Hi Richard, may I ask how far you got with uimaFIT v3 ? Does it already make sense to base new code on the branch you mentioned below or do you think it's still too unstable?
I am in the early stages of a text analysis project that might use UIMA, uimaFIT, and (possibly) DKPro. If a release of uimaFIT v3 is in sight, then I'd rather save myself a rewrite. On the other hand, I am a UIMA newbie; I might have a hard time to tell errors in my code apart from things that are broken due to the upgrade. Cheers! Christoph on 2018-10-02, Richard Eckart de Castilho wrote in in https://narkive.com/2yAjsNFi.2: > uimaFIT v3 is on my todo list for some time already. A few > weeks ago, I almost announced a vote for a release candidate, > but then Apache policy regarding checksums changed, mandating > a switch to SHA 256/512, so I dropped the RC again. > > We are now about to release a new version of the UIMA parent > POM which will produce SHA 512 checksums for our artifacts, > then I'll go back to running the uimaFIT v2 and v3 releases. > > There are no more significant changes scheduled for uimaFIT > 2.5.0 and 3.0.0 - it's basically just doing the release now. > > If you are brave enough to use unreleased version, you could > just check out the uimaFIT 3.0.x branch [1] temporarily and > build the SNAPSHOTs locally.
