mqttk 1.3.1

A lightweight MQTT client GUI written in pure python

• Introduction
• Features
  • Connection profile management
  • Subscribe interface
  • Publish interface
  • Topic browser
  • Broker stats
  • Log tab
  • Import MQTT.fx configuration
• Planned features
  • V1.4
  • V1.5
• Software dependencies
• Installation
  • macOS - app
  • On macOS from source
    • macOS Dependencies
    • macOS - acquiring and installing the package from source
    • macOS - installing via pip
    • macOS - running MQTTk
  • Windows - as an executable
  • Windows - from source
    • Windows - dependencies
    • Windows - acquiring and installing the package from source
    • Windows - installing it via pip
    • Windows - running MQTTk
  • On Linux from source
    • Linux - dependencies
    • Linux - acquiring and installing the package from source
    • Linux - installing it via pip
    • Linux - Running MQTTk
• Building the app from source
  • pypi package
  • macOS appimage
    • Dependencies
    • Building the macOS app
  • Windows executable
• How to contribute
  • Reporting bugs
  • macOS universal2 appimage
  • Linux binary package or app

Introduction

MQTTk is a very lightweight MQTT GUI client that looks retarded, but it does the job fast in a native fashion, without bloated and sluggish browser, java and javascript based rubbish that may look good, but are a pain to use especially in a professional environment.

It intends to replicate most features and functionality of MQTT.fx which is no longer free and the free version is no longer maintained. Since upgrading my computer, it was crashing every minute, practically becoming useless. I always found it more useful than other MQTT GUI clients, which mostly update the values of topics as they come in, in my work, being able to track message exchange over time is as important as the content of the messages themselves.

Since I couldn't find a similar tool out there, I decided to make my own and share it with whoever is interested. The project is written in Tk/ttk. I don't have time to learn some fancy-pancy GUI environment, it was quick and easy to knock out, and it should run on anything including the kitchen sink without too much pain.

Features

Connection profile management

MQTTk allows the user to create and manage multiple connection profiles. For each connection profile, the broker configuration, the topics that have been subscribed to along with the associated colour, the topics in which messages have been published and the message templates are saved. From these connection profiles, the broker connection configuration and the associated subscribe/publish history and message templates can be exported and imported separately.

Once a connection has been configured, it can be connected to. Upon successful connection, the different interfaces for subscription, publish and topic inspection become available.

The configuration files and logs are saved in the following locations:

macOS: ~/Library/ApplicationSupport/MQTTk/

Windows: %LOCALAPPDATA/MQTTk/

Linux: ~/.config/MQTTk/

Configuration interface

Export subscribe/publish history interface

Subscribe interface

On the subscribe tab, topics can be subscribed to. The $SYS topics and both the # and + wildcards are supported. Once subscribed, the messages arriving in the topic(s) are listed in the listbox with the time of their arrival, topic, QOS, retained state and their ID in MQTTk. Topics and topic patterns subscribed to get a colour assigned to them, messages arriving in these topis appear in the colour associated to the topic pattern. The colour can be changed on the fly and the new colour gets applied to all previous messages that arrived in these topics. Activating the Autoscroll checkbox will cause the last message that arrived to be selected automatically and its details to be shown immediately in the message details section of the interface. Topic subscriptions can be temporarily muted using the Mute checkbox on the subscription widget.

Once messages arrived, they can be selected in the listbox. Selected message details appear in the lower right part of the interface. Here, different decoders are available to quickly decode or format the most common message types in the message content textbox. So far, a JSON pretty formatter and a hex decoder are available, but decoders can be added in the future on demand.

The Attempt to decompress option will try to decompress the payload using the most common compression algorithms (currently zlib and bz2 are supported, but these can be extended in the future).

Messages that have arrived, can be exported in .CSV and .JSON formats. Message payload is exported as unicode text if possible, otherwise it is encoded in base64.

Subscribe interface

Publish interface

On this interface, messages can be published. Once a topic is input, a message payload can be inserted, the QoS of the message selected and if needed be, the message can be made retained. Once a message is published, the topic will be saved in the topic drop-down for future use. If a message or a payload is needed often, the message can be saved as a template with a custom name, and published by just a click of the publish button on the message template widget. Selecting the message template will fill the relevant fields on the interface so the content can be modified and/or saved as a new template.

Publish interface

Topic browser

The topic browser allows to subscribe to a topic pattern and organises all incoming messages in a tree format, split by the / in message the topic. The most important message information (time of arrival of the last message, QoS, retained status, payload) are also shown. The message payload is decoded into a string if possible, otherwise it remains the bytestring as it arrived. Right clicking on the selected message allows the topic and the payload to be copied on the clipboard. The Ignore retained messages option will ignore all retained messages, only freshly arrived messages will make it into the topic browser.

Topic browser

Broker stats

This allows the statistics of the broker connected to, to be viewed in a similar fashion as the topic browser.

Broker stats

Log tab

The log may contain useful information in case something isn't working with the app as expected. The log is also output in a file, which is in the same directory as the configuration files. The log tab text will change to * Log * when error or exception level messages get inserted to it, to indicate an issue. Upon clicking on the tab, the text returns to normal.

Import MQTT.fx configuration

If MQTT.fx was already installed on the computer, the "MQTT.fx config" option in the "Import" will try to find and import it. If MQTTk cannot find it, the file can also be browsed for. This feature has only been tested with my MQTT.fx configs and although it worked, there may be config files out there that may fail to import.

Planned features

V1.4

• Option to turn on notifications when new messages arrive
• Notifications only for specific subscriptions

V1.5

• option to encrypt the configuration file and decrypt it at application launch or use an alternative unencrypted config in the current session
• option to encrypt exported broker configuration

Software dependencies

The project is written in pure python, powered by the below projects:

• python3.7+
• Tkinter/ttk
• Eclipse paho-mqtt python client
• xmltodict

That't it, nothing fancy. Give the above projects a big thumbs up!

Installation

macOS - app

:warning: The built appimage is experimental! Use it at your own risk, it may cause kernel panics and app crashes.

:warning: The app image is only for M1 (ARM64, Apple Silicon) macs! With my current knowledge I can't produce a working intel or universal2 app image.

Download the latest release from the GitHub releases page and install it like any other apps.

The system may complain about not being able to verify the developer. You can find more information about me here, so you can verify the developer yourself instead. To run the app, follow the instructions provided by Apple here.

On macOS from source

macOS Dependencies

You must have Python 3.7+ and pip installed. On some versions of macOS or the python package, Tk/ttk is not included, in which case the python-tk package is also needed.

The easiest way to install these, is to use homebrew. The commands below may be different on your system. Open the terminal and issue these commands:

$ brew install python python-tk

:warning: When installing/running the app, use the system interpreter, or the interpreter available via homebrew. During my testing, Conda or other interpreters occasionally caused my system to crash entirely (kernel panic) and when using mission control to switch between apps, MQTTk crashed regularly. The cause is outside the code of this software. This crash happens under certain circumstances when switching to the app via mission control or the dock, there's nothing I can do about it, unfortunately. Therefore, use other interpreters at your own risk!

macOS - acquiring and installing the package from source

Download the latest release from the GitHub repository and install it using pip.

$ pip3 install mqttk-x.y.tar.gz

macOS - installing via pip

Issue the following command:

$ pip3 install mqttk

macOS - running MQTTk

To run the software, just issue the mqttk command from the terminal.

$ mqttk

If the app fails to start, or crashes randomly, try re-launching it using the

$ mqttk-console

command. This will leave a console window, which might provide additional debug information when something goes tits up.

Windows - as an executable

Download the latest release from the GitHub repository and install/run like any other apps.

Windows - from source

Windows - dependencies

Download python3 from the official website and install it like any other apps.

Windows - acquiring and installing the package from source

Download the latest release from the GitHub repository and then install it using pip and the command line:

> pip3 install mqttk-x.y.tar.gz

Windows - installing it via pip

Issue the following command in the command line:

> pip3 install mqttk

Windows - running MQTTk

In the command line issue the command:

> mqttk

If the app fails to start, or crashes randomly, try re-launching it using the

$ mqttk-console

command. This will leave a console window, which might provide additional debug information when something goes tits up.

On Linux from source

Linux - dependencies

You need to install python3, python3-pip and in some cases the python3-tk packages. The process will be different depending on your distribution, refer to your distributions package manager or community. As an example, on ubuntu you'd need to install the below packages using apt. On other distros, python3 might be default, in which case the "3" suffix won't be needed on the packages.

$ sudo apt install python3 python3-pip python3-tk

Linux - acquiring and installing the package from source

Download the latest release from the GitHub repository and install it using pip.

$ pip3 install mqttk-x.y.tar.gz

Linux - installing it via pip

Just issue the command

$ pip3 install mqttk

Linux - Running MQTTk

From the command line issue the command

$ mqttk

If the app fails to start, or crashes randomly, try re-launching it using the

$ mqttk-console

command. This will leave a console window, which might provide additional debug information when something goes tits up.

Building the app from source

pypi package

issue the following command in the project root to build the sdist package.

$ python3 setup.py sdist

The built package will appear in the dist/ directory.

macOS appimage

:warning: This is highly experimental and needs refiniement!

:warning: When building the app, use the system interpreter, or the interpreter available via homebrew.

Conda or other interpreters can cause your system to crash entirely (kernel panic) which issue is outside of the code of this software. This crash happens under certain circumstances when switching to the app via mission control or the dock. Use other interpreters at your own risk!

Dependencies

You need to have xcode installed. Use the app store to do that if you don't have it yet. You will also need the xcode command line tools to be installed. You can do that from the terminal:

$ xcode-select —install

Just like when running the app, you need python3, pip and python3-tk. Install these as explained above.

In addition, you need the pyinstaller package. Install it using pip:

$ pip install pyinstaller

Building the macOS app

I was not able to build a universal app image for MACs that ran native on both Intel and M1 architectures, so I only built the ARM64 package.

Navigate to the project root and issue

$  pyinstaller mqttk.spec

Windows executable

Just like when running the app, you need python3, pip and python3-tk. Install these as explained above. In addition, you need the pyinstaller package, use the command line:

> pip3 install pyinstaller

Navigate to the project root and issue the following command:

> pyinstaller mqttk.spec

How to contribute

Reporting bugs

Use the GitHub issue reporting page of the project to help me squish bugs.

macOS universal2 appimage

My time and knowledge is limited to figure how to properly build a universal2 app image (intel + ARM). I managed to build an M1 only version, with which I'm not entirely happy, it takes a long time to start up for some reason. Furthermore, I had issues with the app when not running on the system interpreter on my M1 mac, causing regular crashes and kernel panics when switching to and from MQTTk. I would appreciate help with building the app and testing the resulting image out on other machines.

Linux binary package or app

There are more ways to distribute apps on various linux distros than stars on the sky. I'd appreciate recommendations on what format to use and maybe a helping hand figuring it out and getting things set up.