Just trying to understand why we need to have hadoop, nutch, tika and mahout releases lying aorund when nobody will come to Lucene for them anymore, and I doubt anyone would prefer to use those versions.
So no pressing “need”, just general housekeeping. Both Hadoop and Mahout projects have copied over all releases to their respective release folders and lucene/hadoop folder is completely gone from the mirros, although it still remains in the archive. For Tika, I see that you link to the lucene archives so people will actually know that these releases exist :) Worse so for Nutch that have neither a complete archive themselves nor link to the lucene archives. I can see the argument that there may still be some automated scripts pulling down 0.7 somewhere. Do we have download stats per file available somewhere? -- Jan Høydahl, search solution architect Cominvent AS - www.cominvent.com > 16. feb. 2017 kl. 00.30 skrev Chris Mattmann <mattm...@apache.org>: > > Hi Jan, > > Thanks for the message, but frankly I don’t think we should do this. The > releases > have been made and should be canonical, even if they were made from Lucene > since > Tika originated there. Many sites over many years have likely mirror’ed > thoese URLs and > I think this could screw that up. Is there some driving motivation to do > this? For example, > the releases are likely small-ish, and don’t take a lot of space. > > Just trying to understand. > > Cheers, > Chris > > > > > On 2/15/17, 3:28 PM, "Jan Høydahl" <jan....@cominvent.com> wrote: > > Hi, > > The Lucene release archive still hosts Tika releases v0.2-0.7, se > https://archive.apache.org/dist/lucene/tika/. These do not exist in > https://archive.apache.org/dist/tika/ > > We’d prefer if you would copy the missing pieces over to Tika’s release > area and we’ll then remove it from Lucene. > > For reference, see https://issues.apache.org/jira/browse/LUCENE-7696 > > -- > Jan Høydahl, search solution architect > Cominvent AS - www.cominvent.com > > > >
