master branch contains dangerous bugs about datajoint-matlab HOT 5 CLOSED

Yeah, I agree. I've been using future for ages and periodically we've added little things and fixed some issues. We should put out a new version soon.

from datajoint-matlab.

The only reason I haven't made the release is that I haven't tested issue #2 extensively: multiple datajoint schemas in the same database. Other than that, future has been stable for some time. It implements multiple important improvements and fixes. I think we can wait for ahoenselaar to finish adding the syntax for declaring additional indices and support for the Parallel Computing Toolbox. It looks like he is about to finish these features.

from datajoint-matlab.

In the future, bugs of such impact should trigger a bugfix release to master. We could use a major.minor.bugfix versioning scheme for releases. A [major.minor] release will remain feature- and API-stable but receive important bugfixes.
I am going to finish implementing secondary indexes and PCT support and put some effort into testing issue #2 .
Then we should be ready for 2.6 (if we adopt the new versioning scheme) or 2.5.2 in the old scheme.

from datajoint-matlab.

Agreed. Let's release 2.6 in the next few days and adopt better practices going forward since now we have a larger user base. Henceforth, let's keep master updated with latest bug fixes; future will be used for new features to be included in the upcoming release.

from datajoint-matlab.

just released version 2.6 (tag v2.6). I have tested most features, but there have been quite a few recent changes, so let's test it extensively.

from datajoint-matlab.