News for hackers. Stuff that matters.
Re:OptaPlanner – constraint solver for optimizing planning a... by antman on Monday October 12, 2020 @10:48PM (#24761758)
Looks interesting, this domain has a lot of players such as Google's coin-or so some quick description of its advantages or a benchmark would promote it better.
Re:OptaPlanner – constraint solver for optimizing planning a... by leetrout on Monday October 12, 2020 @11:36PM (#24762018)
One strong advantage is how well it integrates with Drools so if you’re a Java shop and happen to be using Drools you can just drop in optaplanner and have it optimize around facts.

Because it’s Java you can use it with all the various libraries and any compatible languages that support Java libs like kotlin and Scala.

Re:OptaPlanner – constraint solver for optimizing planning a... by victor106 on Monday October 12, 2020 @11:56PM (#24762114)
Drools is awesome and we use it in some of our projects. but it just feels very monolithic and not a cloud native solution. What are some good lightweight alternatives to drools?
Re:OptaPlanner – constraint solver for optimizing planning a... by ge0ffrey on Tuesday October 13, 2020 @07:07AM (#24764316)
For decision tables, definitly look at Kogito and Drools-DMN as alternatives.

To declare incremental constraints in OptaPlanner, take a look at ConstraintStreams if you want to avoid learning DRL: https://www.optaplanner.org/blog/2020/04/07/ConstraintStream... It's very similar to Java 8 Streams (but incremental and under the hood it uses Drools).

Re:OptaPlanner – constraint solver for optimizing planning a... by tappio on Tuesday October 13, 2020 @04:43AM (#24763552)
Kogito is a nice upgrade that we've now been experimenting with for some months https://kogito.kie.org/
Re:OptaPlanner – constraint solver for optimizing planning a... by evacchi on Tuesday October 13, 2020 @06:26AM (#24764097)
Kogito dev here if you folks have questions :)
Re:OptaPlanner – constraint solver for optimizing planning a... by victor106 on Tuesday October 13, 2020 @09:11AM (#24765229)
Thank you.

Is Kogito ready for production use? Does it have a UI that lets users develop their own rules?

Re:OptaPlanner – constraint solver for optimizing planning a... by evacchi on Tuesday October 13, 2020 @11:41AM (#24766637)
we are working hard to deliver a stable version. stay tuned :) but it is still currently under heavy development. If you are interested in visually design rules, you might want to look into DMN!
Re:OptaPlanner – constraint solver for optimizing planning a... by victor106 on Tuesday October 13, 2020 @03:56PM (#24769632)
OKAY, thanks. Does DMN integrate with Kogito? (maybe that's not a valid question?)
Re:OptaPlanner – constraint solver for optimizing planning a... by tappio on Wednesday October 14, 2020 @05:27AM (#24774757)
DMN is integral part of Kogito and they work really well. That was actually the main reason for us looking into that direction.
Re:OptaPlanner – constraint solver for optimizing planning a... by karreiro_ on Wednesday October 14, 2020 @11:21AM (#24777529)
Yeah, it does. If you want to learn more about it, check this: http://learn-dmn-in-15-minutes.com (it's a quick course I created to cover DMN on Kogito world)
Re:OptaPlanner – constraint solver for optimizing planning a... by victor106 on Wednesday October 14, 2020 @02:45PM (#24780078)
This is really cool. Will keep coming back to it and share it with users as we start using this more.
Re:OptaPlanner – constraint solver for optimizing planning a... by nurettin on Tuesday October 13, 2020 @01:30AM (#24762563)
Drools looks and sounds very good, but we already have a standard specification for workflow graphs called scxml and there are libraries which implement it and editors which can create it visually. I was hoping it would get more traction outside of Qt.
Re:OptaPlanner – constraint solver for optimizing planning a... by nazgul17 on Tuesday October 13, 2020 @01:23AM (#24762527)
Roll your own incremental score calculator ;)
Re:OptaPlanner – constraint solver for optimizing planning a... by 7thaccount on Tuesday October 13, 2020 @11:11AM (#24766325)
Not sure if you meant to have a comma in there or if you're incorrectly saying google made coin-or.

Coin-Or is a separate project. It's the fastest free/open source LP & MILP solver I've seen, but has a terrible CLI if you don't want to use it as a C++ library. It's easiest as part of the Open Solver Excel plugin. Lp_Solve isn't as fast I don't think, but the CLI is much easier to use and better documented. I prefer to use a scripting language language to write my own .lp models and send that directly to the solver. There are lots of libraries to do that, but it's easier for me to write code I fully understand than to try to learn someone's library.

Google has the "Or-tools" solver which I haven't used, but have friends which do use it and like it.

Re:OptaPlanner – constraint solver for optimizing planning a... by antman on Tuesday October 13, 2020 @12:38PM (#24767207)
Exactly, Thanks!
Re:OptaPlanner – constraint solver for optimizing planning a... by Frost1x on Tuesday October 13, 2020 @12:07AM (#24762181)
Is COIN-OR from Google? I'm pretty sure I've met some lead developers of the group in person and they weren't from Google IIRC.
Re:OptaPlanner – constraint solver for optimizing planning a... by quico on Tuesday October 13, 2020 @12:26AM (#24762263)
Coin-OR is not from Google. I suspect that OP confused Coin-OR [0] with Google's OR-tools [1].

[0]: https://www.coin-or.org/ [1]: https://developers.google.com/optimization/

Re:OptaPlanner – constraint solver for optimizing planning a... by eru on Tuesday October 13, 2020 @04:40AM (#24763538)
Google's OR-tools can use coin-or under the hood. So it's many people's main way to interact with coin-or.
Re:OptaPlanner – constraint solver for optimizing planning a... by antman on Tuesday October 13, 2020 @01:04PM (#24767486)
Exactly thanks, sorry cannot edit it now!