I wouldn't regard the seed support in Jaccard as a new feature, but instead as a bug fix. You aren't changing an existing API, you aren't adding a new API or functionality. Instead, you are making something that should work given the API return a correct value instead of an error. Based on my understanding of semver, that would indicate it's a patch, not a minor rev.
I haven't looked at the KS test stuff, but that sounds like actual new functionality, which would push this into the minor rev category instead of a patch. Will <http://www.verizonmedia.com> Will Lauer Senior Principal Architect, Audience & Advertising Reporting Data Platforms & Systems Engineering M 508 561 6427 1908 S. First St Champaign, IL 61822 <http://www.facebook.com/verizonmedia> <http://twitter.com/verizonmedia> <https://www.linkedin.com/company/verizon-media/> <http://www.instagram.com/verizonmedia> On Thu, May 27, 2021 at 12:11 PM Alexander Saydakov <[email protected]> wrote: > Looking at the history I see the following changes since 3.0.0 release: > - massive clean up to avoid compiler warnings > - iterator fix in REQ sketch > - iterator fix in KLL sketch > - seed support in Jaccard similarity (Theta and Tuple) > - Kolmogorov-Smirnov test for KLL sketches > - Theta and Tuple union fix > > KS test and seed support in Jaccard are new features, which might justify > a minor version release 3.1.0 > > > On Thu, May 27, 2021 at 9:46 AM Jon Malkin <[email protected]> wrote: > >> Hi all, >> >> A PR yesterday fixes a bug in the C++ theta sketch union: >> https://github.com/apache/datasketches-cpp/pull/218 >> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_apache_datasketches-2Dcpp_pull_218&d=DwMFaQ&c=sWW_bEwW_mLyN3Kx2v57Q8e-CRbmiT9yOhqES_g_wVY&r=0TpvE_u2hS1ubQhK3gLhy94YgZm2k_r8JHJnqgjOXx4&m=CvhQFP9I5tEG8iuFQ2IPKkURZxI90QYKdAW0Bib_H6M&s=j8TQzD4RzSKfmcQZMBguH1vYUB23HNkDnuDD35MNC8A&e=> >> >> The theta sketch is probably the most heavily-used sketch, so I suggest >> this justifies a new release. At least a patch release with that fix, but I >> think we've added a few things (jaccard similarity, seed support for more >> operations) that could justify a regular minor version release. >> >> Anyone second that? And if so, patch or bump the minor version? >> >> jon >> >
