Hi all,
given the recent merge and upcoming name change, I'd like to share a
draft of our board report (due May 13). If you have any concerns, let's
discuss them.
## Description:
SystemML is a declarative large-scale machine learning (ML) system that
allows the flexible specification of ML algorithms via an R-like syntax
and automatically generates hybrid runtime plans of single node,
in-memory operations as well as distributed operations on top of Apache
Spark. SystemDS extends this scope to the end-to-end data science
lifecycle (from data cleaning and preparation, over model training, to
debugging and deployment) as well as privacy-preserving federated learning.
## Issues:
After an internal discussion and positive feedback, the SystemDS fork of
Apache SystemML has been merged back into the Apache repository by
2020-03-27. This merge preserved the history of individual commits (430)
and respective contributors (16 of 17 new to SystemML).
The name change to Apache SystemDS has been approved
(PODLINGNAMESEARCH-179), and we intend (after this report has been
approved) to work with the INFRA team to change the Github repo, JIRA, etc.
Once the name change has been completed, we intend to release Apache
SystemDS 2.0 (to continue the Apache SystemML releases, but clearly
communicate the new scope).
## Membership Data:
- Apache SystemML was founded 2017-05-16 (incubator process entered
2015-11-02)
- There are currently 30 committers and 23 PMC members in the project.
- 4 new committers have been added 2020-05-01 (Arnab Phani, Mark Dokter,
Shafaq Siddiqi, and Kevin Innerebner).
## Activity and Health:
- Code activity is healthy with 149 commits in the last 3 months.
- Community growth is healthy with 4 new committers and 20 active
contributors in the last 3 months
- Communication is healthy but mailing list activity can be improved.
## Releases:
- Apache SystemML 1.2.0 was released on 2018-08-24.
- SystemDS 0.2.0 was released (outside ASF) on 2020-03-24
Regards,
Matthias
