oxyplot: the version is computed, and the package feed is cluttered
A cross-platform plotting library for .NET
At a glance
- What is it?
- OxyPlot is a plotting library for .NET that splits a chart into a model object and a view control, so adopting it means adding one control and assigning one property. The details worth reading are in the packaging: the version is derived from git tags, the default branch is the development line, and the stable package feed still lists packages the project has unlisted.
- Who is it for?
- OxyPlot fits a .NET desktop application that needs a chart and already has a user interface toolkit, because the whole integration is one view control and one assignment, and the model is plain data you can build in a test. Two things to check before you pin a version.
- Can I use it commercially?
- Yes. MIT is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
- Is it still maintained?
- Yes. The repository last received commits 29 days ago.
- What is it written in?
- Mainly C#, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.
Star history
- oxyplot/oxyplot3,542 ★ · +2
| Repository | 2026-09-24 | 2026-11-15 |
|---|---|---|
| oxyplot/oxyplot | 3.5k | 3.5k |
2 Stars in the last 52 days, measured from the daily GitHub snapshot. Compare with other projects
Editorial analysis
Two types and one property assignment
The getting started section is four steps, and they tell you the whole architecture.
First, add a reference to the library with the NuGet package manager, with a pointer to a later section if you want the pre-release packages. Second, add a view control to your user interface. Third, create a plot model in your code. Fourth, bind the model to the view's model property.
So there are two objects and one assignment. The model holds the chart: the axes, the series, whatever you configured. The view is the control that renders it, which means the library is not tied to one user interface toolkit and you keep whatever windowing and layout your application already uses.
That split is why the model is the interesting half for testing. A chart's data can be constructed and asserted on without a window, and the view can be attached later where a user interface is available.
The fourth step is also the whole binding story. There is no observable collection, no change notification and no data binding framework in the middle; you assign the model and the view renders it.
The stable feed still lists packages the project has unlisted
The packaging section has a paragraph that reads like an apology, and it is the most useful thing in the readme.
Packages come from two feeds. The latest pre-release packages are pushed to a separate package feed, and installing them means setting that feed as a package source with a specific path and remembering a pre-release flag. The stable releases go to the public package repository.
Then the note: there are a lot of old calendar-versioned packages and pre-release packages on the stable feed, and sometimes these show up even if they are unlisted.
Two things follow from that sentence. A tool that enumerates the feed, whether that is an IDE, a build script or a dependency audit, will see versions the project no longer offers. And a package reference without an explicit version can resolve to something you did not intend to install.
The calendar-versioned packages also tell you something about the project's history: the feed carries 2015-era versions alongside the current two-series versions, so the numbering has been through at least two schemes and both are still visible.
The sentence itself has a small editing slip in it, a repeated word, which is a fair indicator of how often the paragraph has been touched.
The version is derived from git tags, and the default branch is develop
The tree contains a configuration file for a version computation tool, which explains the whole release numbering story.
When a version is derived from git history rather than typed into a manifest, the tag is the source of truth and the version string is generated at build time. That is why the releases here are plain version tags with no accompanying changelog file per release, and why the changelog lives as a single file at the root instead.
The other structural detail is the default branch, which is the development line rather than a stable line. A clone of this repository therefore gives you the branch where work happens, not a frozen stable point. For a library that is a meaningful difference, and it is not the convention most repositories follow.
The release dates are irregular. The three visible releases are 2.1.0 in October 2021, 2.1.2 in December 2022 and 2.2.0 in September 2024. The last recorded push to the branch is dated 2026-07-30, so there are commits well above the newest tag.
For an adopter that means the newest published version is about two years older than the newest commit, and the two feeds are how you choose between them.
Examples live inside the library repository
The readme points at one folder for examples, and it is inside the code repository rather than beside it.
That is a small decision with a large convenience attached. The samples are versioned with the library, so an example that compiled against one release keeps compiling against the next, and a reader can see how the API was meant to be used at a specific version by checking out the matching tag.
The cost is that the repository carries sample code it does not ship. The tree confirms the shape: one source directory holding the code and the examples, plus a licence, a changelog, an authors file, a contributors file, a code of conduct, an icons directory and a build pipeline definition. There is no separate samples repository and no documentation tree.
The documentation itself is elsewhere, which is the other half of this arrangement. The readme links a documentation site hosted on a third-party service with a language segment and a latest-version path, so the prose documentation is not versioned alongside the code it describes. The examples are; the manuals are not.
Site, documentation and chat are three separate properties
The more-information block has five links, and they point at five different services.
There is a project web page on a GitHub pages address, a documentation site on a read-the-docs style service with a language path and a latest path, the package feed, a question tag on a developer forum, a hashtag on a social network, and a chat room.
The split between the site and the documentation is the one that matters in practice. The site is a GitHub pages address served from this organisation, so it is versioned with something. The documentation is on a third-party documentation host, which is what makes the path segment for language and version visible in the link rather than in a release.
So the two are maintained on different cadences. A change to the library's API appears on the documentation host when its maintainers rebuild it, and in the examples directory when the next commit lands. For a reader trying to match documentation to a version, the version segment in the documentation link is the only handle available.
The remaining links are support channels rather than documentation: a forum tag, a hashtag and a chat room. None of them is a substitute for versioned reference material.
The contribution guide is on GitHub while the build is on Azure Pipelines
The root of the repository tells you how this project is run, and the arrangement is slightly unusual.
Two directories carry people: an authors file and a contributors file, both at the root, alongside a code of conduct and a changelog. Having both an authors and a contributors list is a small signal about how a long-lived library tracks who did what.
Then the tooling. The contribution guide lives inside the GitHub configuration directory, and the continuous integration definition is a pipeline file for Microsoft's hosted build service rather than a GitHub Actions workflow. So the instructions for contributing are on GitHub while the build that validates them is somewhere else entirely.
The version computation configuration sits in the same layer, alongside an attributes file for line endings, an ignore file, an icons directory and the licence.
There is no test project in the visible tree, no package manifest and no build script in the conventional sense, which is consistent with a library whose build is driven by the version tool and the pipeline configuration rather than by a checked-in project file. It also means a contributor cloning the repository has to read the pipeline definition to find out how to build it at all.
There is no feature list, and everything else is on another site
The readme is about 170 words, and it is worth being precise about what that buys.
It contains a title, the getting started steps, a pointer to the examples folder, the packaging section with its two feeds and its note about unlisted packages, a link to a wiki page about the available packages, five links for more information, and a pointer to the contribution guide.
That is the whole document. There is no list of what the library can plot, no configuration reference, no mention of export formats, no theming section, no compatibility statement about which .NET versions are supported, and no note about which user interface toolkits the view control works with.
For a plotting library, that last omission is the one to notice. The whole appeal of the model and view split is that the view is not tied to one toolkit, and the readme does not say which toolkits are supported or how a different one would be added.
None of that makes the project unusable. It makes the readme a signpost rather than a manual, and the five links are the manual. A reader who stops at the readme will know how to bind a model to a view and nothing about what to put in the model.
Editorial conclusion
OxyPlot fits a .NET desktop application that needs a chart and already has a user interface toolkit, because the whole integration is one view control and one assignment, and the model is plain data you can build in a test. Two things to check before you pin a version. The newest release in this repository is from September 2024 while commits continue, so the branch is ahead of the last tag and the version you get depends on which feed you install from. And the stable feed is not a clean list of supported versions: it still carries calendar-versioned packages from 2015 and unlisted entries, so any tooling that enumerates the feed will show you more than the project intends to offer. The documentation lives on a third-party service rather than in the repository, which is the first place to look when the readme feels too short.
Frequently asked questions
how to use oxyplot
Add a package reference with the NuGet package manager, add a PlotView to your user interface, create a PlotModel in your code, and bind the model to the Model property of the view. Worked examples live in the Source/Examples folder of the repository.
what is oxyplot dll
The readme does not describe a standalone library file. It distributes OxyPlot as NuGet packages, with stable releases on the public package repository and the latest pre-release packages on a separate feed, where installing them means setting that feed as a package source and using the pre-release flag.
oxyplot alternatives
This repository does not list alternatives. It describes OxyPlot as a cross-platform plotting library for .NET and points to its own web page, a documentation site hosted separately, a wiki page about the available packages, and a question tag on a developer forum.
How do I get pre-release OxyPlot packages?
They are pushed to a separate package feed. You set that feed as a package source using its oxyplot path and remember the pre-release flag. The readme also notes that the stable feed still carries many old calendar-versioned and pre-release packages, and that some of them appear even though they are unlisted.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/oxyplot-oxyplot)Star history chart for your README
[](https://hysenlabs.com/projects/oxyplot-oxyplot)Was this useful?