using the blog sample

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

using the blog sample

jamiecampbell
Valentin, Bartoz,

Thanks for your helpful comments.  I got things working (I think) and
wrote up a tutorial based on the adventure, at
http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample 
.. it's an unformatted plain text monstrosity, but, hopefully has some
useful information for folks to reference in future.

Part of what it includes is a shell command implementing bundle.  I got
kind of stuck on what namespace to use for it.  I didn't want to use
org.apache.aries.samples.blog because that would imply that it was
apache endorsed.  And I didn't want to use a namespace that was too
obscure to clearly connect to the fact that it DOES provide
functionality for that namespace.  So I went with
org.parit.aries.samples.blog.  If that namespace is still too close for
comfort, I can take another try at a different name, and am open to
suggestions :)

-Jamie
Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

Valentin Mahrwald
Very nice!

I followed it all the way through and except for a small race condition between the blog datasource service registration and the jpa extender creating the EntityManager everything worked like a charm. I am not quite sure what the right fix for the race condition is, I am not sure tweaking start orders would be necessarily any more reliable.

One thought would be that making the references to the sample files hyperlinks would make it a bit more consumable ;)

Thanks for posting this.

Valentin

On 29 Jun 2010, at 21:46, jamie campbell wrote:

> Valentin, Bartoz,
>
> Thanks for your helpful comments.  I got things working (I think) and wrote up a tutorial based on the adventure, at http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample .. it's an unformatted plain text monstrosity, but, hopefully has some useful information for folks to reference in future.
>
> Part of what it includes is a shell command implementing bundle.  I got kind of stuck on what namespace to use for it.  I didn't want to use org.apache.aries.samples.blog because that would imply that it was apache endorsed.  And I didn't want to use a namespace that was too obscure to clearly connect to the fact that it DOES provide functionality for that namespace.  So I went with org.parit.aries.samples.blog.  If that namespace is still too close for comfort, I can take another try at a different name, and am open to suggestions :)
>
> -Jamie

Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

Bartosz Kowalewski
Hi Jamie,

I followed your guide and I think it's great :).

Below are some minor comments:

1. You could base your Maven assembly descriptor on the ones that are
avilable in the Apache Felix project. This would eliminate the need to
change the existing sections of config.properties and
startup.properties. See:
http://svn.apache.org/viewvc/karaf/trunk/assembly/src/main/descriptors/unix-bin.xml?revision=958856&view=markup

2. Unfortunately, the released version of Apache Aries
(0.1-incubating) uses Derby DB with disk storage. This complicates
your quide a bit. The version of samples that is available in the
trunk uses in-memory DB. I think that it would make sense to update
your guide when a new version of Aries is released.

3. Warnings related to namespace handlers are ok. All shell commands
are initialized eventually by the Blueprint Extender.

4. Maybe you could also provide a package with a fully working
solution? It would be really useful for those wanting to take the
shortest path possible. :)

Best regards,
  Bartek

2010/7/4 Valentin Mahrwald <[hidden email]>:

> Very nice!
>
> I followed it all the way through and except for a small race condition between the blog datasource service registration and the jpa extender creating the EntityManager everything worked like a charm. I am not quite sure what the right fix for the race condition is, I am not sure tweaking start orders would be necessarily any more reliable.
>
> One thought would be that making the references to the sample files hyperlinks would make it a bit more consumable ;)
>
> Thanks for posting this.
>
> Valentin
>
> On 29 Jun 2010, at 21:46, jamie campbell wrote:
>
>> Valentin, Bartoz,
>>
>> Thanks for your helpful comments.  I got things working (I think) and wrote up a tutorial based on the adventure, at http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample .. it's an unformatted plain text monstrosity, but, hopefully has some useful information for folks to reference in future.
>>
>> Part of what it includes is a shell command implementing bundle.  I got kind of stuck on what namespace to use for it.  I didn't want to use org.apache.aries.samples.blog because that would imply that it was apache endorsed.  And I didn't want to use a namespace that was too obscure to clearly connect to the fact that it DOES provide functionality for that namespace.  So I went with org.parit.aries.samples.blog.  If that namespace is still too close for comfort, I can take another try at a different name, and am open to suggestions :)
>>
>> -Jamie
>
>
Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

Lin Sun
In reply to this post by jamiecampbell
Hi

Very nice documentation.  I was able to follow through.

One prob I encountered was the derby database setup.  I basically went
to http://incubator.apache.org/aries/blog-sample.html and followed the
instruction there.  I don't think it is intuitive for a first time
user to know where/how to run the ij script, in addition to setting
the DERBY_INSTALL variable.

Lin

On Tue, Jun 29, 2010 at 4:46 PM, jamie campbell <[hidden email]> wrote:

> Valentin, Bartoz,
>
> Thanks for your helpful comments.  I got things working (I think) and wrote
> up a tutorial based on the adventure, at
> http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample
> .. it's an unformatted plain text monstrosity, but, hopefully has some
> useful information for folks to reference in future.
>
> Part of what it includes is a shell command implementing bundle.  I got kind
> of stuck on what namespace to use for it.  I didn't want to use
> org.apache.aries.samples.blog because that would imply that it was apache
> endorsed.  And I didn't want to use a namespace that was too obscure to
> clearly connect to the fact that it DOES provide functionality for that
> namespace.  So I went with org.parit.aries.samples.blog.  If that namespace
> is still too close for comfort, I can take another try at a different name,
> and am open to suggestions :)
>
> -Jamie
>
Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

zoe slattery
Lin Sun wrote:

> Hi
>
> Very nice documentation.  I was able to follow through.
>
> One prob I encountered was the derby database setup.  I basically went
> to http://incubator.apache.org/aries/blog-sample.html and followed the
> instruction there.  I don't think it is intuitive for a first time
> user to know where/how to run the ij script, in addition to setting
> the DERBY_INSTALL variable.
>  
Hi Lin - just had another look at that page, what would you suggest to
make it clearer? I'm probably a bit to familiar with it to see what the
problem is:-/
As  Bartek mentioned, the sample in the development stream uses in
memory Derby so this step won't be necessary after the next release.

Z

> Lin
>
> On Tue, Jun 29, 2010 at 4:46 PM, jamie campbell <[hidden email]> wrote:
>  
>> Valentin, Bartoz,
>>
>> Thanks for your helpful comments.  I got things working (I think) and wrote
>> up a tutorial based on the adventure, at
>> http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample
>> .. it's an unformatted plain text monstrosity, but, hopefully has some
>> useful information for folks to reference in future.
>>
>> Part of what it includes is a shell command implementing bundle.  I got kind
>> of stuck on what namespace to use for it.  I didn't want to use
>> org.apache.aries.samples.blog because that would imply that it was apache
>> endorsed.  And I didn't want to use a namespace that was too obscure to
>> clearly connect to the fact that it DOES provide functionality for that
>> namespace.  So I went with org.parit.aries.samples.blog.  If that namespace
>> is still too close for comfort, I can take another try at a different name,
>> and am open to suggestions :)
>>
>> -Jamie
>>
>>    
>
>  

Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

Bartosz Kowalewski
Zoe,

I think that Lin wanted to say that she had been forced to take a look
at http://incubator.apache.org/aries/blog-sample.html when following
Jamie's guide :). ( => The
http://incubator.apache.org/aries/blog-sample.html guide is clear
enough.)

Right?

Bartek

2010/7/8 zoe slattery <[hidden email]>:

> Lin Sun wrote:
>>
>> Hi
>>
>> Very nice documentation.  I was able to follow through.
>>
>> One prob I encountered was the derby database setup.  I basically went
>> to http://incubator.apache.org/aries/blog-sample.html and followed the
>> instruction there.  I don't think it is intuitive for a first time
>> user to know where/how to run the ij script, in addition to setting
>> the DERBY_INSTALL variable.
>>
>
> Hi Lin - just had another look at that page, what would you suggest to make
> it clearer? I'm probably a bit to familiar with it to see what the problem
> is:-/
> As  Bartek mentioned, the sample in the development stream uses in memory
> Derby so this step won't be necessary after the next release.
>
> Z
>>
>> Lin
>>
>> On Tue, Jun 29, 2010 at 4:46 PM, jamie campbell <[hidden email]> wrote:
>>
>>>
>>> Valentin, Bartoz,
>>>
>>> Thanks for your helpful comments.  I got things working (I think) and
>>> wrote
>>> up a tutorial based on the adventure, at
>>>
>>> http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample
>>> .. it's an unformatted plain text monstrosity, but, hopefully has some
>>> useful information for folks to reference in future.
>>>
>>> Part of what it includes is a shell command implementing bundle.  I got
>>> kind
>>> of stuck on what namespace to use for it.  I didn't want to use
>>> org.apache.aries.samples.blog because that would imply that it was apache
>>> endorsed.  And I didn't want to use a namespace that was too obscure to
>>> clearly connect to the fact that it DOES provide functionality for that
>>> namespace.  So I went with org.parit.aries.samples.blog.  If that
>>> namespace
>>> is still too close for comfort, I can take another try at a different
>>> name,
>>> and am open to suggestions :)
>>>
>>> -Jamie
>>>
>>>
>>
>>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

zoe slattery
Bartek
D'oh - thanks :-)
Z

> Zoe,
>
> I think that Lin wanted to say that she had been forced to take a look
> at http://incubator.apache.org/aries/blog-sample.html when following
> Jamie's guide :). ( => The
> http://incubator.apache.org/aries/blog-sample.html guide is clear
> enough.)
>
> Right?
>
> Bartek
>
> 2010/7/8 zoe slattery <[hidden email]>:
>  
>> Lin Sun wrote:
>>    
>>> Hi
>>>
>>> Very nice documentation.  I was able to follow through.
>>>
>>> One prob I encountered was the derby database setup.  I basically went
>>> to http://incubator.apache.org/aries/blog-sample.html and followed the
>>> instruction there.  I don't think it is intuitive for a first time
>>> user to know where/how to run the ij script, in addition to setting
>>> the DERBY_INSTALL variable.
>>>
>>>      
>> Hi Lin - just had another look at that page, what would you suggest to make
>> it clearer? I'm probably a bit to familiar with it to see what the problem
>> is:-/
>> As  Bartek mentioned, the sample in the development stream uses in memory
>> Derby so this step won't be necessary after the next release.
>>
>> Z
>>    
>>> Lin
>>>
>>> On Tue, Jun 29, 2010 at 4:46 PM, jamie campbell <[hidden email]> wrote:
>>>
>>>      
>>>> Valentin, Bartoz,
>>>>
>>>> Thanks for your helpful comments.  I got things working (I think) and
>>>> wrote
>>>> up a tutorial based on the adventure, at
>>>>
>>>> http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample
>>>> .. it's an unformatted plain text monstrosity, but, hopefully has some
>>>> useful information for folks to reference in future.
>>>>
>>>> Part of what it includes is a shell command implementing bundle.  I got
>>>> kind
>>>> of stuck on what namespace to use for it.  I didn't want to use
>>>> org.apache.aries.samples.blog because that would imply that it was apache
>>>> endorsed.  And I didn't want to use a namespace that was too obscure to
>>>> clearly connect to the fact that it DOES provide functionality for that
>>>> namespace.  So I went with org.parit.aries.samples.blog.  If that
>>>> namespace
>>>> is still too close for comfort, I can take another try at a different
>>>> name,
>>>> and am open to suggestions :)
>>>>
>>>> -Jamie
>>>>
>>>>
>>>>        
>>>      
>>    
>
>  

Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

Lin Sun
In reply to this post by Bartosz Kowalewski
Hi Bartosz,

Yes, thanks for reading my mind :)   Zoe, your instruction is great!

Lin

On Thu, Jul 8, 2010 at 3:54 AM, Bartosz Kowalewski
<[hidden email]> wrote:

> Zoe,
>
> I think that Lin wanted to say that she had been forced to take a look
> at http://incubator.apache.org/aries/blog-sample.html when following
> Jamie's guide :). ( => The
> http://incubator.apache.org/aries/blog-sample.html guide is clear
> enough.)
>
> Right?
>
> Bartek
>
> 2010/7/8 zoe slattery <[hidden email]>:
>> Lin Sun wrote:
>>>
>>> Hi
>>>
>>> Very nice documentation.  I was able to follow through.
>>>
>>> One prob I encountered was the derby database setup.  I basically went
>>> to http://incubator.apache.org/aries/blog-sample.html and followed the
>>> instruction there.  I don't think it is intuitive for a first time
>>> user to know where/how to run the ij script, in addition to setting
>>> the DERBY_INSTALL variable.
>>>
>>
>> Hi Lin - just had another look at that page, what would you suggest to make
>> it clearer? I'm probably a bit to familiar with it to see what the problem
>> is:-/
>> As  Bartek mentioned, the sample in the development stream uses in memory
>> Derby so this step won't be necessary after the next release.
>>
>> Z
>>>
>>> Lin
>>>
>>> On Tue, Jun 29, 2010 at 4:46 PM, jamie campbell <[hidden email]> wrote:
>>>
>>>>
>>>> Valentin, Bartoz,
>>>>
>>>> Thanks for your helpful comments.  I got things working (I think) and
>>>> wrote
>>>> up a tutorial based on the adventure, at
>>>>
>>>> http://parit.ca/documentation/tutorials/apache/aries/openjpa-karaf-aries-derby-blog-sample
>>>> .. it's an unformatted plain text monstrosity, but, hopefully has some
>>>> useful information for folks to reference in future.
>>>>
>>>> Part of what it includes is a shell command implementing bundle.  I got
>>>> kind
>>>> of stuck on what namespace to use for it.  I didn't want to use
>>>> org.apache.aries.samples.blog because that would imply that it was apache
>>>> endorsed.  And I didn't want to use a namespace that was too obscure to
>>>> clearly connect to the fact that it DOES provide functionality for that
>>>> namespace.  So I went with org.parit.aries.samples.blog.  If that
>>>> namespace
>>>> is still too close for comfort, I can take another try at a different
>>>> name,
>>>> and am open to suggestions :)
>>>>
>>>> -Jamie
>>>>
>>>>
>>>
>>>
>>
>>
>
Reply | Threaded
Open this post in threaded view
|

Re: using the blog sample

jamiecampbell
In reply to this post by Valentin Mahrwald
Valentin: Hyperlinkification complete

Bartek:
    1) I had a peek at it.  The only unfortunate thing is that it sort
of trades complexities.  For the purposes of the tutorial I'd prefer to
have a simple 4 lines in the dependencySets section of the assembly
descriptor, rather than a section 155 lines long.  I think the startup
and config properties files are easier to visually compare for the
reader, who will say "uh huh, different slash conventions but otherwise
identical".

    2) Okie doke, will do

    3) Updated the guide to reflect this

    4) Done, mostly : I  didn't provide a package with all the
dependencies pre-grabbed, because the site I'm serving it from would
prefer not to have 32 meg grabs, but I did provide a package (2 meg
size) where all the user has to do is mvn assembly:assembly, unzip
archive, then run

Lin: Guide now has a link to the derby install/configure tutorial page

-Jamie