You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@marmotta.apache.org by Sebastian Schaffert <ss...@apache.org> on 2013/03/19 21:49:02 UTC
[VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Dear all,
Andy rightly suggested to restart the vote for the 3.0.0-incubating release
with the correct information. So I'd like to cancel the previous vote and
restart it. Please reply to the following vote:
=====
A candidate for the Marmotta 3.0.0-incubating release is available at:
https://dist.apache.org/repos/dist/dev/incubator/marmotta/3.0.0-incubating/
The release candidate is a zip archive of the sources in:
https://git-wip-us.apache.org/repos/asf/incubator-marmotta.git
with the tag "3.0.0-incubating".
The SHA1 checksum of the archive is
38a0c67603a229a1e69df9e0f84cf60a4c7f998c.
A staged Maven repository is available for review at:
https://repository.apache.org/**content/repositories/**
orgapachemarmotta-006/<https://repository.apache.org/content/repositories/orgapachemarmotta-006/>
Please vote on releasing this package as Apache Marmotta 3.0.0-incubating.
The vote is open for the next 72 hours and passes if a majority of at
least three +1 Marmotta PMC votes are cast.
[ ] +1 Release this package as Apache Marmotta 3.0.0-incubating
[ ] 0 I don't feel strongly about it, but I'm okay with the release
[ ] -1 Do not release this package because...
Re: [CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Andy Seaborne <an...@apache.org>.
On 21/03/13 14:01, Sebastian Schaffert wrote:
> I am cancelling this vote because we are preparing a new release candidate
> with the mentioned issue regarding onejar fixed. I will open a new vote
> this afternoon (I'll also try taking into account the comments wrt the RAT
> plugin).
>
> Greetings,
>
> Sebastian
You're doing just great - having to pull a vote and restart when there
is a large amount of new stuff is just all to normal.
The first release is a significant point in the evolution of a podling
and is "high ceremony". It is as much about process as content.
All the PPMC are encouraged to vote. While only IPMC members (mentors)
have binding votes on the release, everyone is encouraged to become
familiar with the voting process and put in a +1/0/-1.
Andy
Re: [CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Andy Seaborne <an...@apache.org>.
On 21/03/13 14:15, Sergio Fernández wrote:
>
>
> On 21/03/13 15:08, Fabian Christ wrote:
>> Just a minor thing. It is good practice in such cases to change the
>> subject of the thread to [CANCELLED][VOTE] ... or something like that.
>
> and "[CANCEL][VOTE]" not? ok, sorry
Yes - it's OK
(GMail bites again? it does not make it very clear when a thread changes
subject line)
Andy
Re: [CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sergio Fernández <se...@salzburgresearch.at>.
On 21/03/13 15:08, Fabian Christ wrote:
> Just a minor thing. It is good practice in such cases to change the
> subject of the thread to [CANCELLED][VOTE] ... or something like that.
and "[CANCEL][VOTE]" not? ok, sorry
--
Sergio Fernández
Salzburg Research
+43 662 2288 318
Jakob-Haringer Strasse 5/II
A-5020 Salzburg (Austria)
http://www.salzburgresearch.at
Re: [CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
2013/3/21 Fabian Christ <ch...@googlemail.com>:
> 2013/3/21 Sebastian Schaffert <se...@gmail.com>:
>> I am cancelling this vote [...]
>
> Just a minor thing. It is good practice in such cases to change the
> subject of the thread to [CANCELLED][VOTE] ... or something like that.
Sorry for that one. My bad! I did not recognize that you did that
already. As Andy assumed was GMail hiding the changed subject from me.
--
Fabian
http://twitter.com/fctwitt
Re: [CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
2013/3/21 Sebastian Schaffert <se...@gmail.com>:
> I am cancelling this vote [...]
Just a minor thing. It is good practice in such cases to change the
subject of the thread to [CANCELLED][VOTE] ... or something like that.
The same for vote results. Just add [RESULT][VOTE] or the like.
--
Fabian
http://twitter.com/fctwitt
[CANCEL] [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <se...@gmail.com>.
I am cancelling this vote because we are preparing a new release candidate
with the mentioned issue regarding onejar fixed. I will open a new vote
this afternoon (I'll also try taking into account the comments wrt the RAT
plugin).
Greetings,
Sebastian
2013/3/21 Sergio Fernández <se...@salzburgresearch.at>
> I quickly check for other projects with similar issue, and the solution
> they adopted to solve CAMEL-1765 (just do not use it) is not valid for us.
>
>
> On 21/03/13 10:43, Sebastian Schaffert wrote:
>
>> Hi Nandana,
>>
>> I take back my suggestion, because the OneJar license says:
>>
>> "Including this file inside the built One-JAR file conforms with these
>> terms."
>>
>>
>> The Maven OneJar plugin automatically adds the license to the jar file
>> (in the doc/ subdirectory). So this should at least be enough for the
>> OneJar people. Is it also ok for the ASF?
>>
>>
>> Greetings,
>>
>>
>> Sebastian
>>
>>
>>
>> 2013/3/21 Sebastian Schaffert<se...@gmail.com>
>> >
>>
>> Hi Nandana,
>>>
>>> thanks for pointing this out. Actually these files are added by the Maven
>>> OneJar Plugin, which builds the jar file for the LDPath CLI. I was not
>>> aware that it adds additional source files to the jar file that are not
>>> from the project. To solve this issue, I propose that I create a new
>>> release build including the license that you pointed out. Do you think it
>>> is necessary to again restart the vote afterwards?
>>>
>>> Greetings,
>>>
>>> Sebastian
>>>
>>>
>>> 2013/3/21 Nandana Mihindukulasooriya<na...@gmail.com>
>>> >
>>>
>>> Hi all,
>>>>
>>>> Kudos for all the hard work pushing forward the release and sorry for
>>>> not
>>>> being able to check and provide feedback earlier on. I am a bit stuck
>>>> catching up stuff after one week of travelling.
>>>>
>>>> One question, ldpath-3.0.0-incubating.jar (inside
>>>> apache-marmotta-3.0.0-**incubating-ldpath.tar.gz) contains some source
>>>> files
>>>> from http://one-jar.sourceforge.**net/<http://one-jar.sourceforge.net/>.
>>>> Was this intentional ?
>>>>
>>>> I find no reference to one jar licence in the LICENCE.txt
>>>>
>>>> Best Regards,
>>>> Nandana
>>>>
>>>> [1] - http://one-jar.sourceforge.**net/one-jar-license.html<http://one-jar.sourceforge.net/one-jar-license.html>
>>>>
>>>>
>>>> On Wed, Mar 20, 2013 at 3:09 PM, Fabian Christ<
>>>> christ.fabian@googlemail.com
>>>>
>>>>> wrote:
>>>>>
>>>>
>>>> 2013/3/20 Fabian Christ<ch...@googlemail.com>
>>>>> >:
>>>>>
>>>>>> Summary vote: 0 (for the moment) because I am unsure about the rat
>>>>>> check. At least I am not really convinced that the license headers are
>>>>>> fine in all cases.
>>>>>>
>>>>>
>>>>> After the very constructive discussions on the RAT topic I came to the
>>>>> conclusion that the podling is handling the license headers in a
>>>>> profound way.
>>>>>
>>>>> Therefore, I withdraw my previous +0 vote and vote with +1 for the
>>>>>
>>>> release.
>>>>
>>>>>
>>>>> --
>>>>> Fabian
>>>>> http://twitter.com/fctwitt
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Best Regards,
>>>> Nandana Mihindukulasooriya
>>>>
>>>>
>>>
>>>
>>
> --
> Sergio Fernández
> Salzburg Research
> +43 662 2288 318
> Jakob-Haringer Strasse 5/II
> A-5020 Salzburg (Austria)
> http://www.salzburgresearch.at
>
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sergio Fernández <se...@salzburgresearch.at>.
I quickly check for other projects with similar issue, and the solution
they adopted to solve CAMEL-1765 (just do not use it) is not valid for us.
On 21/03/13 10:43, Sebastian Schaffert wrote:
> Hi Nandana,
>
> I take back my suggestion, because the OneJar license says:
>
> "Including this file inside the built One-JAR file conforms with these terms."
>
>
> The Maven OneJar plugin automatically adds the license to the jar file
> (in the doc/ subdirectory). So this should at least be enough for the
> OneJar people. Is it also ok for the ASF?
>
>
> Greetings,
>
>
> Sebastian
>
>
>
> 2013/3/21 Sebastian Schaffert<se...@gmail.com>
>
>> Hi Nandana,
>>
>> thanks for pointing this out. Actually these files are added by the Maven
>> OneJar Plugin, which builds the jar file for the LDPath CLI. I was not
>> aware that it adds additional source files to the jar file that are not
>> from the project. To solve this issue, I propose that I create a new
>> release build including the license that you pointed out. Do you think it
>> is necessary to again restart the vote afterwards?
>>
>> Greetings,
>>
>> Sebastian
>>
>>
>> 2013/3/21 Nandana Mihindukulasooriya<na...@gmail.com>
>>
>>> Hi all,
>>>
>>> Kudos for all the hard work pushing forward the release and sorry for not
>>> being able to check and provide feedback earlier on. I am a bit stuck
>>> catching up stuff after one week of travelling.
>>>
>>> One question, ldpath-3.0.0-incubating.jar (inside
>>> apache-marmotta-3.0.0-incubating-ldpath.tar.gz) contains some source files
>>> from http://one-jar.sourceforge.net/. Was this intentional ?
>>>
>>> I find no reference to one jar licence in the LICENCE.txt
>>>
>>> Best Regards,
>>> Nandana
>>>
>>> [1] - http://one-jar.sourceforge.net/one-jar-license.html
>>>
>>>
>>> On Wed, Mar 20, 2013 at 3:09 PM, Fabian Christ<
>>> christ.fabian@googlemail.com
>>>> wrote:
>>>
>>>> 2013/3/20 Fabian Christ<ch...@googlemail.com>:
>>>>> Summary vote: 0 (for the moment) because I am unsure about the rat
>>>>> check. At least I am not really convinced that the license headers are
>>>>> fine in all cases.
>>>>
>>>> After the very constructive discussions on the RAT topic I came to the
>>>> conclusion that the podling is handling the license headers in a
>>>> profound way.
>>>>
>>>> Therefore, I withdraw my previous +0 vote and vote with +1 for the
>>> release.
>>>>
>>>> --
>>>> Fabian
>>>> http://twitter.com/fctwitt
>>>>
>>>
>>>
>>>
>>> --
>>> Best Regards,
>>> Nandana Mihindukulasooriya
>>>
>>
>>
>
--
Sergio Fernández
Salzburg Research
+43 662 2288 318
Jakob-Haringer Strasse 5/II
A-5020 Salzburg (Austria)
http://www.salzburgresearch.at
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <se...@gmail.com>.
Hi Nandana,
I take back my suggestion, because the OneJar license says:
"Including this file inside the built One-JAR file conforms with these terms."
The Maven OneJar plugin automatically adds the license to the jar file
(in the doc/ subdirectory). So this should at least be enough for the
OneJar people. Is it also ok for the ASF?
Greetings,
Sebastian
2013/3/21 Sebastian Schaffert <se...@gmail.com>
> Hi Nandana,
>
> thanks for pointing this out. Actually these files are added by the Maven
> OneJar Plugin, which builds the jar file for the LDPath CLI. I was not
> aware that it adds additional source files to the jar file that are not
> from the project. To solve this issue, I propose that I create a new
> release build including the license that you pointed out. Do you think it
> is necessary to again restart the vote afterwards?
>
> Greetings,
>
> Sebastian
>
>
> 2013/3/21 Nandana Mihindukulasooriya <na...@gmail.com>
>
>> Hi all,
>>
>> Kudos for all the hard work pushing forward the release and sorry for not
>> being able to check and provide feedback earlier on. I am a bit stuck
>> catching up stuff after one week of travelling.
>>
>> One question, ldpath-3.0.0-incubating.jar (inside
>> apache-marmotta-3.0.0-incubating-ldpath.tar.gz) contains some source files
>> from http://one-jar.sourceforge.net/. Was this intentional ?
>>
>> I find no reference to one jar licence in the LICENCE.txt
>>
>> Best Regards,
>> Nandana
>>
>> [1] - http://one-jar.sourceforge.net/one-jar-license.html
>>
>>
>> On Wed, Mar 20, 2013 at 3:09 PM, Fabian Christ <
>> christ.fabian@googlemail.com
>> > wrote:
>>
>> > 2013/3/20 Fabian Christ <ch...@googlemail.com>:
>> > > Summary vote: 0 (for the moment) because I am unsure about the rat
>> > > check. At least I am not really convinced that the license headers are
>> > > fine in all cases.
>> >
>> > After the very constructive discussions on the RAT topic I came to the
>> > conclusion that the podling is handling the license headers in a
>> > profound way.
>> >
>> > Therefore, I withdraw my previous +0 vote and vote with +1 for the
>> release.
>> >
>> > --
>> > Fabian
>> > http://twitter.com/fctwitt
>> >
>>
>>
>>
>> --
>> Best Regards,
>> Nandana Mihindukulasooriya
>>
>
>
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <se...@gmail.com>.
Hi Nandana,
thanks for pointing this out. Actually these files are added by the Maven
OneJar Plugin, which builds the jar file for the LDPath CLI. I was not
aware that it adds additional source files to the jar file that are not
from the project. To solve this issue, I propose that I create a new
release build including the license that you pointed out. Do you think it
is necessary to again restart the vote afterwards?
Greetings,
Sebastian
2013/3/21 Nandana Mihindukulasooriya <na...@gmail.com>
> Hi all,
>
> Kudos for all the hard work pushing forward the release and sorry for not
> being able to check and provide feedback earlier on. I am a bit stuck
> catching up stuff after one week of travelling.
>
> One question, ldpath-3.0.0-incubating.jar (inside
> apache-marmotta-3.0.0-incubating-ldpath.tar.gz) contains some source files
> from http://one-jar.sourceforge.net/. Was this intentional ?
>
> I find no reference to one jar licence in the LICENCE.txt
>
> Best Regards,
> Nandana
>
> [1] - http://one-jar.sourceforge.net/one-jar-license.html
>
>
> On Wed, Mar 20, 2013 at 3:09 PM, Fabian Christ <
> christ.fabian@googlemail.com
> > wrote:
>
> > 2013/3/20 Fabian Christ <ch...@googlemail.com>:
> > > Summary vote: 0 (for the moment) because I am unsure about the rat
> > > check. At least I am not really convinced that the license headers are
> > > fine in all cases.
> >
> > After the very constructive discussions on the RAT topic I came to the
> > conclusion that the podling is handling the license headers in a
> > profound way.
> >
> > Therefore, I withdraw my previous +0 vote and vote with +1 for the
> release.
> >
> > --
> > Fabian
> > http://twitter.com/fctwitt
> >
>
>
>
> --
> Best Regards,
> Nandana Mihindukulasooriya
>
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Nandana Mihindukulasooriya <na...@gmail.com>.
Hi all,
Kudos for all the hard work pushing forward the release and sorry for not
being able to check and provide feedback earlier on. I am a bit stuck
catching up stuff after one week of travelling.
One question, ldpath-3.0.0-incubating.jar (inside
apache-marmotta-3.0.0-incubating-ldpath.tar.gz) contains some source files
from http://one-jar.sourceforge.net/. Was this intentional ?
I find no reference to one jar licence in the LICENCE.txt
Best Regards,
Nandana
[1] - http://one-jar.sourceforge.net/one-jar-license.html
On Wed, Mar 20, 2013 at 3:09 PM, Fabian Christ <christ.fabian@googlemail.com
> wrote:
> 2013/3/20 Fabian Christ <ch...@googlemail.com>:
> > Summary vote: 0 (for the moment) because I am unsure about the rat
> > check. At least I am not really convinced that the license headers are
> > fine in all cases.
>
> After the very constructive discussions on the RAT topic I came to the
> conclusion that the podling is handling the license headers in a
> profound way.
>
> Therefore, I withdraw my previous +0 vote and vote with +1 for the release.
>
> --
> Fabian
> http://twitter.com/fctwitt
>
--
Best Regards,
Nandana Mihindukulasooriya
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
2013/3/20 Fabian Christ <ch...@googlemail.com>:
> Summary vote: 0 (for the moment) because I am unsure about the rat
> check. At least I am not really convinced that the license headers are
> fine in all cases.
After the very constructive discussions on the RAT topic I came to the
conclusion that the podling is handling the license headers in a
profound way.
Therefore, I withdraw my previous +0 vote and vote with +1 for the release.
--
Fabian
http://twitter.com/fctwitt
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
2013/3/20 Sebastian Schaffert <se...@gmail.com>:
> For the future I'd like to install a policy that such files should only be
> added to the source tree if they are created originally by the committer.
> Once we have the Jenkins server running I will add the RAT plugin to the
> default build cycle so noone can add anything without it being checked -
> then is also the time to restrict the rules further (and possibly create
> one or more separate exclude files with more explicit patterns). For now,
> all excluded files have been checked manually before the release (this is
> what I spent my last Friday with...).
Okay I think we have stressed the point enough. I see that the podling
is aware of the problem and has a plan how to proceed and handle it.
--
Fabian
http://twitter.com/fctwitt
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <se...@gmail.com>.
Hi Fabian,
2013/3/20 Fabian Christ <ch...@googlemail.com>
> Regarding the RAT check:
>
> 2013/3/20 Sebastian Schaffert <se...@gmail.com>:
> > I was looking at how Stanbol does it and was not too happy about it,
> > because it creates a maintenance headache with too many exclude files
> > distributed over the source tree. I prefer having an overview over what
> is
> > excluded so it is easy to fix.
>
> Yes, it is a maintenance headache. But otherwise I have not seen a way
> to control it properly.
Well, I consider the way we did it to "control it properly". I have the
same options in the parent POM that I have in individual exclude files. I
can also explicitly specify individual files or directories there without
applying a general rule.
The reason for the general rules is as I explained - these files do not
support a comment format.
> We (in Stanbol) are doing it this way based on
> the comments by our mentor Bertrand during incubation. We even place a
> README somewhere near the excluded file that informs about the license
> of that file.
>
> I understand your viewpoint but how does the Marmotta podling itself
> check the headers? How do you convince yourself?
By checking each of the excluded files individually before the release and
tracking where they came from - same as I do for other dependencies that
are explicitly excluded (like jQuery). For most it is easy (kwrl and ldpath
are our own formats, so they will be under our copyright). For some it
requires checking who added them when.
For the future I'd like to install a policy that such files should only be
added to the source tree if they are created originally by the committer.
Once we have the Jenkins server running I will add the RAT plugin to the
default build cycle so noone can add anything without it being checked -
then is also the time to restrict the rules further (and possibly create
one or more separate exclude files with more explicit patterns). For now,
all excluded files have been checked manually before the release (this is
what I spent my last Friday with...).
> The fact that some
> file formats do not support comments does not prevent you from the
> burden to specify the license of such a specific file somehow. The
> problem is that there is AFAIK no default for files without any
> license information.
>
Hmm, I checked the documentation on the Apache website and it does not
provide much information about it. We could treat such files as "binary
files", which would simply require NOTICE in case they come from a third
party, and otherwise they are already covered by the main LICENSE file.
This is the way we are handling the situation at the moment.
http://www.apache.org/legal/src-headers.html
> But since I am also rather new in the Apache world, this may be a
> situation where in practice it could be handled as you did. I would
> like to hear the opinion of the other mentors on this.
>
>
Andy, any suggestions?
Greetings,
Sebastian
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
Regarding the RAT check:
2013/3/20 Sebastian Schaffert <se...@gmail.com>:
> I was looking at how Stanbol does it and was not too happy about it,
> because it creates a maintenance headache with too many exclude files
> distributed over the source tree. I prefer having an overview over what is
> excluded so it is easy to fix.
Yes, it is a maintenance headache. But otherwise I have not seen a way
to control it properly. We (in Stanbol) are doing it this way based on
the comments by our mentor Bertrand during incubation. We even place a
README somewhere near the excluded file that informs about the license
of that file.
I understand your viewpoint but how does the Marmotta podling itself
check the headers? How do you convince yourself? The fact that some
file formats do not support comments does not prevent you from the
burden to specify the license of such a specific file somehow. The
problem is that there is AFAIK no default for files without any
license information.
But since I am also rather new in the Apache world, this may be a
situation where in practice it could be handled as you did. I would
like to hear the opinion of the other mentors on this.
--
Fabian
http://twitter.com/fctwitt
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
2013/3/20 Sebastian Schaffert <se...@gmail.com>:
>>
>> LICENSE looks good for all packages
>> NOTICE same as above. "Additionally, it includes the following 3rd
>> party modules:" The list is not necessary. The NOTICE is not a file to
>> list dependencies or included modules. You may create an additional
>> file for such information.
>>
>
> So now I have a bit of a contradiction. First Andy and you said we need to
> add all the 3rd party libraries included in the distribution there that are
> not licensed by the Apache Software Foundation. Now remove them again?
Maybe there was a misunderstanding but I have never said that this
list should be included as NOTICE.
In [1] it says: "The remainder of the NOTICE file is to be used for
required third-party notices.". Listing the included third-party libs
is not such a required notice, according to [2].
[1] http://www.apache.org/legal/src-headers.html
[2] http://apache.org/legal/resolved.html#required-third-party-notices
I would put the list of modules in a DEPENDENCY or MODULES file or
something similar.
But, as I said, for this release it should be okay and something to be
fixed in future releases.
--
Fabian
http://twitter.com/fctwitt
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <se...@gmail.com>.
Hi Fabian,
trying to answer to some of your doubts in the hope you change your vote to
+1 ;-)
2013/3/20 Fabian Christ <ch...@googlemail.com>
>
>
> ###################
>
> Checking ./apache-marmotta-3.0.0-incubating-src.tar.gz
>
> NOTICE.txt
>
> The following notice is not required and could be removed. I think it
> is okay for this release but it should be removed to keep the NOTICE
> minimal. The reason for removal is that these are not legal statements
> that are required by the licenses of the those bundles.
>
> In addition, the source distribution of this product contains:
> - software based on IzPack licensed under Apache License 2.0
> - software based on D3.js licensed under New BSD License
> - software based on Dracula Graph Library licensed under MIT License
> - software based on strftime licensed under New BSD License
> - software based on JQuery licensed under MIT License
> - software based on CodeMirror licensed under Apache License 2.0
> - software based on D2R Snorql licensed under Apache License 2.0
> - software based on script.aculo.us licensed under Apache License 2.0
> - software based on Prototype licensed under Apache License 2.0
> - software based on SPARQL Flint Editor licensed under Apache License 2.0
> - software based on Sgvizler license under a MIT-style license
> - software based on MigLayout licensed under New BSD License
> - software based on SwingBits licensed under New BSD License
> - software based on rometools licensed under the Apache License 2.0
> - software based on jsonld-java licensed under the New BSD License
>
> Note: "Sgvizler license under a MIT-style license" -> In LICENSE.txt
> it is listed as "MIT". Homepage also just says "MIT" and not
> "MIT-style".
>
We will remove this in the future (or put it in a different file). We
thought it would be good to have somewhere a place where we list all 3rd
party modules we are using.
>
> RAT check: mvn apache-rat:check
> The apache-rat check always uses the following excludes defined in the
> parent POM for all artifacts
> Exclude: **/*.txt
> Exclude: **/atlassian-ide-plugin.xml
> Exclude: **/README.*
> Exclude: **/NOTICE.*
> Exclude: **/.*
> Exclude: **/.*/*
> Exclude: **/.*/**/*
> Exclude: **/target/**
> Exclude: doc/api/**/*
> Exclude: .gitignore
> Exclude: .git/**
> Exclude: *.sh
> Exclude: .git/**
> Exclude: .idea/**
> Exclude: .eclipse/**
> Exclude: **/*.iml
> Exclude: **/*.ipr
> Exclude: **/*.iws
> Exclude: **/*.psd
> Exclude: **/*.out
> Exclude: **/*.log
> Exclude: **/jquery*.js
> Exclude: **/jquery*.css
> Exclude: **/foundation.js
> Exclude: **/strftime.js
> Exclude: **/customforms.js
> Exclude: **/pageguide.js
> Exclude: **/raphael-dracula.pack.min.js
> Exclude: **/dygraph/**
> Exclude: **/sgvizler/**
> Exclude: **/js/lib/**
> Exclude: **/src/ext/**
> Exclude: **/src/main/resources/task-dialog*.properties
> Exclude:
> **/src/test/resources/org/apache/marmotta/commons/sesame/rio/jsonld/**
> Exclude:
> **/src/test/resources/org/apache/marmotta/commons/sesame/rio/vcard/*.vcf
> Exclude: **/META-INF/**
> Exclude: **/*.kwrl
> Exclude: **/*.ldpath
> Exclude: **/*.search
> Exclude: **/*.ics
> Exclude: **/*.json
>
> The problem with this configuration is that you will not be informed
> about a missing license header, e.g., in some *.json file. This way
> someone may use in the future a JSON file that is missing license
> information (may even have another license) and the rat check will
> pass. With this RAT report I can not be really sure that everything is
> fine. Just too many global excludes. IMHO excludes should be
> configured and documented on a per artifact level. This is at least
> what we do in Stanbol.
>
I configured the RAT plugin to the extent that is reasonable. If file
extensions are excluded in general, it is because these formats do not have
guaranteed support for comments, so we cannot add a license header. This is
true for
- KWRL is our reasoner rule specification (.kwrl)
- LDPath is our query language specification (.ldpath and .search)
- ICal Calender (.ics)
- JSON (.json)
- TXT (.txt)
Note that according to the JSON specification, there is no comment format
for JSON guaranteed.
In addition, we excluded "known 3rd party" modules. The remaining excluded
files are those that are also in .gitignore (e.g. IntelliJ files).
I was looking at how Stanbol does it and was not too happy about it,
because it creates a maintenance headache with too many exclude files
distributed over the source tree. I prefer having an overview over what is
excluded so it is easy to fix.
>
> I do not know if this is a blocker. Is there another way to check that
> all files have proper license headers?
>
>
The RAT plugin alone can in any case only support in the release process.
In the end, it will always be necessary to do a manual check.
> ###################
>
> Checking ./apache-marmotta-3.0.0-incubating-installer.tar.gz
> Checking ./apache-marmotta-3.0.0-incubating-ldpath.tar.gz
> Checking ./apache-marmotta-3.0.0-incubating-webapp.tar.gz
>
> While extracting the tar.gz I realized that all the different archives
> are extracted to "apache-marmotta-3.0.0-incubating". This is a bit
> confusing. Why not use "apache-marmotta-3.0.0-incubating-installer",
> "apache-marmotta-3.0.0-incubating-ldpath" etc?
>
Also thought about it. Currently, it simply takes the default value (which
is the name without classifier). Should be easy to change in the future.
>
> LICENSE looks good for all packages
> NOTICE same as above. "Additionally, it includes the following 3rd
> party modules:" The list is not necessary. The NOTICE is not a file to
> list dependencies or included modules. You may create an additional
> file for such information.
>
So now I have a bit of a contradiction. First Andy and you said we need to
add all the 3rd party libraries included in the distribution there that are
not licensed by the Apache Software Foundation. Now remove them again?
>
> ###################
>
> Summary vote: 0 (for the moment) because I am unsure about the rat
> check. At least I am not really convinced that the license headers are
> fine in all cases.
>
I hope I could convince you that the files that are excluded are excluded
because
- the format does not support comments, so no license header possible OR
- the files come from a 3rd party module its own license OR
- the files are temporary configuration (e.g. from the IDE) or generated
(e.g. target, doc) files
;-)
>
> Beside that, really great work guys! Really impressive how fast you
> are adopting to the Apache way and learning all these details. The
> issues that I raised are only minor and easy to fix. Maybe I am
> nit-picking a bit but it is just to make you aware of certain details.
> I think, you are on very good way.
>
Thanks ;-)
Greetings,
Sebastian
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Andy Seaborne <an...@apache.org>.
On 20/03/13 10:00, Fabian Christ wrote:
> Checking ./apache-marmotta-3.0.0-incubating-src.tar.gz
>
> NOTICE.txt
>
> The following notice is not required and could be removed. I think it
> is okay for this release but it should be removed to keep the NOTICE
> minimal. The reason for removal is that these are not legal statements
> that are required by the licenses of the those bundles.
>
> In addition, the source distribution of this product contains:
> - software based on IzPack licensed under Apache License 2.0
> - software based on D3.js licensed under New BSD License
> - software based on Dracula Graph Library licensed under MIT License
> - software based on strftime licensed under New BSD License
> - software based on JQuery licensed under MIT License
> - software based on CodeMirror licensed under Apache License 2.0
> - software based on D2R Snorql licensed under Apache License 2.0
> - software based on script.aculo.us licensed under Apache License 2.0
> - software based on Prototype licensed under Apache License 2.0
> - software based on SPARQL Flint Editor licensed under Apache License 2.0
> - software based on Sgvizler license under a MIT-style license
> - software based on MigLayout licensed under New BSD License
> - software based on SwingBits licensed under New BSD License
> - software based on rometools licensed under the Apache License 2.0
> - software based on jsonld-java licensed under the New BSD License
>
> Note: "Sgvizler license under a MIT-style license" -> In LICENSE.txt
> it is listed as "MIT". Homepage also just says "MIT" and not
> "MIT-style".
This is for included source in the source-release, which makes it
different to depending on a binary.
http://www.apache.org/dev/licensing-howto.html#permissive-deps is about
bundling.
When it's included source code, in the source-release, I'm not clear
whether that makes a need for something in NOTICE.
Ah ...
http://www.apache.org/dev/licensing-howto.html#mod-notice
No need if the original material has a copyright notice in place already.
So if you're taking some part of BSD-type source code, and the copyright
notice isn't in the source-release files, then something in NOTICE is
required. If the copyright notice of the BSD item is in-place, no need
to put something in NOTICE.
Andy
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Fabian Christ <ch...@googlemail.com>.
Hi,
here is my release check report:
Checking signatures and digests:
./apache-marmotta-3.0.0-incubating-installer.zip
gpg: GOOD
md5 : GOOD (ea8139b858465f698c5dbb0d78c16d59)
sha1 : GOOD (d79a51b21fa9a1b4c6e3c3a340ae4fb91a8768ff)
./apache-marmotta-3.0.0-incubating-src.zip
gpg: GOOD
md5 : GOOD (10e85e05e70f1516064dcd6275c29d87)
sha1 : GOOD (38a0c67603a229a1e69df9e0f84cf60a4c7f998c)
./apache-marmotta-3.0.0-incubating-installer.tar.gz
gpg: GOOD
md5 : GOOD (b65ada108e4a5a77d7ee9b082f541e09)
sha1 : GOOD (46878a7d526d1f2772cae3a09f1bf6b80b0f320f)
./apache-marmotta-3.0.0-incubating-ldpath.zip
gpg: GOOD
md5 : GOOD (8f9ea8941e4b95e75d9401407fb2584c)
sha1 : GOOD (375919d247506418d7cd665eb476a167e5aaddbc)
./apache-marmotta-3.0.0-incubating-webapp.tar.gz
gpg: GOOD
md5 : GOOD (357650a1ce9d9c152809b09b7d6c6746)
sha1 : GOOD (a39a22d81a39e44b87eb602c167df5445576c6c6)
./apache-marmotta-3.0.0-incubating-webapp.zip
gpg: GOOD
md5 : GOOD (5d2fb7b0e7f7aeb5a0af63593e92ee02)
sha1 : GOOD (b205e7fd4ad7417296a52dd5a7971566dced9eab)
./apache-marmotta-3.0.0-incubating-ldpath.tar.gz
gpg: GOOD
md5 : GOOD (5def4c8565f16a2d8cdde0583813594e)
sha1 : GOOD (3c859a0f799f0c3226d08e26551d327f12d14d5a)
./apache-marmotta-3.0.0-incubating-src.tar.gz
gpg: GOOD
md5 : GOOD (1aaa20cd9141a064bd2db5703bdbcff2)
sha1 : GOOD (81ea6669516ae30f9a2a56489e673e9635c5eab7)
###################
Checking ./apache-marmotta-3.0.0-incubating-src.tar.gz
NOTICE.txt
The following notice is not required and could be removed. I think it
is okay for this release but it should be removed to keep the NOTICE
minimal. The reason for removal is that these are not legal statements
that are required by the licenses of the those bundles.
In addition, the source distribution of this product contains:
- software based on IzPack licensed under Apache License 2.0
- software based on D3.js licensed under New BSD License
- software based on Dracula Graph Library licensed under MIT License
- software based on strftime licensed under New BSD License
- software based on JQuery licensed under MIT License
- software based on CodeMirror licensed under Apache License 2.0
- software based on D2R Snorql licensed under Apache License 2.0
- software based on script.aculo.us licensed under Apache License 2.0
- software based on Prototype licensed under Apache License 2.0
- software based on SPARQL Flint Editor licensed under Apache License 2.0
- software based on Sgvizler license under a MIT-style license
- software based on MigLayout licensed under New BSD License
- software based on SwingBits licensed under New BSD License
- software based on rometools licensed under the Apache License 2.0
- software based on jsonld-java licensed under the New BSD License
Note: "Sgvizler license under a MIT-style license" -> In LICENSE.txt
it is listed as "MIT". Homepage also just says "MIT" and not
"MIT-style".
LICENSE.txt looks good.
Checking release matches tag:
Only in tags/3.0.0-incubating/: .git
Only in tags/3.0.0-incubating/: .gitignore
Only in tags/3.0.0-incubating//platform/marmotta-sparql/src/main/resources/web/admin/editor:
.gitignore
-> Looks good.
Checking build: mvn clean install
- > Build Success
RAT check: mvn apache-rat:check
The apache-rat check always uses the following excludes defined in the
parent POM for all artifacts
Exclude: **/*.txt
Exclude: **/atlassian-ide-plugin.xml
Exclude: **/README.*
Exclude: **/NOTICE.*
Exclude: **/.*
Exclude: **/.*/*
Exclude: **/.*/**/*
Exclude: **/target/**
Exclude: doc/api/**/*
Exclude: .gitignore
Exclude: .git/**
Exclude: *.sh
Exclude: .git/**
Exclude: .idea/**
Exclude: .eclipse/**
Exclude: **/*.iml
Exclude: **/*.ipr
Exclude: **/*.iws
Exclude: **/*.psd
Exclude: **/*.out
Exclude: **/*.log
Exclude: **/jquery*.js
Exclude: **/jquery*.css
Exclude: **/foundation.js
Exclude: **/strftime.js
Exclude: **/customforms.js
Exclude: **/pageguide.js
Exclude: **/raphael-dracula.pack.min.js
Exclude: **/dygraph/**
Exclude: **/sgvizler/**
Exclude: **/js/lib/**
Exclude: **/src/ext/**
Exclude: **/src/main/resources/task-dialog*.properties
Exclude: **/src/test/resources/org/apache/marmotta/commons/sesame/rio/jsonld/**
Exclude: **/src/test/resources/org/apache/marmotta/commons/sesame/rio/vcard/*.vcf
Exclude: **/META-INF/**
Exclude: **/*.kwrl
Exclude: **/*.ldpath
Exclude: **/*.search
Exclude: **/*.ics
Exclude: **/*.json
The problem with this configuration is that you will not be informed
about a missing license header, e.g., in some *.json file. This way
someone may use in the future a JSON file that is missing license
information (may even have another license) and the rat check will
pass. With this RAT report I can not be really sure that everything is
fine. Just too many global excludes. IMHO excludes should be
configured and documented on a per artifact level. This is at least
what we do in Stanbol.
I do not know if this is a blocker. Is there another way to check that
all files have proper license headers?
###################
Checking ./apache-marmotta-3.0.0-incubating-installer.tar.gz
Checking ./apache-marmotta-3.0.0-incubating-ldpath.tar.gz
Checking ./apache-marmotta-3.0.0-incubating-webapp.tar.gz
While extracting the tar.gz I realized that all the different archives
are extracted to "apache-marmotta-3.0.0-incubating". This is a bit
confusing. Why not use "apache-marmotta-3.0.0-incubating-installer",
"apache-marmotta-3.0.0-incubating-ldpath" etc?
LICENSE looks good for all packages
NOTICE same as above. "Additionally, it includes the following 3rd
party modules:" The list is not necessary. The NOTICE is not a file to
list dependencies or included modules. You may create an additional
file for such information.
###################
Summary vote: 0 (for the moment) because I am unsure about the rat
check. At least I am not really convinced that the license headers are
fine in all cases.
Beside that, really great work guys! Really impressive how fast you
are adopting to the Apache way and learning all these details. The
issues that I raised are only minor and easy to fix. Maybe I am
nit-picking a bit but it is just to make you aware of certain details.
I think, you are on very good way.
Best,
- Fabian
2013/3/19 Sebastian Schaffert <ss...@apache.org>:
> Dear all,
>
> Andy rightly suggested to restart the vote for the 3.0.0-incubating release
> with the correct information. So I'd like to cancel the previous vote and
> restart it. Please reply to the following vote:
>
> =====
> A candidate for the Marmotta 3.0.0-incubating release is available at:
>
> https://dist.apache.org/repos/dist/dev/incubator/marmotta/3.0.0-incubating/
>
> The release candidate is a zip archive of the sources in:
>
> https://git-wip-us.apache.org/repos/asf/incubator-marmotta.git
>
> with the tag "3.0.0-incubating".
>
> The SHA1 checksum of the archive is
> 38a0c67603a229a1e69df9e0f84cf60a4c7f998c.
>
> A staged Maven repository is available for review at:
>
> https://repository.apache.org/**content/repositories/**
> orgapachemarmotta-006/<https://repository.apache.org/content/repositories/orgapachemarmotta-006/>
>
> Please vote on releasing this package as Apache Marmotta 3.0.0-incubating.
> The vote is open for the next 72 hours and passes if a majority of at
> least three +1 Marmotta PMC votes are cast.
>
> [ ] +1 Release this package as Apache Marmotta 3.0.0-incubating
> [ ] 0 I don't feel strongly about it, but I'm okay with the release
> [ ] -1 Do not release this package because...
--
Fabian
http://twitter.com/fctwitt
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sergio Fernández <wi...@apache.org>.
Hi,
- signature: OK
- build: OK
- dependencies: OK
- license and notice: OK
So +1 (binding)
Cheers,
On 19/03/13 21:49, Sebastian Schaffert wrote:
> Dear all,
>
> Andy rightly suggested to restart the vote for the 3.0.0-incubating release
> with the correct information. So I'd like to cancel the previous vote and
> restart it. Please reply to the following vote:
>
> =====
> A candidate for the Marmotta 3.0.0-incubating release is available at:
>
> https://dist.apache.org/repos/dist/dev/incubator/marmotta/3.0.0-incubating/
>
> The release candidate is a zip archive of the sources in:
>
> https://git-wip-us.apache.org/repos/asf/incubator-marmotta.git
>
> with the tag "3.0.0-incubating".
>
> The SHA1 checksum of the archive is
> 38a0c67603a229a1e69df9e0f84cf60a4c7f998c.
>
> A staged Maven repository is available for review at:
>
> https://repository.apache.org/**content/repositories/**
> orgapachemarmotta-006/<https://repository.apache.org/content/repositories/orgapachemarmotta-006/>
>
> Please vote on releasing this package as Apache Marmotta 3.0.0-incubating.
> The vote is open for the next 72 hours and passes if a majority of at
> least three +1 Marmotta PMC votes are cast.
>
> [ ] +1 Release this package as Apache Marmotta 3.0.0-incubating
> [ ] 0 I don't feel strongly about it, but I'm okay with the release
> [ ] -1 Do not release this package because...
>
--
Sergio Fernández
Re: [VOTE] Release Apache Marmotta 3.0.0-incubating (RESTARTED)
Posted by Sebastian Schaffert <ss...@apache.org>.
>From my side also
+1 (binding)
P.S.: Sergio: ignore my previous mail, apparently the reply-to-all did not
work
2013/3/19 Sebastian Schaffert <ss...@apache.org>
> Dear all,
>
> Andy rightly suggested to restart the vote for the 3.0.0-incubating
> release with the correct information. So I'd like to cancel the previous
> vote and restart it. Please reply to the following vote:
>
> =====
> A candidate for the Marmotta 3.0.0-incubating release is available at:
>
> https://dist.apache.org/repos/dist/dev/incubator/marmotta/3.0.0-incubating/
>
> The release candidate is a zip archive of the sources in:
>
> https://git-wip-us.apache.org/repos/asf/incubator-marmotta.git
>
> with the tag "3.0.0-incubating".
>
> The SHA1 checksum of the archive is
> 38a0c67603a229a1e69df9e0f84cf60a4c7f998c.
>
> A staged Maven repository is available for review at:
>
> https://repository.apache.org/**content/repositories/**
> orgapachemarmotta-006/<https://repository.apache.org/content/repositories/orgapachemarmotta-006/>
>
> Please vote on releasing this package as Apache Marmotta 3.0.0-incubating.
> The vote is open for the next 72 hours and passes if a majority of at
> least three +1 Marmotta PMC votes are cast.
>
> [ ] +1 Release this package as Apache Marmotta 3.0.0-incubating
> [ ] 0 I don't feel strongly about it, but I'm okay with the release
> [ ] -1 Do not release this package because...
>