Re: [RESULT][APE] Proposal for Enhancing Geometry Processing in AsterixDB with JTS instead of ESRI Library

[Suryaa Charan Shivakumar <ss...@ucr.edu.INVALID>]

Hello AsterixDB Dev Community,

The vote to accept the proposal for migrating from ESRI's geometry library
to JTS has passed with
3 binding votes from
- Mike Carey
- Ian Maxon
- Till Westmann

Thanks for voting!

Vote thread:
https://lists.apache.org/thread/mchlz7xhlqjtm2xhd9xm1qk4sykq1l7p

Best,
Suryaa (on behalf of the AsterixDB PMC)

On Sun, Aug 4, 2024 at 10:59 AM Mike Carey <dt...@gmail.com> wrote:

> Agreed, nice proposal!  +1
>
> (Be sure to including testing with both row and columnnar formats.)
>
> On 7/31/24 3:18 PM, Till Westmann wrote:
> > Nice proposal: +1
> >
> > On 2024/07/30 00:59:46 Ian Maxon wrote:
> >> Sounds good to me. +1
> >>
> >> On Jul 29, 2024 at 16:12:09, Suryaa Charan Shivakumar<ss...@ucr.edu>
> >> wrote:
> >>
> >>> Hello AsterixDB Dev Community,
> >>>
> >>> I hope this message finds you well. I am excited to share a proposal
> aimed
> >>> at advancing our geometric processing capabilities in AsterixDB. The
> >>> proposal, identified with JIRA epic *ASTERIXDB-3412*, is targeted for
> the
> >>> 9.10.0 release and involves transitioning from the current
> >>> com.esri.core.geometry.ogc.OGCGeometry to the Java Topology Suite (JTS)
> >>> library.
> >>>
> >>> Key Motivations:
> >>>
> >>>     1. Enhanced Compatibility & Maintainability: Transitioning to JTS
> >>>     offers a more robust support for geometric operations, allowing
> >>>     extensibility and custom implementation over spatial functions.
> >>>     2. Potential Performance Improvements: JTS is expected to provide
> not
> >>>     only compatibility but also performance enhancements over the
> existing
> >>>     geometry library.
> >>>
> >>> Proposed Changes Include:
> >>>
> >>>     1.  Comprehensive Research & Documentation of current OGCGeometry
> use
> >>>     and dependencies.
> >>>     2.  Thorough Evaluation of JTS Library to ensure it meets our needs
> >>>     and integrates well with the existing codebase.
> >>>     3. Implementation and Testing across the codebase to ensure
> seamless
> >>>     functionality and performance.
> >>>     4. Updates and Revisions in Documentation to reflect the new
> changes.
> >>>     5. Performance Benchmarking to confirm and enhance the benefits of
> JTS
> >>>     integration.
> >>>
> >>> Known Limitations and Considerations:
> >>>
> >>>     1. CRS and Dimensionality Support: Address JTS’s native
> limitations on
> >>>     CRS and higher-dimensional (3d,4d) geometries through custom
> >>>     implementations or additional libraries.
> >>>        1. JTS does not consider or support CRS while doing geometric
> >>>        operations. However, CRS is important to ensure precision and
> correctness.
> >>>        We may have to evaluate the integration of CRS functionalities
> through
> >>>        libraries like GeoTools, minimizing dependencies while ensuring
> robust
> >>>        support for varied CRS applications.
> >>>        2. Issues with JTS while handling 3D or 4D geometry,
> >>>        Does not have some function implementations for Z, eg. Finding
> min
> >>>        z or max z coordinate in a given geometry. Solution - To
> implement this is
> >>>        fairly easy. Other known documented issues in JTS w.r.t
> dimensions are
> >>>        listed below,
> >>>        https://github.com/locationtech/jts/issues/733
> >>>        <
> https://urldefense.com/v3/__https://github.com/locationtech/jts/issues/733__;!!CzAuKJ42GuquVTTmVmPViYEvSg!PiGdnbM37SsJM5MdFI4o05dsJmQQjNbeKfB04azc7-SO85bhaS0Vj3T6RwCeRXOYOajJFvWWw1duH8tT$
> >
> >>>        https://locationtech.github.io/jts/jts-faq.html#B4
> >>>        <
> https://urldefense.com/v3/__https://locationtech.github.io/jts/jts-faq.html*B4__;Iw!!CzAuKJ42GuquVTTmVmPViYEvSg!PiGdnbM37SsJM5MdFI4o05dsJmQQjNbeKfB04azc7-SO85bhaS0Vj3T6RwCeRXOYOajJFvWWwx1gF13X$
> >
> >>>        2.  Dependency Management: Carefully manage new dependencies to
> >>>     prevent over-complexity in the system architecture.
> >>>
> >>> We invite you to review the detailed proposal, provide your insights,
> and
> >>> discuss any potential impacts or improvements. Your feedback is
> invaluable
> >>> as we refine and move forward with this initiative.
> >>>
> >>> Thank you for your continued support and collaboration.
> >>>
> >>> Best,
> >>> Suryaa Charan
> >>>