v3 Ecosystem — NuGet feed
Every sciBASIC# release ships through our own NuGet feed — add one source, pull the whole ecosystem.
sciBASIC# runs its own package server at nuget.scibasic.net — a NuGet v3 feed powered by the ecosystem's own Fluteway/JSql stack, served by VB.NET on .NET 10. Point Visual Studio or any NuGet client at it once and every release library the tutorials use — the data-mining kernels, the plotting stack, the UMAP manifold learner, the IL→CUDA bridge — becomes a one-click install.
01 The feed
One server, the whole ecosystem
Browse
The Package Index lists every published package with its version history and download links — plus server statistics and download graphs.
Inspect
Pick a package to walk its version timeline; every release is served through the standard NuGet v3 flat-container endpoints.
Download
Anonymous by design: the official Visual Studio NuGet client and the dotnet CLI
fetch .nupkg files without any credentials.
Publish
Releases are pushed to the PackagePublish endpoint; uploads are gated behind an
experimental TOTP handshake driven by the xGet CLI.
02 Visual Studio
Wire the feed into the NuGet package manager
Open settings
In Visual Studio go to Tools → NuGet Package Manager → Package Manager
Settings, then select the Package Sources node.
Add the source
Press ➕, name it sciBASIC# and set the source to
https://nuget.scibasic.net/v3/index.json, then Update → OK.
Manage packages
Right-click your solution in Solution Explorer and choose Manage NuGet Packages for Solution.
Switch & browse
In the package manager, pick sciBASIC# in the Package source dropdown, open the Browse tab and search for the module you need.
Install
Select the package, tick the projects it should land in and press Install — from then on
dotnet restore resolves it like any other NuGet source.
03 Command line
Same feed, no GUI required
Visual Studio is only one client. The feed speaks standard NuGet v3, so dotnet, nuget.exe, CI pipelines and any NuGet-compatible tool work unchanged. Drop a nuget.config next to your solution, or register the source once per machine:
<?xml version="1.0" encoding="utf-8"?>
<configuration>
<packageSources>
<add key="sciBASIC#" value="https://nuget.scibasic.net/v3/index.json" />
</packageSources>
</configuration>
$ dotnet nuget add source "https://nuget.scibasic.net/v3/index.json" --name "sciBASIC#"
# install exactly where the tutorials need it, e.g. the data-mining stack:
$ dotnet add package Microsoft.VisualBasic.DataMining.Framework \
--source "sciBASIC#"
# or from the Package Manager Console inside Visual Studio:
PM> Install-Package Microsoft.VisualBasic.DataMining.Framework \
-Source "sciBASIC#"
04 Module families
What the feed ships
The packages mirror the sciBASIC# module tree — the same assemblies the six
tutorials #include. The live
index is always the source of truth for what is currently published:
| Family | Modules | What it gives you | Tutorials |
|---|---|---|---|
| Language kernel & runtime | Microsoft.VisualBasic core | The VB language kernel, the vbs.exe script engine and the runtime every other module sits on. | 01–06 |
| Data & statistics | Data.Framework · Math.Statistics.ANOVA | DataFrameResolver table loading, statistics and hypothesis testing. | 01 |
| Data mining | DataMining.Framework · DataMining.UMAP | k-means clustering, PCA and UMAP manifold learning. | 01 · 03 |
| Visualization & imaging | Data.DataPlot · Data.visualize.Network.* · Drawing | Skia-backed ScatterPlot / HeatmapPlot, the network visualizer with HOLA orthogonal layout. | 01 · 04 |
| NLP | Data.NLP · Data.NLP.Word2Vec | TF-IDF vectorizer, paragraph segmentation and CBOW word-vector training. | 05 · 06 |
| GPU acceleration | Computing.ILCuda | The IL → AST → CUDA pipeline with NVRTC just-in-time compilation. | 02 |
#include in their demo scripts resolves to a package family listed above. Add the feed once
and the whole toolchain installs from a single source.