Skip to content

User guide (introductory)

Stack is a modern, cross-platform build tool for Haskell code.

This introductory guide takes a new Stack user through the typical workflows. This guide will not teach Haskell or involve much code, and it requires no prior experience with the Haskell packaging system or other build tools. Terms used in the guide are defined in the glossary.

Some of Stack's features will not be needed regularly or by all users. See the advanced user's guide for information about those features.

Stack's functions

Stack handles the management of your toolchain (including GHC — the Glasgow Haskell Compiler — and, for Windows users, MSYS2), building and registering libraries, building build tool dependencies, and more. While it can use existing tools on your system, Stack has the capacity to be your one-stop shop for all Haskell tooling you need. This guide will follow that Stack-centric approach.

What makes Stack special?

The primary Stack design point is reproducible builds. If you run stack build today, you should get the same result running stack build tomorrow. There are some cases that can break that rule (changes in your operating system configuration, for example), but, overall, Stack follows this design philosophy closely. To make this a simple process, Stack uses curated package sets called snapshots.

Stack has also been designed from the ground up to be user friendly, with an intuitive, discoverable command line interface. For many users, simply downloading Stack and reading stack --help will be enough to get up and running. This guide provides a more gradual tour for users who prefer that learning style.

To build your project, Stack uses a project-level configuration file, named stack.yaml, in the root directory of your project as a sort of blueprint. That file contains a reference, called a resolver, to the snapshot which your package will be built against.

Finally, Stack is isolated: it will not make changes outside of specific Stack directories. Stack-built files generally go in either the Stack root directory or ./.stack-work directories local to each project. The Stack root directory holds packages belonging to snapshots and any Stack-installed versions of GHC. Stack will not tamper with any system version of GHC or interfere with packages installed by other build tools, such as Cabal (the tool).

Downloading and Installation

The documentation dedicated to downloading Stack has the most up-to-date information for a variety of operating systems. Instead of repeating that content here, please go check out that page and come back here when you can successfully run stack --version.

We also assume that the directory reported by stack path --local-bin has been added to the PATH.

Hello World Example

With Stack installed, let's create a new project from a template and walk through the most common Stack commands.

In this guide, an initial $ represents the command line prompt. The prompt may differ in the terminal on your operating system. Unless stated otherwise, the working directory is the project's root directory.

The stack new command

We'll start off with the stack new command to create a new project, that will contain a Haskell package of the same name. So let's pick a valid package name first:

A package is identified by a globally-unique package name, which consists of one or more alphanumeric words separated by hyphens. To avoid ambiguity, each of these words should contain at least one letter.

(From the Cabal users guide)

We'll call our project helloworld, and we'll use the new-template project template. This template is used by default, but in our example we will refer to it expressly. Other templates are available. For further information about templates, see the stack templates command documentation.

From the root directory for all our Haskell projects, we command:

stack new helloworld new-template

For this first Stack command, there's quite a bit of initial setup it needs to do (such as downloading the list of packages available upstream), so you'll see a lot of output. Over the course of this guide a lot of the content will begin to make more sense.

After creating the project directory, and obtaining and populating the project template, Stack will initialise its own project-level configuration. For further information about setting paramaters to populate templates, see the YAML configuration documentation. For further information about initialisation, see the stack init command documentation. The stack new and stack init commands have options and flags in common.

Info

Pass the --bare flag to cause Stack to create the project in the current working directory rather than in a new project directory.

Info

Parameters to populate project templates can be set at the command line with the --param <key>:<value> (or -p) option.

We now have a project in the helloworld directory! We will change to that directory, with command:

cd helloworld

The stack build command

Next, we'll run the most important Stack command, stack build:

stack build
# installing ... building ...

Stack needs a version of GHC in order to build your project. Stack will discover that you are missing it and will install it for you.

You'll get intermediate download percentage statistics while the download is occurring. This command may take some time, depending on download speeds.

Note

GHC will be installed to your Stack programs directory, so calling ghc on the command line won't work. See the stack exec, stack ghc, and stack runghc commands below for more information.

Once a version of GHC is installed, Stack will then build your project.

The stack exec command

Looking closely at the output of the previous command, you can see that it built both a library called helloworld and an executable called helloworld-exe (on Windows, helloworld-exe.exe). We'll explain more in the next section, but, for now, just notice that the executables are installed in a location in our project's .stack-work directory.

Now, Let's use the stack exec command to run our executable (which just outputs "someFunc"):

stack exec helloworld-exe
someFunc

stack exec works by providing the same reproducible environment that was used to build your project to the command that you are running. Thus, it knew where to find helloworld-exe even though it is hidden in the .stack-work directory. Command stack path --bin-path to see the PATH in the Stack environment.

Info

On Windows, the Stack environment includes the \mingw64\bin, \usr\bin and \usr\local\bin directories of the Stack-supplied MSYS2. If your executable depends on files (for example, dynamic-link libraries) in those directories and you want ro run it outside of the Stack environment, you will need to ensure copies of those files are on the PATH.

The stack test command

Finally, like all good software, helloworld actually has a test suite.

Let's run it with the stack test command:

stack test
# build output ...

Reading the output, you'll see that Stack first builds the test suite and then automatically runs it for us. For both the build and test command, already built components are not built again. You can see this by using the stack build and stack test commands a second time:

stack build
stack test
# build output ...

Inner Workings of Stack

In this subsection, we'll dissect the helloworld example in more detail.

Files in helloworld

Before studying Stack more, let's understand our project a bit better. The files in the directory include:

app/Main.hs
src/Lib.hs
test/Spec.hs
ChangeLog.md
README.md
LICENSE
Setup.hs
helloworld.cabal
package.yaml
stack.yaml
.gitignore

The app/Main.hs, src/Lib.hs, and test/Spec.hs files are all Haskell source files that compose the actual functionality of our project (we won't dwell on them here).

The ChangeLog.md, README.md, LICENSE and .gitignore files have no effect on the build.

The helloworld.cabal file is updated automatically as part of the stack build process and should not be modified.

The files of interest here are Setup.hs, stack.yaml, and package.yaml.

The Setup.hs file is a component of the Cabal build system which Stack uses. It's technically not needed by Stack, but it is still considered good practice in the Haskell world to include it. The file we're using is straight boilerplate:

import Distribution.Simple
main = defaultMain

Next, let's look at our stack.yaml file, which gives our project-level settings. Ignoring comments beginning #, the contents will look something like this:

resolver:
  url: https://raw.githubusercontent.com/commercialhaskell/stackage-snapshots/master/lts/19/28.yaml
packages:
- .

The value of the resolver key tells Stack how to build your package: which GHC version to use, versions of package dependencies, and so on. Our value here says to use LTS Haskell 19.28, which implies GHC 9.0.2 (which is why stack build installs that version of GHC if it is not already available to Stack). There are a number of values you can use for resolver, which we'll cover later.

The value of the packages key tells Stack which local packages to build. In our simple example, we have only a single package in our project, located in the same directory, so '.' suffices. However, Stack has powerful support for multi-package projects, which we'll elaborate on as this guide progresses.

Another file important to the build is package.yaml.

The package.yaml file describes the package in the Hpack format. Stack has in-built Hpack functionality and this is its preferred package format. The default behaviour is to generate the Cabal file (here named helloworld.cabal) from this package.yaml file, and accordingly you should not modify the Cabal file.

It is also important to remember that Stack is built on top of the Cabal build system. Therefore, an understanding of the moving parts in Cabal are necessary. In Cabal, we have individual packages, each of which contains a single Cabal file, named <package_name>.cabal. The Cabal file can define one or more components: a library, executables, test suites, and benchmarks. It also specifies additional information such as library dependencies, default language pragmas, and so on.

In this guide, we'll discuss the bare minimum necessary to understand how to modify a package.yaml file. You can see a full list of the available options at the Hpack documentation. The Cabal User Guide the definitive reference for the Cabal file format.

The location of GHC

As we saw above, the build command installed GHC for us. You can use the stack path command for quite a bit of path information (which we'll play with more later). We'll look at where GHC is installed:

Command:

stack exec -- which ghc
/home/<user_name>/.stack/programs/x86_64-linux/ghc-9.0.2/bin/ghc

Command:

stack exec -- where.exe ghc
C:\Users\<user_name>\AppData\Local\Programs\stack\x86_64-windows\ghc-9.0.2\bin\ghc.exe

As you can see from that path (and as emphasized earlier), the installation is placed to not interfere with any other GHC installation, whether system-wide or even different GHC versions installed by Stack.

Cleaning your project

You can clean up build artifacts for your project using the stack clean and stack purge commands.

The stack clean command

stack clean deletes the local working directories containing compiler output. By default, that means the contents of directories in .stack-work/dist, for all the .stack-work directories within a project.

Use stack clean <specific-package> to delete the output for the package specific-package only.

The stack purge command

stack purge deletes the local stack working directories, including extra-deps, git dependencies and the compiler output (including logs). It does not delete any snapshot packages, compilers or programs installed using stack install. This essentially reverts the project to a completely fresh state, as if it had never been built. stack purge is just a shortcut for stack clean --full

The stack build command

The build command is the heart and soul of Stack. It is the engine that powers building your code, testing it, getting dependencies, and more. Quite a bit of the remainder of this guide will cover more advanced build functions and features, such as building test and Haddocks at the same time, or constantly rebuilding blocking on file changes.

Note

Using the build command twice with the same options and arguments should generally do nothing (besides things like rerunning test suites), and should, in general, produce a reproducible result between different runs.

Adding dependencies

Let's say we decide to modify our helloworld source a bit to use a new library, perhaps the ubiquitous text package. In src/Lib.hs, we can, for example add:

{-# LANGUAGE OverloadedStrings #-}
module Lib
    ( someFunc
    ) where

import qualified Data.Text.IO as T

someFunc :: IO ()
someFunc = T.putStrLn "someFunc"

When we try to build this, things don't go as expected:

stack build
# build failure output (abridged for clarity) ...
src\Lib.hs:6:1: error:
    Could not load module ‘Data.Text.IO’
    It is a member of the hidden package ‘text-1.2.5.0’.
    Perhaps you need to add ‘text’ to the build-depends in your .cabal file.
    Use -v (or `:set -v` in ghci) to see a list of the files searched for.
  |
6 | import qualified Data.Text.IO as T
  | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

This means that the package containing the module in question is not available. To tell Stack to use text, you need to add it to your package.yaml file — specifically in your dependencies section, like this:

dependencies:
- base >= 4.7 && < 5
- text # added here

Now if we rerun stack build, we should get a successful result. Command:

stack build
# build output ...

This output means that the text package was downloaded, configured, built, and locally installed. Once that was done, we moved on to building our local package (helloworld). At no point did we need to ask Stack to build dependencies — it does so automatically.

Listing Dependencies

Let's have Stack add a few more dependencies to our project. First, we'll include two new packages in the dependencies section for our library in our package.yaml:

dependencies:
- base >= 4.7 && < 5
- text
- filepath
- containers

After adding these two dependencies, we can again run stack build to have them installed. Command:

stack build
# build output ...

Finally, to find out which versions of these libraries Stack installed, we can ask Stack to ls dependencies. Command:

stack ls dependencies
# dependency output ...

extra-deps

Let's try a more off-the-beaten-track package: the joke acme-missiles package. Our source code is simple:

module Lib
    ( someFunc
    ) where

import Acme.Missiles

someFunc :: IO ()
someFunc = launchMissiles

Again, we add this new dependency to the package.yaml file like this:

dependencies:
- base >= 4.7 && < 5
- text
- filepath
- containers
- acme-missiles # added

However, rerunning stack build shows us the following error message. Command:

stack build
# build failure output ...

It says that it was unable to construct the build plan.

This brings us to the next major topic in using Stack.

Curated package sets

Remember above when stack new selected some LTS resolver for us? That defined our build plan and available packages. When we tried using the text package, it just worked, because it was part of the LTS package set.

We've specified the acme-missiles package in the package.yaml file (see above), but acme-missiles is not part of that LTS package set, so building failed.

To add acme-missiles to the available packages, we'll use the extra-deps key in the stack.yaml file. That key defines extra packages, not present in the resolver, that will be needed as dependencies. You can add this like so:

extra-deps:
- acme-missiles-0.3 # not in the LTS resolver

Now stack build will succeed.

With that out of the way, let's dig a little bit more into these package sets, also known as snapshots. We mentioned the LTS resolvers, and you can get quite a bit of information about it at https://www.stackage.org/lts, including:

  • The appropriate resolver value (resolver: lts-20.4, as is currently the latest LTS)
  • The GHC version used
  • A full list of all packages available in this snapshot
  • The ability to perform a Hoogle search on the packages in this snapshot
  • A list of all modules in a snapshot, which can be useful when trying to determine which package to add to your package.yaml file.

You can also see a list of all available snapshots. You'll notice two flavors: LTS (for "Long Term Support") and Nightly. You can read more about them on the LTS Haskell GitHub page. If you're not sure which to use, start with LTS Haskell (which Stack will lean towards by default as well).

Resolvers and changing your compiler version

Let's explore package sets a bit further. Instead of lts-19.17, let's change our stack.yaml file to use the latest nightly. Right now, this is currently 2022-12-16 - please see the resolver from the link above to get the latest.

Then, commanding stack build again will produce:

stack build
# Downloaded nightly-2022-12-16 build plan.
# build output ...

We can also change resolvers on the command line, which can be useful in a Continuous Integration (CI) setting, like on Travis. For example, command:

stack --resolver lts-18.28 build
# Downloaded lts-18.28 build plan.
# build output ...

When passed on the command line, you also get some additional "short-cut" versions of resolvers: --resolver nightly will use the newest Nightly resolver available, --resolver lts will use the newest LTS, and --resolver lts-19 will use the newest LTS in the 19.x series. The reason these are only available on the command line and not in your stack.yaml file is that using them:

  1. Will slow down your build (since Stack then needs to download information on the latest available LTS each time it builds)
  2. Produces unreliable results (since a build run today may proceed differently tomorrow because of changes outside of your control)

Changing GHC versions

Finally, let's try using an older LTS snapshot. We'll use the newest 18.x snapshot with the command:

stack --resolver lts-18 build
# build output ...

This succeeds, automatically installing the necessary GHC along the way. So, we see that different LTS versions use different GHC versions and Stack can handle that.

Other resolver values

We've mentioned nightly-YYYY-MM-DD and lts-X.Y values for the resolver. There are actually other options available, and the list will grow over time. At the time of writing:

  • ghc-X.Y.Z, for requiring a specific GHC version but no additional packages
  • Experimental custom snapshot support

The most up-to-date information can always be found in the stack.yaml documentation.

Existing projects

Alright, enough playing around with simple projects. Let's take an open source package and try to build it. We'll be ambitious and use yackage, a local package server using Yesod. To get the code, we'll use the stack unpack command from the root directory for all our Haskell projects:

stack unpack yackage
Unpacked yackage-0.8.1 to <root_directory>/yackage-0.8.1/

You can also unpack to the directory of your liking instead of the current one by issuing the command:

stack unpack yackage --to <desired_directory>

This will create a yackage-0.8.1 directory inside <desired_directory>.

We will change to that directory, with the command:

cd yackage-0.8.1

The stack init command

This new directory does not have a stack.yaml file, so we need to make one first. We could do it by hand, but let's be lazy instead with the stack init command:

stack init
# init output ...

stack init does quite a few things for you behind the scenes:

  • Finds all of the Cabal files in your current directory and subdirectories (unless you use --ignore-subdirs) and determines the packages and versions they require
  • Finds the best combination of snapshot and package flags that allows everything to compile with minimum external dependencies
  • It tries to look for the best matching snapshot from latest LTS, latest nightly, other LTS versions in that order

Assuming it finds a match, it will write your stack.yaml file, and everything will work.

Note

The yackage package does not currently support Hpack, but you can also use hpack-convert should you need to generate a package.yaml file.

Excluded Packages

Sometimes multiple packages in your project may have conflicting requirements. In that case stack init will fail, so what do you do?

You could manually create stack.yaml by omitting some packages to resolve the conflict. Alternatively you can ask stack init to do that for you by specifying --omit-packages flag on the command line. Let's see how that works.

To simulate a conflict we will use acme-missiles-0.3 in yackage and we will also copy yackage.cabal to another directory and change the name of the file and package to yackage-test. In this new package we will use acme-missiles-0.2 instead. Let's see what happens when we command stack init again:

stack init --force --omit-packages
# init failure output ...

Looking at stack.yaml, you will see that the excluded packages have been commented out under the packages field. In case wrong packages are excluded you can uncomment the right one and comment the other one.

Packages may get excluded due to conflicting requirements among user packages or due to conflicting requirements between a user package and the resolver compiler. If all of the packages have a conflict with the compiler then all of them may get commented out.

When packages are commented out you will see a warning every time you run a command which needs the configuration file. The warning can be disabled by editing the configuration file and removing it.

Using a specific resolver

Sometimes you may want to use a specific resolver for your project instead of stack init picking one for you. You can do that by using stack init --resolver <resolver>.

You can also init with a compiler resolver if you do not want to use a snapshot. That will result in all of your project's dependencies being put under the extra-deps section.

Installing the compiler

Stack will automatically install the compiler when you run stack build but you can manually specify the compiler by running stack setup <GHC-VERSION>.

Miscellaneous and diagnostics

Add selected packages: If you want to use only selected packages from your project directory you can do so by explicitly specifying the package directories on the command line.

Duplicate package names: If multiple packages under the directory tree have same name, stack init will report those and automatically ignore one of them.

Ignore subdirectories: By default stack init searches all the subdirectories for Cabal files. If you do not want that then you can use --ignore-subdirs command line switch.

Cabal warnings: stack init will show warnings if there were issues in reading a Cabal file. You may want to pay attention to the warnings as sometimes they may result in incomprehensible errors later on during dependency solving.

Package naming: If the Name field defined in a Cabal file does not match with the Cabal file name then stack init will refuse to continue.

User warnings: When packages are excluded or external dependencies added Stack will show warnings every time the configuration file is loaded. You can suppress the warnings by editing the configuration file and removing the warnings from it. You may see something like this:

stack build
Warning: Some packages were found to be incompatible with the resolver and have been left commented out in the packages section.
Warning: Specified resolver could not satisfy all dependencies. Some external packages have been added as dependencies.
You can suppress this message by removing it from stack.yaml

Different databases

Time to take a short break from hands-on examples and discuss a little architecture. Stack has the concept of multiple databases.

A database consists of a GHC package database (which contains the compiled version of a library), executables, and a few other things as well. To give you an idea, the contents of the parent directory of the stack path --local-pkg-db directory are the directories:

bin
doc
lib
pkgdb

Databases in Stack are layered. For example, the database listing we just gave is called a local database (also known as a mutable database). That is layered on top of a snapshot database (also known as a write-only database). The snapshot database contains the libraries and executables that are considered to be immutable. Finally, GHC itself ships with a number of libraries and executables, also considered to be immutable, which forms the global database.

To get a quick idea of this, we can look at the output of the stack exec -- ghc-pkg list command in our helloworld project:

<stack path --global-pkg-db directory>
    Cabal-3.6.3.0
    Win32-2.12.0.1
    array-0.5.4.0
    base-4.16.2.0
    binary-0.8.9.0
    bytestring-0.11.3.1
    containers-0.6.5.1
    deepseq-1.4.6.1
    directory-1.3.6.2
    exceptions-0.10.4
    filepath-1.4.2.2
    (ghc-9.2.3)
    ghc-bignum-1.2
    ghc-boot-9.2.3
    ghc-boot-th-9.2.3
    ghc-compact-0.1.0.0
    ghc-heap-9.2.3
    ghc-prim-0.8.0
    ghci-9.2.3
    haskeline-0.8.2
    hpc-0.6.1.0
    integer-gmp-1.1
    libiserv-9.2.3
    mtl-2.2.2
    parsec-3.1.15.0
    pretty-1.1.3.6
    process-1.6.13.2
    rts-1.0.2
    stm-2.5.0.2
    template-haskell-2.18.0.0
    text-1.2.5.0
    time-1.11.1.1
    transformers-0.5.6.2
    xhtml-3000.2.2.1

<stack path --snapshot-pkg-db directory>
    acme-missiles-0.3

<stack path --local-pkg-db directory>
    helloworld-0.1.0.0

where <stack path --global-pkg-db directory> refers to the directory output by the command stack path --global-pkg-db, and so on.

Notice that acme-missiles ends up in the snapshot database. Any package which comes from Hackage, an archive, or a repository is considered to be an immutable package.

Anything which is considered mutable, or depends on something mutable, ends up in the local database. This includes your own code and any other packages located on a local file path.

The reason we have this structure is that:

  • it lets multiple projects reuse the same binary builds of immutable packages,
  • but doesn't allow different projects to "contaminate" each other by putting non-standard content into the shared snapshot database.

As you probably guessed, there can be multiple snapshot databases available. See the contents of the snapshots directory in the Stack root.

  • On Unix-like operating systems, each snapshot is in the last of a sequence of three subdirectories named after the platform, a 256-bit hash of the source map (how the package should be built -- including the compiler, options, and immutable dependencies), and the GHC version.

  • On Windows, each snapshot is in a subdirectory that is a shorter hash (eight characters) of the sequence of three directories used on Unix-like operating systems. This is done to avoid problems created by default limits on file path lengths on Windows systems.

These snapshot databases don't get layered on top of each other; they are each used separately.

In reality, you'll rarely — if ever — interact directly with these databases, but it's good to have a basic understanding of how they work so you can understand why rebuilding may occur at different points.

The build synonyms

Let's look at a subset of the stack --help output:

build    Build the package(s) in this directory/configuration
install  Shortcut for 'build --copy-bins'
test     Shortcut for 'build --test'
bench    Shortcut for 'build --bench'
haddock  Shortcut for 'build --haddock'

Four of these commands are just synonyms for the build command. They are provided for convenience for common cases (e.g., stack test instead of stack build --test) and so that commonly expected commands just work.

What's so special about these commands being synonyms? It allows us to make much more composable command lines. For example, we can have a command that builds executables, generates Haddock documentation (Haskell API-level docs), and builds and runs your test suites, with:

stack build --haddock --test

You can even get more inventive as you learn about other flags. For example, take the following command:

stack build --pedantic --haddock --test --exec "echo Yay, it succeeded" --file-watch

This command will:

  • turn on all warnings and errors (the --pedantic flag)
  • build your library and executables
  • generate Haddocks (the --haddock flag)
  • build and run your test suite (the --test flag)
  • run the command echo Yay, it succeeded when that completes (the --exec option)
  • after building, watch for changes in the files used to build the project, and kick off a new build when done (the --file-watch flag)

The stack install command and copy-bins option

It's worth calling out the behavior of the install command and --copy-bins option, since this has confused a number of users (especially when compared to behavior of other tools like Cabal (the tool)). The install command does precisely one thing in addition to the build command: it copies any generated executables to the local binary directory. You may recognize the default value for that path:

On Unix-like operating systems, command:

stack path --local-bin
/home/<user_name>/.local/bin

On Windows, command:

stack path --local-bin
C:\Users\<user_name>\AppData\Roaming\local\bin

That's why the download page recommends adding that directory to your PATH. This feature is convenient, because now you can simply run executable-name in your shell instead of having to run stack exec executable-name from inside your project directory.

Since it's such a point of confusion, let me list a number of things Stack does not do specially for the install command:

  • Stack will always build any necessary dependencies for your code. The install command is not necessary to trigger this behavior. If you just want to build a project, run stack build.
  • Stack will not track which files it's copied to your local binary directory nor provide a way to automatically delete them. There are many great tools out there for managing installation of binaries, and Stack does not attempt to replace those.

  • Stack will not necessarily be creating a relocatable executable. If your executables hard-codes paths, copying the executable will not change those hard-coded paths.

  • At the time of writing, there's no way to change those kinds of paths with Stack, but see issue #848 about --prefix for future plans.

That's really all there is to the install command: for the simplicity of what it does, it occupies a much larger mental space than is warranted.

Targets, locals, and extra-deps

We haven't discussed this too much yet, but, in addition to having a number of synonyms and taking a number of options on the command line, the build command also takes many arguments. These are parsed in different ways, and can be used to achieve a high level of flexibility in telling Stack exactly what you want to build.

We're not going to cover the full generality of these arguments here; instead, there's documentation covering the full build command syntax. Here, we'll just point out a few different types of arguments:

  • You can specify a package name, e.g. stack build vector.
    • This will attempt to build the vector package, whether it's a local package, in your extra-deps, in your snapshot, or just available upstream. If it's just available upstream but not included in your locals, extra-deps, or snapshot, the newest version is automatically promoted to an extra-dep.
  • You can also give a package identifier, which is a package name plus version, e.g. stack build yesod-bin-1.4.14.
    • This is almost identical to specifying a package name, except it will (1) choose the given version instead of latest, and (2) error out if the given version conflicts with the version of a local package.
  • The most flexibility comes from specifying individual components, e.g. stack build helloworld:test:helloworld-test says "build the test suite component named helloworld-test from the helloworld package."
    • In addition to this long form, you can also shorten it by skipping what type of component it is, e.g. stack build helloworld:helloworld-test, or even skip the package name entirely, e.g. stack build :helloworld-test.
  • Finally, you can specify individual directories to build to trigger building of any local packages included in those directories or subdirectories.

When you give no specific arguments on the command line (e.g., stack build), it's the same as specifying the names of all of your local packages. If you just want to build the package for the directory you're currently in, you can use stack build ..

Components, --test, and --bench

Here's one final important yet subtle point. Consider our helloworld package: it has a library component, an executable helloworld-exe, and a test suite helloworld-test. When you run stack build helloworld, how does it know which ones to build? By default, it will build the library (if any) and all of the executables but ignore the test suites and benchmarks.

This is where the --test and --bench flags come into play. If you use them, those components will also be included. So stack build --test helloworld will end up including the helloworld-test component as well.

You can bypass this implicit adding of components by being much more explicit, and stating the components directly. For example, the following will not build the helloworld-exe executable once all executables have been successfully built:

stack clean
stack build :helloworld-test
Building all executables for `helloworld' once. After a successful build of all of them, only specified executables will be rebuilt.
helloworld> configure (lib + exe + test)
Configuring helloworld-0.1.0.0...
helloworld> build (lib + exe + test)
Preprocessing library for helloworld-0.1.0.0..
Building library for helloworld-0.1.0.0..
[1 of 2] Compiling Lib
[2 of 2] Compiling Paths_helloworld
Preprocessing executable 'helloworld-exe' for helloworld-0.1.0.0..
Building executable 'helloworld-exe' for helloworld-0.1.0.0..
[1 of 2] Compiling Main
[2 of 2] Compiling Paths_helloworld
Linking .stack-work\dist\<hash>\build\helloworld-exe\helloworld-exe.exe ...
Preprocessing test suite 'helloworld-test' for helloworld-0.1.0.0..
Building test suite 'helloworld-test' for helloworld-0.1.0.0..
[1 of 2] Compiling Main
[2 of 2] Compiling Paths_helloworld
Linking .stack-work\dist\<hash>\build\helloworld-test\helloworld-test.exe ...
helloworld> copy/register
Installing library in ...\helloworld\.stack-work\install\...
Installing executable helloworld-exe in ...\helloworld\.stack-work\install\...\bin
Registering library for helloworld-0.1.0.0..
helloworld> test (suite: helloworld-test)

Test suite not yet implemented

helloworld> Test suite helloworld-test passed
Completed 2 action(s).

We first cleaned our project to clear old results so we know exactly what Stack is trying to do. Note that it says it is building all executables for helloworld once, and that after a successful build of all of them, only specified executables will be rebuilt. If we change the source code of test/Spec.hs, say to:

main :: IO ()
main = putStrLn "Test suite still not yet implemented"

and command again:

stack build :helloworld-test
helloworld-0.1.0.0: unregistering (local file changes: test\Spec.hs)
helloworld> build (lib + test)
Preprocessing library for helloworld-0.1.0.0..
Building library for helloworld-0.1.0.0..
Preprocessing test suite 'helloworld-test' for helloworld-0.1.0.0..
Building test suite 'helloworld-test' for helloworld-0.1.0.0..
[2 of 2] Compiling Main
Linking .stack-work\dist\<hash>\build\helloworld-test\helloworld-test.exe ...
helloworld> copy/register
Installing library in ...\helloworld\.stack-work\install\...
Installing executable helloworld-exe in ...\helloworld\.stack-work\install\...\bin
Registering library for helloworld-0.1.0.0..
helloworld> blocking for directory lock on ...\helloworld\.stack-work\dist\<hash>\build-lock
helloworld> test (suite: helloworld-test)

Test suite still not yet implemented

helloworld> Test suite helloworld-test passed
Completed 2 action(s).

Notice that this time it builds the helloworld-test test suite, and the helloworld library (since it's used by the test suite), but it does not build the helloworld-exe executable.

And now the final point: in both cases, the last line shows that our command also runs the test suite it just built. This may surprise some people who would expect tests to only be run when using stack test, but this design decision is what allows the stack build command to be as composable as it is (as described previously). The same rule applies to benchmarks. To spell it out completely:

  • The --test and --bench flags simply state which components of a package should be built, if no explicit set of components is given
  • The default behavior for any test suite or benchmark component which has been built is to also run it

You can use the --no-run-tests and --no-run-benchmarks flags to disable running of these components. You can also use --no-rerun-tests to prevent running a test suite which has already passed and has not changed.

Note

Stack doesn't build or run test suites and benchmarks for non-local packages. This is done so that a command like stack test doesn't need to run 200 test suites!

Multi-package projects

Until now, everything we've done with Stack has used a single-package project. However, Stack's power truly shines when you're working on multi-package projects. All the functionality you'd expect to work just does: dependencies between packages are detected and respected, dependencies of all packages are just as one cohesive whole, and if anything fails to build, the build commands exits appropriately.

Let's demonstrate this with the wai-app-static and yackage packages, starting in the root directory for all our Haskell projects. Command:

mkdir multi
cd multi
stack unpack wai-app-static yackage
Unpacked wai-app-static (from Hackage) to .../multi/wai-app-static-3.1.7.4/
Unpacked yackage (from Hackage) to .../multi/yackage-0.8.1/
stack init
Looking for .cabal or package.yaml files to use to init the project.
Using cabal packages:
- wai-app-static-3.1.7.4/
- yackage-0.8.1/

Cabal file warning in .../multi/yackage-0.8.1/yackage.cabal@47:40: version operators used. To use version operators the package needs to specify at least 'cabal-version: >= 1.8'.
Cabal file warning in .../multi/yackage-0.8.1/yackage.cabal@21:36: version operators used. To use version operators the package needs to specify at least 'cabal-version: >= 1.8'.
Selecting the best among 18 snapshots...

* Matches ...

Selected resolver: ...
Initialising configuration using resolver: ...
Total number of user packages considered: 2
Writing configuration to file: stack.yaml
stack build --haddock --test
# Goes off to build a whole bunch of packages

If you look at the stack.yaml file, you'll see exactly what you'd expect:

resolver:
  url: https://raw.githubusercontent.com/commercialhaskell/stackage-snapshots/master/lts/19/17.yaml
packages:
- wai-app-static-3.1.7.4
- yackage-0.8.1

Notice that multiple directories are listed in the packages key.

In addition to local directories, you can also refer to packages available in a Git repository or in a tarball over HTTP/HTTPS. This can be useful for using a modified version of a dependency that hasn't yet been released upstream.

Note

When adding upstream packages directly to your project it is important to distinguish local packages from the upstream dependency packages. Otherwise you may have trouble running stack ghci. See stack.yaml documentation for more details.

Flags and GHC options

There are two common ways to alter how a package will install: with Cabal flags and with GHC options.

Cabal flag management

To change a Cabal flag setting, we can use the command line --flag option. The yackage package has an upload flag that is enabled by default. We can command:

stack build --flag yackage:-upload

This means: when compiling the yackage package, turn off the upload flag (thus the - in -upload). Unlike other tools, Stack is explicit about which package's flag you want to change. It does this for two reasons:

  1. There's no global meaning for Cabal flags, and therefore two packages can use the same flag name for completely different things.
  2. By following this approach, we can avoid unnecessarily recompiling snapshot packages that happen to use a flag that we're using.

You can also change flag values on the command line for extra-dep and snapshot packages. If you do this, that package will automatically be promoted to an extra-dep, since the build plan is different than what the plan snapshot definition would entail.

GHC options

GHC options follow a similar logic as in managing Cabal flags, with a few nuances to adjust for common use cases. Let's consider the command:

stack build --ghc-options="-Wall -Werror"

This will set the -Wall -Werror options for all local targets. Note that this will not affect extra-dep and snapshot packages at all. This design provides us with reproducible and fast builds.

(By the way: the above GHC options have a special convenience flag: --pedantic.)

There's one extra nuance about command line GHC options: Since they only apply to local targets, if you change your local targets, they will no longer apply to other packages. Let's play around with an example from the wai repository, which includes the wai and warp packages, the latter depending on the former. If we command again:

stack build --ghc-options=-O0 wai

It will build all of the dependencies of wai, and then build wai with all optimizations disabled. Now let's add in warp as well. Command:

stack build --ghc-options=-O0 wai warp

This builds the additional dependencies for warp, and then builds warp with optimizations disabled. Importantly: it does not rebuild wai, since wai's configuration has not been altered. Now the surprising case. Command:

stack build --ghc-options=-O0 warp
wai-3.0.3.0-5a49351d03cba6cbaf906972d788e65d: unregistering (flags changed from ["--ghc-options","-O0"] to [])
warp-3.1.3-a91c7c3108f63376877cb3cd5dbe8a7a: unregistering (missing dependencies: wai)
wai-3.0.3.0: configure

You may expect this to be a no-op: neither wai nor warp has changed. However, Stack will instead recompile wai with optimizations enabled again, and then rebuild warp (with optimizations disabled) against this newly built wai. The reason: reproducible builds. If we'd never built wai or warp before, trying to build warp would necessitate building all of its dependencies, and it would do so with default GHC options (optimizations enabled). This dependency would include wai. So when we command:

stack build --ghc-options=-O0 warp

We want its behavior to be unaffected by any previous build steps we took. While this specific corner case does catch people by surprise, the overall goal of reproducible builds is - in the Stack maintainers' views - worth the confusion.

Final point: if you have GHC options that you'll be regularly passing to your packages, you can add them to your stack.yaml file. See the documentation section on ghc-options for more information.

Note

That's it, the heavy content of this guide is done! Everything from here on out is simple explanations of commands. Congratulations!

The stack path command

Generally, you don't need to worry about where Stack stores various files. But some people like to know this stuff. That's when the stack path command is useful. stack path --help explains the available options and, consequently, the output of the command:

--stack-root             Global Stack root directory
--global-config          Global Stack configuration file
--project-root           Project root (derived from stack.yaml file)
--config-location        Configuration location (where the stack.yaml file is)
--bin-path               PATH environment variable
--programs               Install location for GHC and other core tools (see
                         'stack ls tools' command)
--compiler-exe           Compiler binary (e.g. ghc)
--compiler-bin           Directory containing the compiler binary (e.g. ghc)
--compiler-tools-bin     Directory containing binaries specific to a
                         particular compiler (e.g. intero)
--local-bin              Directory where Stack installs executables (e.g.
                         ~/.local/bin (Unix-like OSs) or %APPDATA%\local\bin
                         (Windows))
--extra-include-dirs     Extra include directories
--extra-library-dirs     Extra library directories
--snapshot-pkg-db        Snapshot package database
--local-pkg-db           Local project package database
--global-pkg-db          Global package database
--ghc-package-path       GHC_PACKAGE_PATH environment variable
--snapshot-install-root  Snapshot installation root
--local-install-root     Local project installation root
--snapshot-doc-root      Snapshot documentation root
--local-doc-root         Local project documentation root
--local-hoogle-root      Local project documentation root
--dist-dir               Dist work directory, relative to package directory
--local-hpc-root         Where HPC reports and tix files are stored

In addition, stack path accepts the flags above on the command line to state which keys you're interested in. This can be convenient for scripting. As a simple example, let's find out the sandboxed versions of GHC that Stack installed:

Command:

ls $(stack path --programs)/*.installed
/home/<user_name>/.stack/programs/x86_64-linux/ghc-9.0.2.installed

Command:

dir "$(stack path --programs)/*.installed"

Directory: C:\Users\mikep\AppData\Local\Programs\stack\x86_64-windows

Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
-a---          27/07/2022  5:40 PM              9 ghc-9.0.2.installed
-a---          25/02/2022 11:39 PM              9 msys2-20210604.installed

While we're talking about paths, to wipe our Stack install completely, here's what typically needs to be removed:

  1. the Stack root folder (see stack path --stack-root, before you uninstall);
  2. if different, the folder containing Stack's global YAML configuration file (see stack path --global-config, before you uninstall);
  3. on Windows, the folder containing Stack's tools (see stack path --programs, before you uninstall), which is located outside of the Stack root folder; and
  4. the stack executable file (see which stack, on Unix-like operating systems, or where.exe stack, on Windows).

You may also want to delete .stack-work folders in any Haskell projects that you have built using Stack. The stack uninstall command provides information about how to uninstall Stack.

The stack exec command

We've already used stack exec multiple times in this guide. As you've likely already guessed, it allows you to run executables, but with a slightly modified environment. In particular: stack exec looks for executables on Stack's bin paths, and sets a few additional environment variables (like adding those paths to the PATH, and setting GHC_PACKAGE_PATH, which tells GHC which package databases to use).

If you want to see exactly what the modified environment looks like, try command:

stack exec env

The only issue is how to distinguish flags to be passed to Stack versus those for the underlying program. Thanks to the optparse-applicative library, Stack follows the Unix convention of -- to separate these. For example, command:

stack exec --package stm -- echo I installed the stm package via --package stm

yields output like:

Run from outside a project, using implicit global project config
Using latest snapshot resolver: lts-18.3
Writing global (non-project-specific) config file to: /home/michael/.stack/global/stack.yaml
Note: You can change the snapshot via the resolver field there.
I installed the stm package via --package stm

Flags worth mentioning:

  • --package foo can be used to force a package to be installed before running the given command.
  • --no-ghc-package-path can be used to stop the GHC_PACKAGE_PATH environment variable from being set. Some tools — notably Cabal (the tool) — do not behave well with that variable set.

You may also find it convenient to use stack exec to launch a subshell (substitute bash with your preferred shell) where your compiled executable is available at the front of your PATH. Command:

stack exec bash

The stack ghci or stack repl command

GHCi is the interactive GHC environment, a.k.a. the REPL. You could access it with command:

stack exec ghci

But that won't load up locally written modules for access. For that, use the stack ghci or stack repl commands, which are equivalent. To then load modules from your project in GHCi, use the :module command (:m for short) followed by the module name.

Note

If you have added packages to your project please make sure to mark them as extra deps for faster and reliable usage of stack ghci. Otherwise GHCi may have trouble due to conflicts of compilation flags or having to unnecessarily interpret too many modules. See Stack's project-level configuration to learn how to configure a package as an extra-dep.

For further information, see the REPL environment documentation.

The stack ghc and stack runghc commands

You'll sometimes want to just compile (or run) a single Haskell source file, instead of creating an entire Cabal package for it. You can use stack exec ghc or stack exec runghc for that. As simple helpers, we also provide the stack ghc and stack runghc commands, for these common cases.

Finding project configs, and the implicit global project

Whenever you run something with Stack, it needs a project-level configuration file. The algorithm Stack uses to find such a file is:

  1. Check for a --stack-yaml option on the command line
  2. Check for a STACK_YAML environment variable
  3. Check the current directory and all ancestor directories for a stack.yaml file

The first two provide a convenient method for using an alternate configuration. For example: stack build --stack-yaml stack-ghc-9.2.3.yaml can be used by your CI system to check your code against GHC 9.2.3. Setting the STACK_YAML environment variable can be convenient if you're going to be running commands like stack ghc in other directories, but you want to use the configuration you defined in a specific project.

If Stack does not find a project level configuration file in any of the three specified locations, the implicit global logic kicks in. You've probably noticed that phrase a few times in the output from commands above. Implicit global is essentially a hack to allow Stack to be useful in a non-project setting. When no implicit global configuration file exists, Stack creates one for you with the latest LTS snapshot as the resolver. This allows you to do things like:

  • compile individual files easily with stack ghc
  • build executables without starting a project, e.g. stack install pandoc

Keep in mind that there's nothing magical about this implicit global configuration. It has no effect on projects at all. Every package you install with it is put into isolated databases just like everywhere else. The only magic is that it's the catch-all project whenever you're running Stack somewhere else.

Setting the Stack root location

The Stack root is a directory where Stack stores important files. The location and contents of the directory depend on the operating system and/or whether Stack is configured to use the XDG Base Directory Specification.

The location of the Stack root can be configured by setting the STACK_ROOT environment variable or using Stack's --stack-root option on the command line.

The Stack root contains snapshot packages; tools such as GHC, in a programs directory; and Stack's global YAML configuration file (config.yaml).

The default Stack root is ~/.stack.

The Stack root contains snapshot packages; and Stack's global YAML configuration file (config.yaml). The default location of tools such as GHC and MSYS2 is outside of the Stack root.

The default Stack root is %APPDIR%\stack.

The default location of tools is %LOCALAPPDATA%\Programs\stack.

On Windows, the length of filepaths may be limited (to MAX_PATH), and things can break when this limit is exceeded. Setting a Stack root with a short path to its location (for example, C:\sr) can help.

On Unix-like operating systems and Windows, Stack can be configured to follow the XDG Base Directory Specification if the environment variable STACK_XDG is set to any non-empty value. However, Stack will ignore that configuration if the Stack root location has been set on the command line or the STACK_ROOT environment variable exists.

If Stack is following the XDG Base Directory Specification, the Stack root contains what it would otherwise contain for the operating system, but Stack's global YAML configuration file (config.yaml) may be located elsewhere.

The Stack root is <XDG_DATA_HOME>/stack. If the XDG_DATA_HOME environment variable does not exist, the default is ~/.local/share/stack on Unix-like operating systems and %APPDIR%\stack on Windows.

The location of config.yaml is <XDG_CONFIG_HOME>/stack. If the XDG_CONFIG_HOME environment variable does not exist, the default is ~/.config/stack on Unix-like operating systems and %APPDIR%\stack on Windows.

This approach treats:

  • the project-level YAML configuration file that is common to all projects without another such file in their project directory or its ancestor directories as data rather than as part of Stack's own configuration;

  • the snapshots database as essential data rather than as non-essential data that would be part of a cache, notwithstanding that Stack will rebuild that database as its contents are needed; and

  • the Pantry store as essential data rather than as non-essential data that would be part of a cache, notwithstanding that Stack will download the package index and rebuild the store if it is absent.

The location of the Stack root is reported by command:

stack path --stack-root

The full path of Stack's global YAML configuration file is reported by command:

stack path --global-config

stack.yaml versus Cabal files

Now that we've covered a lot of Stack use cases, this quick summary of stack.yaml versus Cabal files will hopefully make sense and be a good reminder for future uses of Stack:

  • A project can have multiple packages.
  • Each project has a stack.yaml.
  • Each package has a Cabal file, named <package_name>.cabal.
  • The Cabal file specifies which packages are dependencies.
  • The stack.yaml file specifies which packages are available to be used.
  • The Cabal file specifies the components, modules, and build flags provided by a package
  • stack.yaml can override the flag settings for individual packages
  • stack.yaml specifies which packages to include

Comparison to other tools

Stack is not the only tool available for building Haskell code. Stack came into existence due to limitations at that time with some of the existing tools. If you are happily building Haskell code with other tools, you may not need Stack. If you're experiencing problems with other tools, give Stack a try instead.

If you're a new user who has no experience with other tools, we recommend Stack. The defaults match modern best practices in Haskell development, and there are fewer corner cases you need to be aware of. You can develop Haskell code with other tools, but you probably want to spend your time writing code, not convincing a tool to do what you want.

Underlying package format

Before turning to differences, we clarify an important similarity: Stack, Cabal (the tool), and presumably all other tools share the same underlying package format of Cabal (the library). This is a Good Thing: we can share the same set of upstream libraries, and collaboratively work on the same project with Stack, Cabal (the tool), and NixOS. In that sense, we're sharing the same ecosystem.

Curation vs dependency solving

  • Stack uses 'curation' (snapshots and Stack's project-level configuration file (stack.yaml) define precisely the set of packages available for a project). The Stack team firmly believes that the majority of users want to simply ignore dependency resolution nightmares and get a valid build plan from day one. That's why we've made 'curation' the focus of Stack.

  • Cabal (the tool) can use 'curation' too but its origins are in dependency solving.

Emphasis on reproducibility

  • Stack goes to great lengths to ensure that stack build today does the same thing tomorrow. With Stack, changing the build plan is always an explicit decision.

  • Cabal (the tool) does not go to the same lengths: build plans can be affected by the presence of pre-installed packages, and running cabal update can cause a previously successful build to fail.

Automatic building of dependencies

  • Stack's automatically builds dependencies. So for example, in Stack, stack test does the same job as:

    cabal install --enable-tests --only-dependencies
cabal configure --enable-tests
cabal build
cabal test

    (newer versions of Cabal (the tool) may make this command sequence shorter).

  • With Cabal (the tool), you need to use cabal install to trigger dependency building. This is somewhat necessary as building dependencies can, in some cases, break existing installed packages.

Isolation

  • Stack is isolated - provides 'sandboxed' behaviour - by default, via its databases. In other words: when you use Stack, there's no need for sandboxes, everything is (essentially) sandboxed by default.

  • With Cabal (the tool), the default behavior is a non-isolated build where working on two projects can cause the user package database to become corrupted. The Cabal solution to this is sandboxes.

Tools other than Stack and Cabal (the tool)

  • cabal-meta inspired a lot of the multi-package functionality of Stack. Still relevant for Cabal (the tool).
  • cabal-src. Deprecated in favor of Stack in 2016.
  • stackage-cli.Deprecated in favor of Stack in 2015.
  • cabal-dev. Deprecated in favor of Cabal (the tool) in 2013.