Desktop Application for
Branch: master
Clone or download
Jikstra Prepare for v0.101.0
Latest commit 5cdb52f Feb 15, 2019


Desktop Application for

Build Status JavaScript Style Guide

If you are upgrading: please see

Table of Contents

Click to expand


The application can be downloaded from the Releases page. Here you'll find prebuilt releases for all supported platforms. See below for platform specific instructions. If you run into any problems please consult the Troubleshooting section below.



AppImages are a generic way to install software accross most linux distributions and systems.

To install a .AppImage based release:

  • Download the binary
  • Make it executable, e.g. chmod u+x ~/Downloads/deltachat-desktop-x.y.z-x86_64.AppImage
  • Executing the .AppImage will install it on the system in /opt/DeltaChat


  • Click on the link for the .deb file
  • Some systems enable installing it directly by clicking
  • If your system doesn't handle .deb files you can install manually by doing e.g. sudo dpkg -i ~/Downloads/deltachat-desktop_x.y.z_amd64.deb

Arch Linux

WARNING: Currently the AUR package compiles from latest master. This can be more recent as the latest release, introduce new features but also new bugs.

If you have a AUR helper like yay installed, you can install it by running yay -S deltachat-desktop-git and following the instruction in your terminal.

Otherwise you can still do it manually:

# Download the latest snapshot of the PKGBUILD
# extract the archive and rm the archive file afterwards
tar xzfv deltachat-desktop-git.tar.gz && rm deltachat-desktop-git.tar.gz
# cd into extracted folder
cd deltachat-desktop-git
# build package
makepkg -si
# install package (you need to replace <version> with whatever version makepkg built)
sudo pacman -U deltachat-desktop-git-<version>.tar.xz

Mac OS

Simply install the .dmg file as you do it with all other software on mac.

If you are getting an openssl error message at the first start up you need to install openssl.

$ brew install openssl

From Source

Get the code:

$ git clone
$ cd deltachat-desktop

Install dependencies:

$ npm install

Build the app (only needed if the code has changed or if the app has never been built before):

$ npm run build

Start the application:

$ npm start

Configuration and Databases

The configuration files and database are stored at application-config's default filepaths.

Each database is a sqlite file that represents the account for a given email address.


This module builds on top of deltachat-core, which in turn has external dependencies. Instructions below assumes a Linux system (e.g. Ubuntu 18.10).

If you get errors when running npm install, they might be related to the build dependencies meson and ninja.

If meson is missing:

sudo apt-get install python3-pip
sudo pip3 install meson

If ninja is missing:

sudo apt-get install ninja-build

You might also need the following system dependencies:

  • libssl-dev
  • libsasl2-dev
  • libsqlite3-dev
  • zlib1g-dev

To fix these issues do:

sudo apt-get install libssl-dev libsasl2-dev libsqlite3-dev zlib1g-dev

Then try running npm install again.

Please see build instructions for additional information.


Logging Options

Debug messages are disabled by default, enable them with the --log-debug flag.

Flag Effect
--log-debug Log debug messages
--log-to-console Output the log to stout / chrome dev console

Log locations

The logs can be found in:

Linux: ~/.config/DeltaChat/logs/
Mac: ~/Library/Application Support/DeltaChat/logs

You can also access the log folder and the current log file under the View->Developer menu:


The log files have the extension .log, the file name represents the point in time the log started. Basically the log files are tab separated csv-files:

"2019-01-27T13:46:31.801Z"	"main/deltachat"	"INFO"	[]	"dc_get_info"
timestamp location / module level stacktrace arg1 arg2 ...
"2019-01-27T13:46:31.801Z" "main/deltachat" "INFO" [] "dc_get_info" - ...

How to Contribute

Code Structure

Some important folders and files:

├── bin                       # various helper scripts
├── build                     # files needed only at build time
├── css                       # styelsheets which need preprocessing
│   └── conversations         # stylesheets pulled out from signal
├── images                    # image files used in conversations
├── index.js                  # entry point for the main process
├── jenkins                   # pipelines for building on Jenkins
├── _locales                  # translation files in xml and json
│   ├── _experimental-en.json # can contain experimental language strings
│   └── languages.json        # central file which keeps the human readable language names
├── src
│   ├── main                  # javascript for the main process
│   └── renderer              # javascript for the renderer process
├── static
│   ├── bundle.js             # javascript bundle built by webpack
│   └── conversations.css     # css bundle built from conversations scss files
│   ├── fonts                 # fonts
│   ├── main.css              # main css file
│   └── main.html             # main html file in renderer process
├── test
│   ├── integration           # integration tests
│   └── unit                  # unit tests
├── .travis.yml               # build script for Travis
├── .tx                       # configuration for Transifex
└── webpack.config.js         # configuration for webpack

Run the Code

While developing the following command will build the app and start electron in debug mode with http cache disabled:

$ npm run dev

It's also handy to run this watch command in a separate terminal

$ npm run watch

Add experimental language strings

Sometimes you need to add new language strings, but don't want to push them to transifex immediately because it's unsure if the string will be adjusted in the short future or it's unclear if the pr will even get merged or you simply don't have push rights to the transifex language repo. To still be able to implement new language strings, you can add them to the _locales/_experimental-en.json file. You can also overload every other language string if you need to. The syntax is the exact same as for all other _locales/*.json files.

Example: {"foobar_desktop": {"message": "This is a test"}}


Running npm test does the following:

  • runs standard as code linter
  • runs the unit tests

Running npm run test-integaration executes the integration tests.

The integration tests use spectron and tape. They click through the app, taking screenshots and comparing each one to a reference. Why screenshots?

  • Ad-hoc checking makes the tests a lot more work to write
  • Even diffing the whole HTML is not as thorough as screenshot diffing. For example, it wouldn't catch an bug where hitting ESC from a video doesn't correctly restore window size.
  • Chrome's own integration tests use screenshot diffing iirc
  • Small UI changes will break a few tests, but the fix is as easy as deleting the offending screenshots and running the tests, which will recreate them with the new look.
  • The resulting Github PR will then show, pixel by pixel, the exact UI changes that were made! See

For MacOS, you'll need a Retina screen for the integration tests to pass. Your screen should have the same resolution as a 2016 12" Macbook.

For Windows, you'll need Windows 10 with a 1366x768 screen.

When running integration tests, keep the mouse on the edge of the screen and don't touch the mouse or keyboard while the tests are running.


Install the transifex-client and get added to the Delta Chat App project.

And periodically we can run the following command to get the new translation strings from translators:

tx pull --all

Note that this command updated _locales/*.xml. Run the following command to convert from xml to json:

npm run build-translations

When you need to modify language strings, this should be done in _locales/en.xml. Run the following command to sync with Transifex:

tx push --source


For Continuous Integration we currently use both Travis and Jenkins. Travis is used for Mac and Jenkins for Linux. Once we support Windows we will most likely use Travis for Windows.

Deploy Workflow

  1. Create a draft release on github, e.g. vX.Y.Z
  2. Change version field in package.json to X.Y.Z
  3. Commit and push modified package.json (repeat until release is ready)
  4. Once done, publish the release on github, which will create the tag

Also see


Licensed under the GPLv3, see LICENSE file for details.

Copyright © 2019 DeltaChat contributors.