resolution - however the way we go about it looks quite different to
> what bushel (does/proposes?). We perform import package resolution
> using an OSGi resolver outside of ivy and then synthesis ivy modules
> at runtime, I believe bushel is intending on pushing the meta
> information into ivy.xml files and then use the standard ivy resolver.

I think you have clearly summarized the difference here :)

> Not saying either approach is right/wrong I think it'd be great if
> there is any useful collaborations we can build once you're setup on
> apache - let me know?

The code has been imported. I am writing some doc right now. I'll try to probably write something about the difference with sigil. Hopefully I'll write something fair ;)
I'll keep you posted if I write anything about Sigil.

I'm not sure how we can collaborate and share some code, but I would be happy to do so. I'll keep subscribed to felix-dev.

Nicolas

[1] http://svn.apache.org/repos/asf/ant/ivy/core/trunk/src/java/org/apache/ivy/osgi/core/jvm-packages.properties


> 
> Regards,
> 
> Dave
> 
> On Mon, Nov 29, 2010 at 4:19 PM, David Savage <da...@paremus.com> wrote:
>> Hi Nicolas,
>> 
>> The commit you found replaces the profile files from Eclipse with a
>> new file which I generated independently. My understanding is that
>> given this file contains information that is public knowledge to the
>> Java community, is of a different format and uses different file
>> access semantics this constitutes it's own IP. As such I'm happy to
>> apply the ASL header on this basis. If that is acceptable then I will
>> push an update with this clarification, PM's please advise?
>> 
>> Regards,
>> 
>> Dave
>> 
>> 2010/11/29 Nicolas Lalevée <ni...@hibnet.org>:
>>> Hi,
>>> 
>>> I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.
>>> 
>>> I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].
>>> 
>>> But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].
>>> 
>>> Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?
>>> 
>>> cheers,
>>> Nicolas
>>> 
>>> [1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
>>> [2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>>> [3] http://svn.apache.org/viewvc?view=revision&revision=983901
>>> 
>>> 
>>

Re: Missing ASL header

Posted by David Savage <da...@paremus.com>.
Hi Nicolas,

Ok no response on this, but I've decided to update the file in any
case. Richard H pointed me at the
$felix-svn/framework/src/main/resources/default.properties file that
includes basically the same information as jvm-packages.properties.
This includes a version attribute which I think will be useful later,
so I've scrapped the previous contents of jvm-packages.properties and
gone with this new format.

You should find this file (with associated ASL header) here:

http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties

Out of interest I believe bushel and sigil are providing similar
capabilities - sigil has a mapping of OSGi meta data into an ivy
resolution - however the way we go about it looks quite different to
what bushel (does/proposes?). We perform import package resolution
using an OSGi resolver outside of ivy and then synthesis ivy modules
at runtime, I believe bushel is intending on pushing the meta
information into ivy.xml files and then use the standard ivy resolver.
Not saying either approach is right/wrong I think it'd be great if
there is any useful collaborations we can build once you're setup on
apache - let me know?

Regards,

Dave

On Mon, Nov 29, 2010 at 4:19 PM, David Savage <da...@paremus.com> wrote:
> Hi Nicolas,
>
> The commit you found replaces the profile files from Eclipse with a
> new file which I generated independently. My understanding is that
> given this file contains information that is public knowledge to the
> Java community, is of a different format and uses different file
> access semantics this constitutes it's own IP. As such I'm happy to
> apply the ASL header on this basis. If that is acceptable then I will
> push an update with this clarification, PM's please advise?
>
> Regards,
>
> Dave
>
> 2010/11/29 Nicolas Lalevée <ni...@hibnet.org>:
>> Hi,
>>
>> I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.
>>
>> I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].
>>
>> But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].
>>
>> Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?
>>
>> cheers,
>> Nicolas
>>
>> [1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
>> [2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
>> [3] http://svn.apache.org/viewvc?view=revision&revision=983901
>>
>>
>

Re: Missing ASL header

Posted by David Savage <da...@paremus.com>.
Hi Nicolas,

The commit you found replaces the profile files from Eclipse with a
new file which I generated independently. My understanding is that
given this file contains information that is public knowledge to the
Java community, is of a different format and uses different file
access semantics this constitutes it's own IP. As such I'm happy to
apply the ASL header on this basis. If that is acceptable then I will
push an update with this clarification, PM's please advise?

Regards,

Dave

2010/11/29 Nicolas Lalevée <ni...@hibnet.org>:
> Hi,
>
> I was working on the import of Bushel into Ivy [1], and there were some files under the EPL which I had to remove from the import. These files were defining the packages which should be part of the bootstrap classpath depending on the execution environment.
>
> I then thought I would find such similar files in the source of Felix but ASL friendly. I have found them: [2].
>
> But the ASL header is missing. Then I started to wonder if this file was a real ASF one, and I searched svn and I have found a commit that makes me doubt [3].
>
> Considering the IP contained is these files, I think that the new file is safe to be an ASF one. So could you add the ASL header so there won't be any doubt ?
>
> cheers,
> Nicolas
>
> [1] http://www.mail-archive.com/dev@ant.apache.org/msg42079.html
> [2] http://svn.apache.org/repos/asf/felix/trunk/sigil/common/core/profiles/jvm-packages.properties
> [3] http://svn.apache.org/viewvc?view=revision&revision=983901
>
>