Releases: TimefoldAI/timefold-solver
Timefold Solver Community Edition 1.13.0
As the summer sun is slowly baking us, so have we baked something new for you. Timefold Solver 1.13.0 brings new features as well as the usual assortment of bug fixes and improvements.
Featured Update: Constraint Weight Overrides
We bring you a smaller an improved API for setting weights for your constraints. Where previously you had to write a lot of boilerplate to provide an entire ConstraintConfiguration implementation, specifying a constraint weight is now as easy as adding a Map to your planning solution!
Besides that, we are introducing a new type of shadow variable - one that can be updated by a simple method, as opposed to introducing an entire VariableListener. This allows for easy updates to the tail chains of a list variable.
As always, all of these updates are available for all of our Java, Python and Kotlin users.
Changelog
🚀 Features
- 10b9258 Add CascadingUpdateShadowVariable (#954), closes #954
- 2c74a11 Constraint weight override (#946), closes #946
🐛 Fixes
- 3ffb49b Address scale equal to NaN
- ec49da1 Respect merge function in toMap() collector
- d01cd42 Avoid duplicating nearby configuration
- ecae569 Proper casts in Python load_balance collector
🔄️ Changes
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Jurriaan Persyn (@oemebamo)
- Lukáš Petrovický (@triceo)
- Marek Winkler (@winklerm)
- hizhangbo (@hizhangbo)
- lee-carlon
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts
and read our documentation.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver 1.12.0
Summer is here, but we're not resting! Not all of us anyway. This release brings features, as well as an assortment of bug fixes and enhancements.
Featured update: Fairness and Load Balancing
Providing optimized schedules is only one part of the problem. And if our employees see that our schedule isn't being fair to them, because Ann is getting three times as many shifts as Dan is, they will not be very happy and the business will struggle. The business may even run into labour law issues.
That is why we are introducing a load balancing collector for Constraint Streams. You can read all about it in Fairness and load balancing section of the documentation.
Besides that, we have added some quality of life improvements:
ScoreAnalysisnow provides more information about itself, including a useful textual summary of the analysis.- The solver can now trigger an event when the first initialized solution shows up, letting you know when the construction heuristic ends and local search begins.
All of these improvements and bugfixes are also available in Timefold Solver for Python. Have you checked it out yet?
Changelog
🚀 Features
- d07cfb2 Add support for fairness constraints (#918), closes #918
- 7e6102e Improve
ScoreAnalysisdebug information (#923), closes #923 - 84e4ae0 Allow to consume the initialized solution (#915), closes #915
🐛 Fixes
- e7fae26 Only include tail chain move when using a single var (#936), closes #936
- d5ed7e1 Do not fail on solution/entity classes as interfaces (#934), closes #934
- 24b3240 Continue the warmup when one configuration fails
- 03ed42c Update the best score image of the benchmark section
- 5690ec6 Fail fast when local search starts from an uninitialized solution
Contributors
We'd like to thank the following people for their contributions:
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.12.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.12.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver 1.11.0
Nearly a year ago, we released Timefold Solver 1.0 for Java, the first step on our journey to rid the world of wasteful planning. Today, the journey continues as we announce Timefold Solver for Python!
In this first Beta release, we give the Python ecosystem access to Timefold Solver and all of its features. We invite you to try out our new Python quickstarts and let us know what you think. Do you like the APIs? Does it integrate well with your Python projects? How's the performance? We need your feedback!
Timefold Solver for Python will move out of Beta once we've heard back from enough of you. While in Beta, we reserve the right to change APIs and to add or remove functionality based on user feedback. Once we leave Beta, we will provide the same backwards compatibility guarantees as with the Java solver.
And speaking of the Java solver, we haven't forgotten about it either. In this release, we've been mostly fixing bugs and making small tweaks here and there. Going forward, we've got plenty of interesting things in store for the Java and Python solvers, as they will evolve together from now on.
Changelog
🐛 Fixes
- 30980b7 Make the nearbyDistanceMeterClass a visitable class for GIZMO
- 73bfaa6 Use custom thread factory class in SolverManager (#875), closes #875
- d7a5404 Disable nearby configuration for CH
🧰 Tasks
- 9c76f53 Update docs on which time and date types to use
📝 Documentation
- 62603f7 Fix a typo in toConsecutiveSequences example
- e558697 Fix typos in connected ranges
- 83abce6 Fix typos in collectors
- d64b634 Fix ConstraintCollectors.average() example
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Geoffrey De Smet
- Lukáš Petrovický (@triceo)
- Marek Winkler (@winklerm)
- plplmax (@plplmax)
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver in Java or Kotlin
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.11.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.11.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver 1.10.0
In the May '24 release of Timefold Solver, we’ve focused on making everything more stable and reliable. Here’s what’s new:
- Connected Ranges Collector: This new feature helps manage and penalize the use of resources over time.
- Legacy Examples to Quickstarts: We’ve turned old examples into quickstart guides for easier use.
- Throttling Best Solution Consumer: For our Enterprise Edition users, we’ve added a feature to prevent integrations from being overloaded by solver events.
- Maven Repository Authentication: We’re preparing to require authentication for our Enterprise Edition binaries. We’ll notify customers in advance to ensure a smooth transition. Community Edition users won’t be affected, as their binaries will remain freely available on Maven Central.
Changelog
🚀 Features
🐛 Fixes
- 6c4e2e6 Properly detect value ranges with wildcard types (#838), closes #838
- c373c2b Remove too eager fail-fast in concat node's retract (#831), closes #831
- fdeb06a Forbid pinning uninitialized entities when unassigned not allowed (#821), closes #821
- b5ff711 Make unassigned location the last element destination
🔄️ Changes
🧰 Tasks
- a649452 Improve score corruption error message (#822), closes #822
- a210d93 Remove
examplesmodule (#809), closes #809 - 93f5b58 Add support for authenticated Enterprise repo (#811), closes #811
📝 Documentation
- 5bfa796 Document the throttling feature of Enterprise Edition
- 4f19adb Correct property name for node sharing in Spring
- e7f9bb2 Fix list variable unassigned values documentation
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Lukáš Petrovický (@triceo)
- Wouter De Schuyter (@wouterds)
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.10.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.10.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver 1.9.0
Introducing Timefold Solver’s latest release: Broadening Horizons Across Industries.
In our pursuit to cater to a wider array of planning challenges, we have added new quickstarts in our latest release. Tailored to meet the dynamic needs of various industries and use-cases.
- Conference Scheduling: Streamline the complex logistics of organizing successful events.
- Flight Crew Scheduling: Optimize crew assignments for efficiency and compliance.
- Hospital Bed Allocation: Enhance patient care through smarter bed management.
- Meeting Scheduling: Simplify the coordination of meetings to boost productivity and collaboration.
These quickstarts are designed to jump-start your planning processes, offering a fast track to deploying sophisticated scheduling and allocation solutions.
Enterprise Edition
Our customers gain an edge with the introduction of Nearby Selection in Construction Heuristics. This powerful feature enhances the solver’s capability to generate better initial solutions by considering proximity in decision-making, thereby optimizing operational efficiency and resource utilization across your projects.
And as always, this release also includes significant bug fixes and quality of life improvements for all users.
Changelog
🚀 Features
- a99e600 Improve constraint verifier error messages (#758), closes #758
- 8344f4b Down-sampling process for Benchmarker (#714), closes #714
🐛 Fixes
- 3ab0480 Enable Nearby selection for Construction Heuristics (#767), closes #767
- c6f1e4f Benchmark duplicated metrics (#761), closes #163 #761
- a7e7a19 Unimproved terminations skip construction heuristic (#755), closes #755
- 493723e Add rebase support to NoChangeMove and ListUnassignMove
🧰 Tasks
📝 Documentation
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Geoffrey De Smet
- Lukáš Petrovický (@triceo)
- Radovan Synek
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.9.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.9.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver 1.8.1
This is a hotfix release for Timefold customers. Please update if all of the following applies to you:
- You use list variables with unassigned values,
- you use multi-threaded solving,
- and you use construction heuristics.
If any one of the above doesn't apply to you, you do not need to update to Timefold Solver 1.8.1 and can wait for the regularly scheduled release next month.
Changelog
🐛 Fixes
- 43e52bd add rebase support to NoChangeMove and ListUnassignMove
Timefold Solver Community Edition 1.8.0
Spring is in the air, and so is another release of Timefold Solver. And this time, it's a big one! For the Community Edition, we have prepared the following features:
- List variables now allow for unassigned values. Say goodbye to the "dummy vehicle" pattern, say hello to the performance improvements that come from it!
- Spring Boot users among you can now easily generate native images, as was already possible for Quarkus.
- ConstraintVerifier can now test for justifications and indictments, allowing you to write tests that will give you even more confidence in your constraints than was possible before.
- We have exposed new metrics that allow you to better monitor the currently running solver(s).
- We have brought the usual bugfixes and dependency upgrades.
On top of all that, our Enterprise Edition customers can enjoy some exclusive goodies:
- Automatic node sharing. Use cases with a large number of complex constraints may run much faster without any changes to your code. In our benchmarks, we have seen improvements in score calculation speed of up to 100 % on some data sets.
- Nearby Selection can now be enabled with a single switch in your configuration, as opposed to the cumbersome configuration of old. If you're still not using Nearby Selection for your large routing problems, you're missing out on cost savings coming from significantly improved solutions!
Going forward, we will be publishing an Upgrade Recipe to let you know of any things you may or may not run into when upgrading to the latest version of Timefold Solver. It's a good read!
Changelog
🚀 Features
- a0ce0e2 Support unassigned elements for list variables, closes #573
- 8462888 Add support for Spring Boot native image, closes #609
- 53f83bd Automatic node sharing for
ConstraintProvider, closes #685 - 31f66ad Add the
nearby-distance-meter-classproperty for Quarkus and Spring, closes #687 - 3211f8c Enable Nearby Selection for non-empty move selectors configuration, closes #684
- 55c373b Add Nearby configuration by exception, closes #673
- fb946aa Support getting score calculation count and speed for
SolverJob, closes #621 - 388490b Expose entity, variable and value counts in the logs and metrics, closes #671
- 14b7b68 Add penalty methods for
longandBigDecimalvalues without match weighter, closes #656 - 82a9a68 Extend
SingleConstraintAssertionto check Justification/indictment, closes #659 - d2e24d2 Add
withClassLoader(...)method toSolverConfig, closes #691 #695
🐛 Fixes
- b8fdd8d Do not recalculate problem size statistics in Benchmarker
- 6916ba7 Pass provided
ClassLoadertoServiceLoader, closes #690 #692
🧰 Tasks
📝 Documentation
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Geoffrey De Smet
- GitHub
- Lukáš Petrovický (@triceo)
- Pieter De Schepper
- dependabot[bot] (@dependabot[bot])
- marinier
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.8.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.8.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver Community Edition 1.7.0
In this release of Timefold Solver, we focused mostly on bugfixes, improvements to documentation and to the quickstarts.
The release also brings the usual assortment of small tweaks and dependency upgrades.
Featured update: We've made it possible to inject more than one SolverManager in your Quarkus and Spring Boot applications, a feature often requested by the community.
Changelog
🚀 Features
🐛 Fixes
- 964adbd Recommended Fit API needs to ignore any unassigned entities/values (#618), closes #581 #618
- 5c3bcdf Nearby selection pinning support (#614), closes #614
- a9d7c62 Pass all subclasses of Throwable to the exception handler in SolverManager
- 693e321 Ensuring the spent termination property is read in the Quarkus environment (#600), closes #600
- d9122e6 Avoid eager initialization of the Benchmark resources
📝 Documentation
- dc71df8 Adding VRP quickstart documentation (#623), closes #623
- 8c4e1ee Clearly mention that the timegrain pattern doesn't scale (#611), closes #611
- 042be52 Add MoveIteratorFactory example to the docs (#606), closes #606
Contributors
We'd like to thank the following people for their contributions:
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.7.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.7.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver Community Edition 1.6.0
New year, new release, new goodness!
Featured Update - Pinning on list variables, enhancing our support for real-time planning:
- With a new
@PlanningPinToIndexannotation, you can tell the solver that a certain portion of the list variable is pinned and shouldn't be moved. - For example, for a list of customers that a vehicle will visit, this allows you to model situations where certain visits have either already happened, or where the vehicle is already too far into the trip that it would not be practical for it to turn back.
- Previously, pinning was only available for the chained variable. With this, we are ramping up our efforts to bring list variable, which is easier to both understand and implement, on par with the chain variable in terms of features. Stay tuned for more on this front!
Other Enhancements:
- Solver termination can now be overridden without creating a new instance of
SolverorSolverManager, look for theSolverConfigOverrideclass. - The usual assortment of bugfixes and performance improvements.
- Improvements all over the documentation and quickstarts.
- Dependency upgrades.
Changelog
🚀 Features
- 3cb5490 Pinning support for list variable (#491), closes #491
- d9b99b1 Syntactic sugar for ScoreAnalysis (#546), closes #546
- a481fe1 Overriding Solver termination configuration (#542), closes #515 #542
- d2e52cd PlannerBenchmarkFactory.createFromSolverConfig() (#479), closes #479
🐛 Fixes
- 67d4448 Avoid exception when analysing uninitialized solution (#536), closes #536
- a5c3960 Fail fast when @DeepPlanningClone on a record
- 69e3e55 Make MatchAnalysis implement Comparable to fix score corruption analysis (#527), closes #527
- 2b35c89 Updating the School Timetabling quickstarts documentation (#494), closes #98 #494
- 10ba396 Non-existent methods referenced for score explanation
- a8e7a8a Filter trivial 2-opt moves and make KOptListMove public (#485), closes #480 #485
- 0357f3e Do not set shadow variables to null when calculating from scratch in TRACED_FULL_ASSERT (#484), closes #478 #484
🔄️ Changes
- 92ff8fa Reduce allocations in CS
📝 Documentation
- d4a5e1c Add Quickstart code examples for Kotlin (#541), closes #541
- 490490a Rename variable names for collection types (#499), closes #499
- aaa0ebc Use tabs where possible to have better distinction between Maven and Gradle
- 85dd7d9 Maintenance scheduling class diagram rename tagSet to reality
- c8c28a0 Add chapter on CS performance
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Frederico Gonçalves (@zepfred)
- Geoffrey De Smet
- GitHub
- Lukáš Petrovický (@triceo)
- M.Thimm
- Peter Laggner
- Pieter De Schepper
- Radovan Synek
- dependabot[bot] (@dependabot[bot])
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support,
check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, add the ai.timefold.solver : timefold-solver-core : 1.6.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.6.0)
to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated.
It excludes contributions to certain areas of the repository, such as CI and build automation.
This is done for the sake of brevity and to make the user-facing changes stand out more.
Timefold Solver Community Edition 1.5.0
For our final release of 2023, we’ve prepared some exciting things!
Featured Update - Recommended Fit API, a game-changer for appointment scheduling.
- Use case: Imagine you are an operator on a phone line, or an AI chatbot. A customer asks you for an appointment with a sales representative. Where do you put that appointment in their busy calendar?
- Timefold Solver now provides superfast sub-second response times, while still taking all your constraints into account!
- Result: Operators can receive a list of appointment choices, ranked from best fit to worst, for informed decision-making. Ensuring they can make an informed choice that has the least potential to disrupt the schedule as a whole.
Other Enhancements:
- Constraint Streams now simplify solving constraints with sequences of indefinite length.
- Debugging score corruptions is now easier.
- Various other improvements and bug fixes to enhance overall user experience.
Changelog
🚀 Features
- ab44aa6 Introduce the Recommended Fit API, closes #432
- dccbda3 Promote consecutive sequence collector to public API, closes #426
- 4415f4d Add constraint weight to ScoreAnalysis, closes #416
- 86d58e6 Report missing before/after calls on undo corruption, closes #433
- 926bb11 Report shadow variables that differ after recalculating from scratch when undo corruption occurs, closes #430
- 176188b Identify Solver version and edition in Benchmarker, Spring and Quarkus, closes #411
🐛 Fixes
- d8cf97c Shadow entities should be annotated with @PlanningEntityCollectionProperty, closes #445
- ae617f4 Example app logo transparency
- 6f0603f Support repeatable annotations like ShadowVariable in Quarkus
🧰 Tasks
📝 Documentation
- 6de79ff Move development section from user guide to Github CONTRIBUTING
- 663f22d Restructure chapters on using and configuring solver
Contributors
We'd like to thank the following people for their contributions:
- Christopher Chianelli
- Geoffrey De Smet
- Lukáš Petrovický (@triceo)
- Pieter De Schepper
- Radovan Synek
- dependabot[bot] (@dependabot[bot])
Timefold Solver Community Edition is an open source project, and you are more than welcome to contribute as well!
For more, see Contributing.
Should your business need to scale to truly massive data sets or require enterprise-grade support, check out Timefold Solver Enterprise Edition.
How to use Timefold Solver
To see Timefold Solver in action, check out the quickstarts.
With Maven or Gradle, just add the ai.timefold.solver : timefold-solver-core : 1.5.0 dependency in your pom.xml to get started.
You can also import the Timefold Solver Bom (ai.timefold.solver : timefold-solver-bom : 1.5.0) to avoid duplicating version numbers when adding other Timefold Solver dependencies later on.
Additional notes
The changelog and the list of contributors above are automatically generated. They exclude contributions to certain areas of the repository, such as CI and build automation. This is done for the sake of brevity and to make the user-facing changes stand out more.