karaf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Achim Nierbeck <bcanh...@googlemail.com>
Subject Re: bnd files in Decanter Project
Date Thu, 11 Feb 2016 16:06:05 GMT
Hi there,

I'll stick to my

-1

Now the reasons for this:

Even though the opposite has been proclaimed on the list, it did break the
build for me. As without a bnd file the project isn't building
successfully, even though the information needed has been provided by the
pom.

Second the usage of an extra bnd file is error prone, and I don't care
about removed lines of code, this is just not a justification.
Another thing why I opposed to extra bnd files. Only on Eclipse when using
the bnd-tools project you have an extra editor available to use it. And
this only works for people using both, eclipse and the bnd-tools. All
others using a different tool chain are losing the benefit to see everthing
in one go.
So basically people using IntelliJ or Netbeans (both tools to be know to
work far better with maven, then eclipse) are out of this. And I rather
don't rely on a proprietary tooling for this.

So what's next after this? Just sneak the bnd-tools plugins into this
project for "supposedly" better support with eclipse and bnd-tools?
I don't think so. Our basis is still Maven and not Gradle, and this is
because we have a big user-base that feels very comfortable with it.
Using the declarations in the pom has been working nicely and if nothing
needs to be done one just needs to add those 4 lines for the usage of the
maven-bundle-plugin. No extra file needed, everything is in one place and
is in line with all other Karaf projects.
Actually this is one of the reasons I removed those bnd files from the
pax-web project long time ago, because all those project related
information are in one place now, instead of scattered around the project.

Third, and this does outweigh everything else. Christian tried to provide
evidence with this move and now argues this way everything is better for
us.
At this point I can't accept this. I don't think this is the way our
community works and surely this isn't the way we work together on our
sources. At least
this has been my understanding of a successful project.
If I'm wrong on this and everybody can do as he likes, I guess I need think
about the consequences.

regards, Achim




2016-02-11 15:12 GMT+01:00 Jean-Baptiste Onofré <jb@nanthrax.net>:

> I see valid arguments here, and I keep my +1.
>
> Regards
> JB
>
>
> On 02/11/2016 02:40 PM, Christian Schneider wrote:
>
>> I also did a jira issue that explains the change.
>> https://issues.apache.org/jira/browse/KARAF-4300
>>
>> To show the advantage let me compare one of the old configs to the new
>> one:
>> http://apaste.info/gC5
>>
>> This shows that 18 lines of xml are replaced by 2 lines in the bnd file.
>> The syntax is basically the same
>> as in xml just without the brackets. The big advantage is that we can
>> leave out the boilerplate xml that maven
>> needs to redefine a plugin.
>>
>> So I think the change to bnd files makes a lot of sense. I also did this
>> change in Aries a long time ago.
>> Btw. I also added API baselining in this change which helps us when do
>> changes on the API.
>>
>> Christian
>>
>>
>> On 11.02.2016 14:21, Christian Schneider wrote:
>>
>>> As further reference. This is the commit where I did the change:
>>>
>>> https://github.com/apache/karaf-decanter/commit/dabeeecd41f46a0c5344580f65c2ea6877fd6d35
>>>
>>>
>>> As you can see I added 108 lines and removed 753.
>>>
>>> That means the usage of bnd files saves us about 640 lines of xml. I
>>> think this is a strong indicator that it is a good idea to do so. Of
>>> course it might cause problems that I overlooked.
>>>
>>> Christian
>>>
>>> On 11.02.2016 13:57, Christian Schneider wrote:
>>>
>>>> Hi Achim,
>>>>
>>>> it is difficult to predict what changes warrant a discussion. I agree
>>>> with you that I should have discussed this on the list. I talked to
>>>> JB and he was positive so I did not expect any problems. Apparently I
>>>> was wrong :-(
>>>>
>>>> The usage of bnd files for configuration of the imports and exports
>>>> is a very concise replacement for the same configs in xml.
>>>>
>>>> The big advantage is that you can omit the maven-bundle-plugin in the
>>>> pom of each module. So basically you replace about 10-15 lines of xml
>>>> with 0-5 lines in the bnd file.
>>>>
>>>> This does not break any functionality for users. For developers it
>>>> just requires to put an empty bnd.bnd file into each module if it
>>>> does not need special configs. Unfortunately it is not possible to
>>>> define that it uses a bnd file if it is there  and is also happy if
>>>> no such file is there. I plan to suggest this to the felix project as
>>>> it would make this even easier to use.
>>>>
>>>> In what way do you see this as a breaking change? I made sure that
>>>> all code still works and that all tests still pass. I also did manual
>>>> tests of all the modules.
>>>>
>>>> So the only thing you need to do for a new module is to add this
>>>> empty bnd.bnd file and you are fine. If you do not want to use the
>>>> bnd file to configure your OSGi configs you can use the xml like before.
>>>>
>>>> So apart from my bad communication where I fully agree with you.. do
>>>> you really  think this warrants a -1?
>>>> Do you have any technical concerns?
>>>>
>>>> Christian
>>>>
>>>>
>>>> 2016-02-11 9:55 GMT+01:00 Achim Nierbeck <bcanhome@googlemail.com
>>>> <mailto:bcanhome@googlemail.com>>:
>>>>
>>>>     Hi,
>>>>
>>>>     the other day I added another module to the decanter project
>>>>     (cassandra
>>>>     appender).
>>>>     And I've got to say I was quite astonished to see all those bnd
>>>>     files in
>>>>     there, but what
>>>>     really got me stirred. It is mandatory to have those now.
>>>>
>>>>     I can't remember seeing a vote for such a change in development!
>>>>
>>>>     So here is my
>>>>
>>>>     -1
>>>>
>>>>     on this not communicated and breaking functionality change that
>>>>     sneaked in
>>>>     there.
>>>>
>>>>     So whoever changed that needs to revoke this, NOW.
>>>>     It hasn't been discussed up-front and actually I just can't stand
>>>>     such
>>>>     sneaky moves.
>>>>
>>>>     regards, Achim
>>>>
>>>>     --
>>>>
>>>>     Apache Member
>>>>     Apache Karaf <http://karaf.apache.org/> Committer & PMC
>>>>     OPS4J Pax Web <http://wiki.ops4j.org/display/paxweb/Pax+Web/>
>>>>     Committer &
>>>>     Project Lead
>>>>     blog <http://notizblog.nierbeck.de/>
>>>>     Co-Author of Apache Karaf Cookbook <http://bit.ly/1ps9rkS>
>>>>
>>>>     Software Architect / Project Manager / Scrum Master
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> --
>>>> Christian Schneider
>>>> http://www.liquid-reality.de
>>>> <
>>>> https://owa.talend.com/owa/redir.aspx?C=3aa4083e0c744ae1ba52bd062c5a7e46&URL=http%3a%2f%2fwww.liquid-reality.de
>>>> >
>>>>
>>>>
>>>> Open Source Architect
>>>> http://www.talend.com
>>>> <
>>>> https://owa.talend.com/owa/redir.aspx?C=3aa4083e0c744ae1ba52bd062c5a7e46&URL=http%3a%2f%2fwww.talend.com
>>>> >
>>>>
>>>>
>>>
>>> --
>>> Christian Schneider
>>> http://www.liquid-reality.de
>>>
>>> Open Source Architect
>>> http://www.talend.com
>>>
>>
>>
>>
> --
> Jean-Baptiste Onofré
> jbonofre@apache.org
> http://blog.nanthrax.net
> Talend - http://www.talend.com
>



-- 

Apache Member
Apache Karaf <http://karaf.apache.org/> Committer & PMC
OPS4J Pax Web <http://wiki.ops4j.org/display/paxweb/Pax+Web/> Committer &
Project Lead
blog <http://notizblog.nierbeck.de/>
Co-Author of Apache Karaf Cookbook <http://bit.ly/1ps9rkS>

Software Architect / Project Manager / Scrum Master

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message