Hi, The decision to drop the legacy Nutch search web apps was motivated by the following facts : 1 - most people tend to use SOLR anyway 2 - the number of active developers / committers is very limited which means that we couldn't maintain / improve the Nutch searcher and work on other aspects of Nutch 3 - SOLR has more functionality, a larger community - trying to catch up and match the functionalities in Nutch would have been a duplication of efforts and IMHO a waste of time
This is part of a larger attempt in Nutch to delegate to external components : Tika, SOLR and shortly crawler-commons. It makes sense to contribute to these projects and benefits from external contributions. We can then focus on more specific crawling-related issues and make Nutch better, the move towards 2.0 is a good example of this. This is a change in the scope of the project from being an open source large scale search engine to an open source crawler indeed. We should make this clearer on the website. While I agree that moving your search to SOLR represents an investment in the short term, you will benefit from the efforts from the SOLR/Lucene community, for instance the field collapsing in 3.3 is something that should be of interest to a lot of Nutch users. Hope it makes sense Julien On 6 July 2011 18:18, MilleBii <[email protected]> wrote: > I'm not too happy either that 1.3 forces to use Solr. > Which means I have to redevelop all my search web sites, didn't need that > extra work/burden. > > I will start the process to upgrade to 1.2 but 1.3 is a much bigger issue > for me. > > 2011/7/6 Markus Jelsma <[email protected]> > > > I'd still recommend to upgrade. Modifying Solr's example Velocity > template > > is > > really simple. You'll get a complete HTML driven search engine for free! > > > > On Wednesday 06 July 2011 18:57:26 [email protected] wrote: > > > So, I assume that, for now, I and people like me, should stick with > older > > > versions of Nutch, until something comes along that allows us to have > > just > > > a regular safety engine. > > > Thank you for your prompt reply. > > > > > > > Hi, > > > > > > > > With Solr's Velocity integration you can easily get an interface > > without > > > > a lot > > > > of work. It hasn't been tuned to work with Nutch' data yet. It's on > the > > > > todo > > > > list. > > > > > > > > Cheers, > > > > > > > > On Wednesday 06 July 2011 18:29:59 [email protected] wrote: > > > >> Ok, I could do the integration of Nutch-1.3 and Solr 3.3, but the > > > >> interface of Solr seems quite complicated to me. I just want to > have > > a > > > >> search engine with a customizable front and results pages, in full > > html > > > >> (not xml) like Nutch provided, until the last version. It is clear > to > > > >> me > > > >> that I have a lot do learn about Solr, or, may be, Nutch deviated > > from > > > >> the original purpose? I also noticed that the clustering-carrot2 is > > not > > > >> included and this one was very important to me. > > > >> Is there any tutorial that clarify those questions? > > > >> > > > >> Thank you. > > > > > > > > -- > > > > Markus Jelsma - CTO - Openindex > > > > http://www.linkedin.com/in/markus17 > > > > 050-8536620 / 06-50258350 > > > > -- > > Markus Jelsma - CTO - Openindex > > http://www.linkedin.com/in/markus17 > > 050-8536620 / 06-50258350 > > > > > > -- > -MilleBii- > -- * *Open Source Solutions for Text Engineering http://digitalpebble.blogspot.com/ http://www.digitalpebble.com
