View precompilation
Razor views are compiled at runtime by default. The first request after deployment therefore triggers a compilation step that can take noticeable time and potentially fail only in production. Precompiling views at build time avoids that delay and surfaces syntax errors early.
Default configuration
Smartstore ships with Smartstore.Razor.props
which enables build- and publish-time compilation for all projects referencing it. The props file turns on the modern Razor SDK switches and disables the build server:
<PropertyGroup>
<AddRazorSupportForMvc>true</AddRazorSupportForMvc>
<MvcRazorCompileOnBuild>false</MvcRazorCompileOnBuild>
<MvcRazorCompileOnPublish>true</MvcRazorCompileOnPublish>
<RazorCompileOnBuild>true</RazorCompileOnBuild>
<RazorCompileOnPublish>true</RazorCompileOnPublish>
<UseRazorBuildServer>false</UseRazorBuildServer>
</PropertyGroup>
With these settings every dotnet build
or dotnet publish
creates a <project>.Views.dll
alongside the main assembly so views are ready to execute.
Building with precompiled views
To produce precompiled views for the web host simply run:
dotnet publish src/Smartstore.Web/Smartstore.Web.csproj -c Release
The publish output contains Smartstore.Web.dll
and Smartstore.Web.Views.dll
which ASP.NET Core loads automatically at startup.
Skipping compilation in development
During rapid prototyping you can skip view compilation by building the special DebugNoRazorCompile
configuration:
dotnet build src/Smartstore.Web/Smartstore.Web.csproj -c DebugNoRazorCompile
This configuration sets RazorCompileOnBuild
to false
, shortening build time at the cost of a slower first request.
Modules can also opt out individually by adding to their project file:
<PropertyGroup>
<RazorCompileOnBuild>false</RazorCompileOnBuild>
</PropertyGroup>
Benefits
Faster cold start because no runtime view compilation is necessary.
Compilation errors surface during CI/CD rather than at runtime.
Precompiled views can be trimmed and optimized as part of the build pipeline.
Precompiling views is a low‑effort optimization that improves reliability and perceived performance for production deployments.
Last updated
Was this helpful?