For me it was also super amazing... getting first contributions especially in the beginning of October for an Open Source project, which was started at the university back in 2006 or 2007 I guess, but where I worked alone on the project since the end of 2012. But where I'm now more convinced than ever that the ideas borrowed especially from ZFS are great. Versioning done at the storage layer (completely written from scratch) using immutable keyed tries, log-structured and without the overhead of a persistent transaction-log, versioning even the data-pages fine granularly, using a distributed transaction log in the future for sharding...
So I'd love to move forward with the help of the Open Source community and to build a great analytics platform for analysing the past, doing audits, simply revert to a known good state because of human- or application-errors while even being able to stream real time changes :-) The versioning aspect introduces just a little bit of constant overhead, but in general unchanged structures are simply shared between revisions, even unchaned records of a data page.
For me it was also super amazing... getting first contributions especially in the beginning of October for an Open Source project, which was started at the university back in 2006 or 2007 I guess, but where I worked alone on the project since the end of 2012. But where I'm now more convinced than ever that the ideas borrowed especially from ZFS are great. Versioning done at the storage layer (completely written from scratch) using immutable keyed tries, log-structured and without the overhead of a persistent transaction-log, versioning even the data-pages fine granularly, using a distributed transaction log in the future for sharding...
So I'd love to move forward with the help of the Open Source community and to build a great analytics platform for analysing the past, doing audits, simply revert to a known good state because of human- or application-errors while even being able to stream real time changes :-) The versioning aspect introduces just a little bit of constant overhead, but in general unchanged structures are simply shared between revisions, even unchaned records of a data page.
Awesome to hear! All the best with the project as well! I am sure if you pull the word out about your project people would flock to it to help out.