Dear new nutch contributors,
I'm very demotivated to contribute a jmx backbone to nutch since I have the feeling nutch contributors aren't ready for that yet.
I'm sorry to say that, let me explain why I think so.
First of all I collect a set of bad experience until last 10 Months. I takes around 2 months to implement the plugin system but it takes another 3 months to discuss it. In the end I missed to have time for writing the 10 lines ant script and nobody else was able to do it. Today more or less the first version of plugin code is in the cvs - still with dom4j and set of bugs that I had already fixed with a patch that wasn't committed to the cvs. (I will contribute it again, since there was conflicts with some work from Doug)
Please remember that there was already a PDF, Word, Flash, RTF, Open office etc. content extractor (today we call it parser again ) plugins that was working well and most important they comes with junit tests.
To me it looks like contributors start to understand the power of the plugin system, just count the contributions that comes since it is official available.
I have a bad taste in my mouth since all new contributions are new functionality to nutch and comes without any junit test.
Just a critical question, why we write " All Java code modules should be accompanied by JUnit tests." if we just ignore that,
sorry to ask but why is code formatting more important then junit tests?
The jmx stuff is to much work for just discussing it after implementing it or just ignore it as other patches I had done (topic filtering, snmp, host grouping ... *:-)*).
Using Jmx will dramatically change the complete architecture of nutch but it will be a bigger improvement to nutch architecture then the plugin system.
If you don't trust me just ask the Heritrix Guys for experience they use jmx as well and it looks like they invest a lot of time for a clean architecture design.
We haven't tools anymore we will have a server that runs 24/7 and a set of components that can be triggered.
There was just one comment by Doug about my jmx mail and nobody else give any technical comment. That strongly remembers me to the time I was sending my first concept about the plugin system to this list.
Don't get me wrong I would love to contribute but there are two targets we can follow, focus to get stuff we personal need just running or write a strong extensible enterprise software.
The second target of course needs more work.
From my point of view a good architecture is as much important as a good search result.
We will get much more users with a out of the box solution, then with a application that's administration is a full-time job.
So I (as a user and patch contributor ) would be happy to see unit tests from the new contributors and I would be happy to start some serious discussions about architecture issues of nutch.
Best Regards, Stefan
--------------------------------------------------------------- enterprise information technology consulting open technology: http://www.media-style.com open source: http://www.weta-group.net open discussion: http://www.text-mining.org
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Nutch-developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/nutch-developers
