Installing on macOS#
Weblate should run on any contemporary hardware without problems, the following is the minimal configuration required to run Weblate on a single host (Weblate, database and webserver):
3 GB of RAM
2 CPU cores
1 GB of storage space
The more memory the better - it is used for caching on all levels (filesystem, database and Weblate).
Many concurrent users increases the amount of needed CPU cores. For hundreds of translation components at least 4 GB of RAM is recommended.
The typical database storage usage is around 300 MB per 1 million hosted words. Storage space needed for cloned repositories varies, but Weblate tries to keep their size minimal by doing shallow clones.
Actual requirements for your installation of Weblate vary heavily based on the size of the translations managed in it.
Install the dependencies needed to build the Python modules (see Software requirements):
brew install python pango cairo gobject-introspection glib libyaml firstname.lastname@example.org pip install virtualenv
Using older libxmlsec is needed until https://github.com/xmlsec/python-xmlsec/issues/254 is addressed.
Optionally install software for running production server, see Running server, Database setup for Weblate, Background tasks using Celery. Depending on size of your installation you might want to run these components on dedicated servers.
The local installation instructions:
# Web server option 1: NGINX and uWSGI brew install nginx uwsgi # Web server option 2: Apache with ``mod_wsgi`` brew install httpd # Caching backend: Redis brew install redis # Database server: PostgreSQL brew install postgresql # Gettext for the msgmerge add-on brew install gettext
We’re using virtualenv to install Weblate in a separate environment from your system. If you are not familiar with it, check virtualenv User Guide.
Create the virtualenv for Weblate:
Activate the virtualenv for Weblate:
Install Weblate including all optional dependencies:
# Install Weblate with all optional dependencies pip install "Weblate[all]"
Please check Optional dependencies for fine-tuning of optional dependencies.
On some Linux distributions running Weblate fails with libffi error:
ffi_prep_closure(): bad user_data (it seems that the version of the libffi library seen at runtime is different from the 'ffi.h' file seen at compile-time)
This is caused by incompatibility of binary packages distributed via PyPI with the distribution. To address this, you need to rebuild the package on your system:
pip install --force-reinstall --no-binary :all: cffi
The following assumes the virtualenv used by Weblate is activated
. ~/weblate-env/bin/activate). If not, specify the full path
to the weblate command as
Copy the file
Adjust the values in the new
settings.pyfile to your liking. You will need to provide at least the database credentials and Django secret key, but you will want more changes for production setup, see Adjusting configuration.
Create the database and its structure for Weblate (the example settings use PostgreSQL, check Database setup for Weblate for a production-ready setup):
Create an administrator user account
admin, generate its password, and copy it to the clipboard; remember to save it for later use:
If you previously missed/lost the admin password, you can generate a new one with the following command:
weblate createadmin --update
Start the Celery workers. This is not necessary for development purposes, but strongly recommended otherwise. Background tasks using Celery has more info:
Start the development server (Running server details a production setup):
Congratulations, your Weblate server is now running and you can start using it.
You can now access Weblate on
Sign in with admin credentials obtained during installation or register with new users.
You can now run Weblate commands using weblate command when Weblate virtualenv is active, see Management commands.
You can stop the test server with Ctrl+C.
Open the admin interface (
http://localhost:8000/create/project/) and create the project you want to translate. See Project configuration for more details.
All you need to specify here is the project name and its website.
Create a component which is the real object for translation - it points to the VCS repository, and selects which files to translate. See Component configuration for more details.
The important fields here are: Component name, Source code repository, and File mask for finding translatable files. Weblate supports a wide range of formats including GNU gettext, Android string resources, Apple iOS strings, Java properties, Stringsdict format or Fluent format, see Supported file formats for more details.
Once the above is completed (it can be lengthy process depending on the size of your VCS repository, and number of messages to translate), you can start translating.