Hi Jukka, > * The release signing guide [1] recommends to have also SHA1 checksums > for the release.
We'll try to incorporate this in the future. > * The approach to do an svn checkout as a part of the build is a bit > troublesome. The build will fail as soon as Lucene rearranges their > svn tree. I agree, but it seems the least bad of our choices. Since we are building a Lucene/Solr module, our alternatives are: (a) Include the entire Lucene/Solr tree in our package; (b) Check out a copy of lucene/solr and put our stuff in and build it; (c) Download a release source version of Lucene/Solr, unpack it, and then do the same as (b). (a) results in a massive package for little gain. (b) is the approach I finally took. And I didn't do (c) because for Lucene/Solr 4.x there is no package to download yet. > * Would it make sense to contribute this code directly to Solr instead > of having it in ManifoldCF? Especially since the code has no direct > ManifoldCF dependencies. Tried to do this for more than a year. I had 6 committers willing and happy to commit, but then another committer vetoed it just before that happened. I can name names if you like, but the reason was apparently that they want Lucene and Solr to become much more minimal than it is now. Since the politics and stars were not in alignment, I decided we'd have to make the best of it somehow. Karl On Thu, Dec 15, 2011 at 12:35 PM, Jukka Zitting <jukka.zitt...@gmail.com> wrote: > Hi, > > +1 from me too (binding also for the upcoming IPMC vote) > > I checked the apache-manifoldcf-solr-3.x-plugin-0.1-incubating-src.tar.gz > package with SHA1 checksum 14adbae8c05dc589a707208a172901cddd5c19d5. > > Some comments, none blocking: > > * The release signing guide [1] recommends to have also SHA1 checksums > for the release. > * The approach to do an svn checkout as a part of the build is a bit > troublesome. The build will fail as soon as Lucene rearranges their > svn tree. > * Would it make sense to contribute this code directly to Solr instead > of having it in ManifoldCF? Especially since the code has no direct > ManifoldCF dependencies. > > [1] http://www.apache.org/dev/release-signing.html > > BR, > > Jukka Zitting
