You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@uima.apache.org by Peter Klügl <pe...@averbis.com> on 2016/09/09 08:31:37 UTC
Additional material for Ruta
Hi,
I am currently wondering about the best way to provide additional
material for Ruta. Normally, I would say that it is hosted, e.g., in the
example-projects sub folder in the ruta reactor, but Apache
license/principles make it sometime harder than it has to be.
There are two current cases:
1. I like to create a new example how Ruta is able to interact with
DKPro Core in a well set-up maven project, e.g., in a new
german-novel-ner example. However, I do not want to restrict the
dependencies just so that the hosting at Apache is not problematic. Also
the example data is not compatible with the ASL 2.0.
2. Last year, I spent a lot of time thinking about the best way people
can learn Ruta. This resulted (this year) in a set of units with
exercises that we use in-house to train my colleagues. After some months
now, I am pleased with the learning effect and convinced that this is
the correct way to learn Ruta. It looks like I will get the OK of my
boses to contribute the material to the comunity. The training material
consists of a docbook (-> pdf) with the description of the exercises, an
Eclipse workspace containing simple Ruta projects for most exercises
where some initial stuff is already given, an Eclipse workspace with
solutions for most exercises, and an Eclipse installation with the Ruta
Workbench installed. The docbook would not be a problem to host here,
but I don't know if the workspaces are that suitable. (And one unit
contains data that is not compatible with the ASL 2.0).
My current guess is to create a github project containing the project
and workspaces with separate notice/license. I would add the docbook to
the ruta trunk and add links to the github project (problematic?). But
what about the Eclsipe installation?
I am not yet convinced.
Advice and opinions are very welcome.
Best,
Peter
Re: Additional material for Ruta
Posted by Jens Grivolla <j+...@grivolla.net>.
Hi Peter, did you ever get around to publishing this? I don't find it
referenced on https://uima.apache.org/ruta.html
Best,
Jens
On Fri, Sep 9, 2016 at 10:31 AM, Peter Klügl <pe...@averbis.com>
wrote:
> Hi,
>
>
> I am currently wondering about the best way to provide additional
> material for Ruta. Normally, I would say that it is hosted, e.g., in the
> example-projects sub folder in the ruta reactor, but Apache
> license/principles make it sometime harder than it has to be.
>
>
> There are two current cases:
>
> 1. I like to create a new example how Ruta is able to interact with
> DKPro Core in a well set-up maven project, e.g., in a new
> german-novel-ner example. However, I do not want to restrict the
> dependencies just so that the hosting at Apache is not problematic. Also
> the example data is not compatible with the ASL 2.0.
>
>
> 2. Last year, I spent a lot of time thinking about the best way people
> can learn Ruta. This resulted (this year) in a set of units with
> exercises that we use in-house to train my colleagues. After some months
> now, I am pleased with the learning effect and convinced that this is
> the correct way to learn Ruta. It looks like I will get the OK of my
> boses to contribute the material to the comunity. The training material
> consists of a docbook (-> pdf) with the description of the exercises, an
> Eclipse workspace containing simple Ruta projects for most exercises
> where some initial stuff is already given, an Eclipse workspace with
> solutions for most exercises, and an Eclipse installation with the Ruta
> Workbench installed. The docbook would not be a problem to host here,
> but I don't know if the workspaces are that suitable. (And one unit
> contains data that is not compatible with the ASL 2.0).
>
>
> My current guess is to create a github project containing the project
> and workspaces with separate notice/license. I would add the docbook to
> the ruta trunk and add links to the github project (problematic?). But
> what about the Eclsipe installation?
>
>
> I am not yet convinced.
>
>
> Advice and opinions are very welcome.
>
>
>
> Best,
>
>
> Peter
>
>
>
Re: Additional material for Ruta
Posted by Richard Eckart de Castilho <re...@apache.org>.
On 09.09.2016, at 18:01, Marshall Schor <ms...@schor.com> wrote:
>
> I recall the Apache policy: you can have dependencies on artifacts with
> non-approved licenses, but only if they are "optional" dependencies.
I'd argue that examples are entirely optional because they do not form
an essential part of the functionality of the product.
Cheers,
-- Richard
Re: Additional material for Ruta
Posted by Marshall Schor <ms...@schor.com>.
I recall the Apache policy: you can have dependencies on artifacts with
non-approved licenses, but only if they are "optional" dependencies.
-Marshall
On 9/9/2016 10:48 AM, Peter Kl�gl wrote:
> Hi,
>
>
> Am 09.09.2016 um 16:39 schrieb Richard Eckart de Castilho:
>> I don't think it is a problem to have dependencies on artifacts that
>> are not following the Apache policy as long as no release artifact
>> is created that includes these dependencies. So as long as you don't plan
>> to release the examples along with all their dependencies, you should
>> be fine.
> Hmm, you are right, I can add a dependency to Stanford CoreNLP GPL, but
> I do not have a good feeling about that.
>
>> Regarding data, you could download is automatically as part of the example
>> instead of hosting it on the Apache infrastructure. Should also solve the
>> problem.
> It is not that easy, e.g., there is no downloadable archive or it is
> annotated in XMIs.
>
>> Regarding DocBook... you might find it interesting to have a look at
>> AsciiDoctor. Eclipse Neon has some preview support for it, there is
>> a nice maven plugin to render it to HTML and PDF as part of a build,
>> and it is *much* less verbose then DocBook. Personally, I started
>> liking to write documentation again after abandoning DocBook for Asciidoctor.
>> There is also a nice feature in it that allows to pull parts of sources
>> into the documentation. So instead of repeating example code, you can
>> pull it out of the actual example code into the documentation. And if
>> you have unit tests for your example code, then you can make quite sure
>> that your documentation examples are actually working.
> I am used to DocBook and there is no reason for me to switch since the
> text is already written.
>
> You can inlude code in Docbook. The documentation of our analysis
> engines that are implemented in Ruta contains also the actual rules
> (which are automatically included).
>
> Peter
>
>
>> Cheers,
>>
>> -- Richard
>>
>>> On 09.09.2016, at 10:31, Peter Kl�gl <pe...@averbis.com> wrote:
>>>
>>> Hi,
>>>
>>>
>>> I am currently wondering about the best way to provide additional
>>> material for Ruta. Normally, I would say that it is hosted, e.g., in the
>>> example-projects sub folder in the ruta reactor, but Apache
>>> license/principles make it sometime harder than it has to be.
>>>
>>>
>>> There are two current cases:
>>>
>>> 1. I like to create a new example how Ruta is able to interact with
>>> DKPro Core in a well set-up maven project, e.g., in a new
>>> german-novel-ner example. However, I do not want to restrict the
>>> dependencies just so that the hosting at Apache is not problematic. Also
>>> the example data is not compatible with the ASL 2.0.
>>>
>>>
>>> 2. Last year, I spent a lot of time thinking about the best way people
>>> can learn Ruta. This resulted (this year) in a set of units with
>>> exercises that we use in-house to train my colleagues. After some months
>>> now, I am pleased with the learning effect and convinced that this is
>>> the correct way to learn Ruta. It looks like I will get the OK of my
>>> boses to contribute the material to the comunity. The training material
>>> consists of a docbook (-> pdf) with the description of the exercises, an
>>> Eclipse workspace containing simple Ruta projects for most exercises
>>> where some initial stuff is already given, an Eclipse workspace with
>>> solutions for most exercises, and an Eclipse installation with the Ruta
>>> Workbench installed. The docbook would not be a problem to host here,
>>> but I don't know if the workspaces are that suitable. (And one unit
>>> contains data that is not compatible with the ASL 2.0).
>>>
>>> My current guess is to create a github project containing the project
>>> and workspaces with separate notice/license. I would add the docbook to
>>> the ruta trunk and add links to the github project (problematic?). But
>>> what about the Eclsipe installation?
>>>
>>> I am not yet convinced.
>>>
>>> Advice and opinions are very welcome.
>>>
>>> Best,
>>>
>>> Peter
>
Re: Additional material for Ruta
Posted by Peter Klügl <pe...@averbis.com>.
Hi,
Am 09.09.2016 um 16:39 schrieb Richard Eckart de Castilho:
> I don't think it is a problem to have dependencies on artifacts that
> are not following the Apache policy as long as no release artifact
> is created that includes these dependencies. So as long as you don't plan
> to release the examples along with all their dependencies, you should
> be fine.
Hmm, you are right, I can add a dependency to Stanford CoreNLP GPL, but
I do not have a good feeling about that.
> Regarding data, you could download is automatically as part of the example
> instead of hosting it on the Apache infrastructure. Should also solve the
> problem.
It is not that easy, e.g., there is no downloadable archive or it is
annotated in XMIs.
>
> Regarding DocBook... you might find it interesting to have a look at
> AsciiDoctor. Eclipse Neon has some preview support for it, there is
> a nice maven plugin to render it to HTML and PDF as part of a build,
> and it is *much* less verbose then DocBook. Personally, I started
> liking to write documentation again after abandoning DocBook for Asciidoctor.
> There is also a nice feature in it that allows to pull parts of sources
> into the documentation. So instead of repeating example code, you can
> pull it out of the actual example code into the documentation. And if
> you have unit tests for your example code, then you can make quite sure
> that your documentation examples are actually working.
I am used to DocBook and there is no reason for me to switch since the
text is already written.
You can inlude code in Docbook. The documentation of our analysis
engines that are implemented in Ruta contains also the actual rules
(which are automatically included).
Peter
> Cheers,
>
> -- Richard
>
>> On 09.09.2016, at 10:31, Peter Kl�gl <pe...@averbis.com> wrote:
>>
>> Hi,
>>
>>
>> I am currently wondering about the best way to provide additional
>> material for Ruta. Normally, I would say that it is hosted, e.g., in the
>> example-projects sub folder in the ruta reactor, but Apache
>> license/principles make it sometime harder than it has to be.
>>
>>
>> There are two current cases:
>>
>> 1. I like to create a new example how Ruta is able to interact with
>> DKPro Core in a well set-up maven project, e.g., in a new
>> german-novel-ner example. However, I do not want to restrict the
>> dependencies just so that the hosting at Apache is not problematic. Also
>> the example data is not compatible with the ASL 2.0.
>>
>>
>> 2. Last year, I spent a lot of time thinking about the best way people
>> can learn Ruta. This resulted (this year) in a set of units with
>> exercises that we use in-house to train my colleagues. After some months
>> now, I am pleased with the learning effect and convinced that this is
>> the correct way to learn Ruta. It looks like I will get the OK of my
>> boses to contribute the material to the comunity. The training material
>> consists of a docbook (-> pdf) with the description of the exercises, an
>> Eclipse workspace containing simple Ruta projects for most exercises
>> where some initial stuff is already given, an Eclipse workspace with
>> solutions for most exercises, and an Eclipse installation with the Ruta
>> Workbench installed. The docbook would not be a problem to host here,
>> but I don't know if the workspaces are that suitable. (And one unit
>> contains data that is not compatible with the ASL 2.0).
>>
>> My current guess is to create a github project containing the project
>> and workspaces with separate notice/license. I would add the docbook to
>> the ruta trunk and add links to the github project (problematic?). But
>> what about the Eclsipe installation?
>>
>> I am not yet convinced.
>>
>> Advice and opinions are very welcome.
>>
>> Best,
>>
>> Peter
Re: Additional material for Ruta
Posted by Richard Eckart de Castilho <re...@apache.org>.
I don't think it is a problem to have dependencies on artifacts that
are not following the Apache policy as long as no release artifact
is created that includes these dependencies. So as long as you don't plan
to release the examples along with all their dependencies, you should
be fine.
Regarding data, you could download is automatically as part of the example
instead of hosting it on the Apache infrastructure. Should also solve the
problem.
Regarding DocBook... you might find it interesting to have a look at
AsciiDoctor. Eclipse Neon has some preview support for it, there is
a nice maven plugin to render it to HTML and PDF as part of a build,
and it is *much* less verbose then DocBook. Personally, I started
liking to write documentation again after abandoning DocBook for Asciidoctor.
There is also a nice feature in it that allows to pull parts of sources
into the documentation. So instead of repeating example code, you can
pull it out of the actual example code into the documentation. And if
you have unit tests for your example code, then you can make quite sure
that your documentation examples are actually working.
Cheers,
-- Richard
> On 09.09.2016, at 10:31, Peter Klügl <pe...@averbis.com> wrote:
>
> Hi,
>
>
> I am currently wondering about the best way to provide additional
> material for Ruta. Normally, I would say that it is hosted, e.g., in the
> example-projects sub folder in the ruta reactor, but Apache
> license/principles make it sometime harder than it has to be.
>
>
> There are two current cases:
>
> 1. I like to create a new example how Ruta is able to interact with
> DKPro Core in a well set-up maven project, e.g., in a new
> german-novel-ner example. However, I do not want to restrict the
> dependencies just so that the hosting at Apache is not problematic. Also
> the example data is not compatible with the ASL 2.0.
>
>
> 2. Last year, I spent a lot of time thinking about the best way people
> can learn Ruta. This resulted (this year) in a set of units with
> exercises that we use in-house to train my colleagues. After some months
> now, I am pleased with the learning effect and convinced that this is
> the correct way to learn Ruta. It looks like I will get the OK of my
> boses to contribute the material to the comunity. The training material
> consists of a docbook (-> pdf) with the description of the exercises, an
> Eclipse workspace containing simple Ruta projects for most exercises
> where some initial stuff is already given, an Eclipse workspace with
> solutions for most exercises, and an Eclipse installation with the Ruta
> Workbench installed. The docbook would not be a problem to host here,
> but I don't know if the workspaces are that suitable. (And one unit
> contains data that is not compatible with the ASL 2.0).
>
> My current guess is to create a github project containing the project
> and workspaces with separate notice/license. I would add the docbook to
> the ruta trunk and add links to the github project (problematic?). But
> what about the Eclsipe installation?
>
> I am not yet convinced.
>
> Advice and opinions are very welcome.
>
> Best,
>
> Peter