| 1. Unit testing for individual or a small set of classes
2. Integration testing in a simulated environment
3. Integration testing in a real time environment
4. System testing in a simulated environment |
Many product specific targets for simulated and real time environment |
1. C++ editor, compiler and debugger
2. Matlab / Simulink |
All pull requests to master branch must be approved by one colleague |
11-30 seconds |
20% |
60% |
20% |
Show
|
| depends on the project. Hardly even pure TDD, a bit of mixture of soft TDD and test created after coding. Both unit and integration tests on PC, the same on HW + System test. |
Embedded (ARM, etc.) |
All sort of Editors (depends on client) |
used a number of tools |
1-5 minutes |
40 |
30 |
30 |
Show
|
| Unit tests with TDD
Host integration test
Target end to end tests |
Embedded systems, some running CentOS or Debian Linux, some bare metal. C and C++ applications |
CLion, CMake, Git, Jenkins, Docker |
Not required very often as dev is done using pair programming. Reviewed by senior devs during dev. |
5-30 minutes |
10 |
40 |
50 |
Show
|
| Catch2 unit test, end to end test using python |
varies |
CLion, Git, Gitlab |
continuous pair programming |
11-30 seconds |
33 |
33 |
34 |
Show
|
| Post-hoc with heavy mocking |
STM32 series microcontrollers |
CMake, GCC, CLion, cpplint, sonarqube, gitlab |
Two mandatory reviews per pull request, one dev, one QA |
11-30 seconds |
5 |
3 |
2 |
Show
|
| Unit and integration tests in Catch2
End to End tests running on hardware in Pytest |
STM32 microcontroller |
Clion
STprogrammer and STlink |
Minimal due to heavy use of pair programming. However diffs are reviewed between releases. |
11-30 seconds |
20 |
30 |
50 |
Show
|
| Combination of various ad hoc |
Embedded C on AMR M4 |
Eclipse |
Doesn't exist |
Under 10 seconds |
70 |
10 |
20 |
Show
|
| Catch2 with google mock |
Linux |
GCC |
use gitlab |
31-60 seconds |
40 |
40 |
20 |
Show
|
| Always test behavior in system after code is written. Varies depending on function, but may include using printf statements or debugger breakpoints. |
Asset tracking device using ARM Cortex MCU |
GNU C Compiler, GDB debugger |
We use a web-based code reviewer to perform peer reviews before every commit to our trunk. |
1-5 minutes |
60 |
25 |
15 |
Show
|
| unit tests, integration tests, e2e tests, performance tests |
N/A |
Git, Jenkins, just, GitHub Actions |
We do not often have code reviews, we use pair programming |
Under 10 seconds |
20 |
60 |
20 |
Show
|
| Unit test, manual testing with hardware, automated tests (unit tests and hw tests) |
Embedded system with ultrascale soc and different peripherals. |
Git, ide, debugger sw, confluence, build tools |
We have code reviews when pull requests are being done. |
5-30 minutes |
20% |
30% |
50% |
Show
|
| Functional Testing. |
WPF Apps, and Embedded STM32 Devices. |
GitHub, Jira, Microsoft Visual Studio, STM32CubeIDE |
Short pull request reviews for minor changes, and a larger system review prior to first release. |
11-30 seconds |
50 |
30 |
20 |
Show
|
| breakpoints; set up constants in code |
windows |
GitLab, Jira |
very unofficial reviews between peers |
1-5 minutes |
50 |
20 |
30 |
Show
|
| CppUTest |
STM32xx w/ on-target tests in IAR and GCC + on-host tests w/ 32-bit windows GCC build |
IAR, GCC |
Pull request + CI automation checks |
1-5 minutes |
30 |
30 |
40 |
Show
|
| Off-target unit-testing, on-target HIL tests. |
Distributed system with internal interfaces (UART), external Debug CLI. hard realtime. |
IDE
Code coverage
Lint
|
Ad-hoc and retrospective. |
30-60 minutes |
40 |
10 |
50 |
Show
|
| build then test on target *_*, if it fails debug then repeat |
MCU based target |
IDE provided by MCU manufacturer |
there are almost no peer review, if I found something that may cause a problem I will fix it. |
5-30 minutes |
30% |
10% |
60% |
Show
|
| By writing unit and integration tests |
Don't have one.
It can change at any point. |
IDEs and Code Editors.
Version Control.
Package and dependency managers.
Debuggers
|
They go okay |
31-60 seconds |
40 |
50 |
10 |
Show
|
| Java unit tests. Build C++ code for some testing |
Java in a Linux environment |
IntelliJ, Qt |
Essentially non-existent, we don't have the time. |
31-60 seconds |
30 |
20 |
50 |
Show
|
| Debugger |
? |
Keil MDK ARM, Visual Studio |
No review |
Under 10 seconds |
5 |
5 |
5 |
Show
|
| Custom test harness. |
hetero multicore SoC |
Linux tool chains |
Ad hoc |
31-60 seconds |
5 |
0 |
15 |
Show
|
| Legacy code. Only perform UAT for any custom changes made to the code. Tried to implement Unit testing framework unity to the legacy code but didn't go well. |
We use 8bit Pic Microcontroller |
MPLAB X |
nil |
30-60 minutes |
5/10 |
3/10 |
2/10 |
Show
|
| Sandbox functions - test functions by writing my own unit tests |
IoT systems |
MPLabX, Cypress Creator (PSOC), Simplicity Studio, Source meters, DMM, |
Review changes with other programmers. Categorize into low, med & high risk changes |
Under 10 seconds |
30 |
40 |
30 |
Show
|
| CppUTest |
AMD X86 embedded, Arm V7 & V8, Alpine musl for static builds |
Make, CMake, VS Code, Sublime Text |
Unless I see a major reason for a change I generally only ask probing questions make recommendations |
31-60 seconds |
7 |
1 |
2 |
Show
|
| Unit and integration |
Embedded PIC16,PIC32 and ESP32 |
Notepad++ |
What code review? |
1-5 minutes |
80 |
10 |
10 |
Show
|
| TDD, testing locally before pushing to CI.
Tests run in CI using Jenkins pipelines. |
Data diodes and their surrounding software for customer ease of use. |
Nvim, GitLab, Jenkins, Docker, Jira |
We use pair programming for the purposes of code reviews. |
1-5 minutes |
10 |
10 |
80 |
Show
|
| Tdd mostly since last year, mncks, DLP |
Embedded Xilinx devices |
Qemu, libcheck, mocks |
Deep code review, understanding of requirements |
Under 10 seconds |
50 |
40 |
10 |
Show
|
| Rust unit test, Boost Test |
Linux based embedded system |
Rust standard tools, c/c++ linter etc |
1) Does it follow the standard 2) Understand the problem 3) Evaluate the proposed solution |
5-30 minutes |
40 |
20 |
40 |
Show
|
| Test to functional specifications and ensure the device functions as intended. Additionally some longer term testing is setup but its all quite unstructured. |
Battery powered small embedded systems. Usually containing a sensor and RF |
STM IDE |
Every pull request to a develop branch or to release goes through a pull request and is reviewed. |
11-30 seconds |
20 |
50 |
30 |
Show
|
| unit test |
Multiple peripheral system |
STM32 CUBE IDE |
2 CYCLE REVIEW |
Under 10 seconds |
30 |
30 |
40 |
Show
|
| TDD frame work |
stm32f427 |
STM32CUBEIDE |
no |
5-30 minutes |
50 |
30 |
20 |
Show
|
| With IDE. |
STM32 Microcontroller. |
STM32CubeIDE |
I don't have a code review experience. |
1 day or more |
10 |
10 |
80 |
Show
|
| Debugging tools |
STM32 and ESP32, we will have some other chips in the future. |
STM32CubeIDE
Visual Studio Code and a variety of plugins |
At work they have been minimal I hope this course helps us establish better practices. |
31-60 seconds |
20 |
40 |
40 |
Show
|
| Debugging on embedded hardware |
STM32 |
STM32CubeIDE, STM32CubeProgrammer, STM32CubePowerMonitor |
Peer reviews on pull request |
Under 10 seconds |
40 |
10 |
50 |
Show
|
| Manual testing |
STM32 |
STMCube IDE |
We dont have that process setup yet |
31-60 seconds |
4 |
2 |
4 |
Show
|
| I attempt to do TDD, but sometimes revert back to not writing tests (particularly when I'm rushing) |
Desktop, Web, and Embedded Cortex M processor for an IoT application |
Simplicity Studio |
Every feature change goes through code reviews before merging with the main code branch |
1-5 minutes |
50 |
20 |
30 |
Show
|
| Limited independent peer review, personal stress testing for edge conditions and critical logical conditions |
Not sure what is meant here. I don't have any specific procedure. |
Stock IDE tools for code generation and basic debugging. |
Informal and without a written description. Reviews performed by developers of the code. |
Under 10 seconds |
25 |
25 |
50 |
Show
|
| Functional tests as I'm programming |
Linux |
VS Code
GCC
Docker
Git |
We don't have a lot of them, and don't have a process established yet. |
5-30 minutes |
50 |
25 |
25 |
Show
|
| On Device Black box testing |
Multiple ARM 32-bit systems. One multi core WiFi/BLE, one BLE, one low level hardware driver. |
VS Code
GDB
Various hardware debug tools
|
One Person Firmware team. Self Review and sometimes AI tools |
31-60 seconds |
75 |
10 |
15 |
Show
|
| pytest + python + powershell |
Arm Cortex M7, comms mostly via CAN |
VS Code
MCUXpresso
Segger J-Link
Picoscope
|
non exist |
31-60 seconds |
30 |
40 |
30 |
Show
|
| Unit testing |
Armv7 SoC |
VS Code + language-specific extensions; git, Jira, valgrind |
They depend on who is doing the reviewing. |
1-5 minutes |
30 |
40 |
30 |
Show
|
| Simple running and debugging the code |
I can't say for a certain |
VS Code for coding |
Just reading the code and test it |
1-2 hours |
20 |
30 |
50 |
Show
|
| My role is not currently development, as I am mainly on Support and with some devops. I write code occasionally for demostrations an recreating customers issues.
I normally write test cases. |
No target on this |
VS Code or JetBrains Tools. I have used others in the past. |
No comment |
Under 10 seconds |
60 |
20 |
20 |
Show
|
| Unit testing, Integration Testing, End-to-End Testing |
Windows/AMD64 |
VS Code, CMake, MSVC BuildTools, GoogleTest, Git, OpenCppCoverage, BitBucket |
2-person review on pull-requests |
1-5 minutes |
40 |
20 |
10 |
Show
|
| In simulator with JSON based inputs. |
Embedded system running embedded software. |
VS Code, Git etc. |
Code is thoroughly tested and reviewed. |
5-30 minutes |
45% |
40% |
15% |
Show
|
| unit tests, script/automation tests and manual tests |
embedded systems running Linux |
VS Code, Pycharm
Colaboration tools: Confluence, Jira, Bitbucket, Jenkins |
mandatory in Bitbucket pull-requests |
30-60 minutes |
50% |
30% |
20% |
Show
|
| Unit tests and integration tests run in CI/CD pipeline |
Linux, running as a controller for attached devices |
VS code |
Reviews are done by team members before any code is merged into the development branch. |
30-60 minutes |
4 |
2 |
4 |
Show
|
| NUnit |
WINCE, moving to Linux later this year
ARM based |
VS, VScode, Resharper |
Jira ticket -> Dev-branch -> PR -> CodeReview |
11-30 seconds |
40 |
50 |
10 |
Show
|
| Writing unit tests, integration tests and end-to-end tests |
N/A |
VSCode |
Talking through with another developer |
11-30 seconds |
7 |
1 |
2 |
Show
|
| Yes |
Embedded Linux |
VSCode |
Bitbucket is used for code reviews. |
Under 10 seconds |
6 |
2 |
2 |
Show
|
| Manual functional / behavioral test suite
I started at Whisker as a FW QA Engineer in July, I have become familiar with Ceedling and am in process of adding unit test capabilities to FW team. |
Main MCU is PIC but is being phased out with STM32, all targets have companion ESP32 for Wi-Fi. |
VSCode
Espressif's ESP-IDF
STM32CubeIDE |
Have not participated in one, but have recently set forth a plan to make this a regular occurance. |
1-5 minutes |
n/a |
n/a |
n/a |
Show
|
| Write the code and if time permits write unit tests.
Then update integration tests to make sure code base works as a whole.
Run robot framework on the output to make sure the behaviour is correct.
QA manually sign off that everything is working. |
The target is docker containers which run the data pipeline |
VSCode with Python Extensions
Copilot |
Done through the PR interface. Usually offline but if it is complex, done on a call |
Under 10 seconds |
30 |
10 |
60 |
Show
|
| For C++ code Cppunit and using mocks for hardware drivers. For FPGA I use System Verilog with assertions to implement TDD to verify behaviours to reduce the risk of regressions when making changes. |
Almost all my projects have an FPGA and chips around them that we interface to. |
VSCode with TerosHDL for FPGA work and VSCode with C/C++ extensions for linting/code completion etc. |
Generally I like to use checklists and have reviewers read standards beforehand. |
1-5 minutes |
2 |
4 |
4 |
Show
|
| QA Feature regession testing for functionality. Firmware engineer edgecase testing. |
ESP32 WROOM 32E SoC |
VSCode with extensions for C and python |
Currently we have 1 other FW engineer to a Pull Request |
1-5 minutes |
30 |
50 |
20 |
Show
|
| Unit testing with CppUTest before merging and periodically performing integration tests |
Testing features of medical devices |
VSCode, IAR, QtCreator, Bitbucket |
Reviews completed before merging code |
1-5 minutes |
35% |
25% |
40% |
Show
|
| CppUTest |
custom embedded platform (Arm and x86) |
VSCode, Makefiles, git, Jenkins |
PRs are reviewed before merging to main. Who reviews is chosen by the PR author. |
31-60 seconds |
20% |
5% |
10% |
Show
|
| BDD, Manual QA, Black Box, Automation Tests of functionality, Integration testing, System Testing, Regression testing. |
Any complex software application |
VSCode, Notepad++, SQL, Docker, Cucumber, |
I don't do any. |
11-30 seconds |
1 |
7 |
2 |
Show
|
| Automated test cases |
Telecom optical equipment |
VSCode, Vim |
Constructive feedback will be provided and most of the corner cases will be evaluated |
1-5 minutes |
60 |
25 |
15 |
Show
|
| A mix of unit tests and manual testing |
- |
VSCode, Visual Studio |
- |
11-30 seconds |
4 |
2 |
4 |
Show
|
| On the current project we have a blend of Unit Tests, Integration tests, System Tests, performance tests and, unfortunately, a heavy contingent of manual tests still. |
Python Data Pipeline which pushes data into a Postgres Database fronted by a .NET / JS web app. |
VSCode, Visual Studio, Git, GitlabCI, Docker, Azure and associated tooling |
All code goes through peer review and gets signed off through a formal QA process on client data |
11-30 seconds |
50 |
30 |
20 |
Show
|
| I am the coach and faciliator I am not actively programming :) |
x86 |
VScode |
we used to 2 level code reviews and prefer code |
1-2 hours |
5 |
5 |
5 |
Show
|
| Using TDD - unit, integration, e2e. Automated CI to run tests on hardware |
Linux based systems, Raspberry Pis, cloud |
VScode, Jenkins, Docker |
Usually pair program so the code review is done during development of a feature rather than after |
Under 10 seconds |
3 |
6 |
1 |
Show
|
| I test edge cases and then come up with a test plan for a technician to run. |
embedded devices |
VScode, eclipse based IDEs |
I have never done one |
1-5 minutes |
3 |
3 |
4 |
Show
|
| Varies |
Varies, but mostly Linux VMs (not embedded) |
Varies |
Varies |
1-5 minutes |
50 |
20 |
10 |
Show
|
| Unit, integration & end-to-end tests, CI & manual |
Various embedded systems |
Various: VSCode, functional safety compilers, CI |
Scrum |
11-30 seconds |
25 |
10 |
25 |
Show
|
| For code touching hardware, most is done with custom code that allows hardware to be exercised to verify functionality |
Windows PC and/or embedded target |
Vendor provided (MPLAB / Keil)
Visual Studio Code
|
Internal use - informal or None depending on scope
External - Formal peer reviews |
11-30 seconds |
30 |
40 |
30 |
Show
|
| As I generate code, I create test functions at the bottom of the same file. The tests are intended to catch edge cases and compare function output to known good output. The test functions are wrapped in #ifdef for easy removal. |
Embedded C on microcontrollers & some scripting with Python |
Vendor provided IDEs (Code Composer Studio, STM32CubeIDE, etc.) |
After verifying functionality, I look to senior programmers for feedback on style and implementation |
11-30 seconds |
40 |
20 |
40 |
Show
|
| Unitary tests, corner tests, functional tests and (if applicable) performance tests. |
Embedded Linux-based systems. GNSS receivers. System on chips. |
Visual Studio Code IDE (previously Eclipse).
gcov/lcov, cppcheck, Valgrind, clang-tidy/format |
Deep code reviews about the functionality of the code, and docs, convention, etc done by CI |
Under 10 seconds |
4 |
4 |
2 |
Show
|
| individual developers define their own Ad hoc tests unless specifics requirements are defined by the customer |
Linux |
Visual Studio Code, Sublime, other IDEs. Gitlab/Github |
Over the shoulder/ virtual screen share code reviews. Usually limited to a few hundred lines. |
11-30 seconds |
70% |
10% |
20% |
Show
|
| Creating Test Cases using the same language and IDE |
Microcontrollers (C/C++) and desktops (C#) |
Visual Studio, Spyder, Arduino |
I like to circulate code, then have an in-person review. |
1-5 minutes |
30 |
30 |
40 |
Show
|
| "Automated" testing by writing code to test code, not using a unit test framework such as the one provided by .NET. |
Windows machines, Xilinx devices |
Visual Studio, Vivado, SDK |
Using either a lead dev on a project or a group of people to review code prior to merging to main. |
1-5 minutes |
15% |
35% |
50% |
Show
|
| Test Benches, Prayer |
ARM processors on Xilinx MPSoCs |
Vivado, Vitis, Visual Studio, VS Code |
None |
5-30 minutes |
2 |
3 |
5 |
Show
|
| manually |
game consoles, pcs, etc. |
all and any |
solo developer so it's new terrain |
5-30 minutes |
50 |
40 |
10 |
Show
|
| unit test (gtest), top level tests (system integratoin tests) |
huge legacy model (several million lines) that has existed over a decade, maintained by many teams |
c/c++/python/assembly |
insanely strict; including a must-add test case |
5-30 minutes |
25 |
35 |
40 |
Show
|
| catch unit test
python integration and end to end tests |
microcontrollers embedded |
clion
gcc tool chain |
pair programming or formal document |
5-30 minutes |
30 |
50 |
20 |
Show
|
| see if it does what i want it to |
arm microcontroller |
code composer studio |
haven't done any |
Under 10 seconds |
40 |
20 |
40 |
Show
|
| Try to use TDD where Possible |
CN0510
https://github.com/graial/ad5940-examples-adjust-output |
cross core embedded studio (eclipse based) |
nil, one man team |
Under 10 seconds |
3 |
5 |
2 |
Show
|
| some of the legacy code here is tested using tools developed after Grenning training (10+ years ago?). I know zero details beyond: "the tests run at build time" |
various embedded systems, more towards the 32 bit microcontroller |
eclipse, xdc tools |
informal if existent |
11-30 seconds |
3 |
3 |
3 |
Show
|
| unit tests or python scripts |
embeeded device, with multiple applications mainly coded in c++ |
gcc, gdb, coverity, ctc, valgrind, visualcode, vim, libasan, cppcheck, codescene |
we have code reviews for each pull request |
31-60 seconds |
60 |
30 |
10 |
Show
|
| Unit test, Integration Test, System Test, Regression Test |
x86, arm |
gcc, vscode, make, cmake |
NTR |
31-60 seconds |
40 |
20 |
40 |
Show
|
| unit tests up to a minimum coverage, robot framework for system level tests, "manual" functional tests |
satellites on-board computer |
git, cmake as a build system, vscode, self-hosted LLM |
code reviews are requested by developers when they think they need it, focused on design |
1-2 hours |
40 |
30 |
30 |
Show
|
| unit test with cFS, system test with robot framework, "manual" functional test |
spacecraft on-board computer |
git, cmake build system, vscode |
code reviews are requested by developers when they finish a task, they are mainly based on design |
1-2 hours |
30% |
30% |
40% |
Show
|
| varied: Unit, integration, E2E tests. Very few human-in-the loop manual tests. |
varied: FPGA, MCU, Embedded C, Linux |
git, gcc, CMake, CLion, JIRA, Jenkins, Docker, Artifactory |
Pair program instead |
Under 10 seconds |
10 |
10 |
20 |
Show
|
| I try to cover as much of my code as I can with unit tests, and run functional and integration tests too. |
I primarily write code for embedded ARM processors in avionics systems. |
git, vscode, docker, gdb, github. |
My goal when going through code reviews is to keep them as small in size and scope as possible. |
11-30 seconds |
60% |
10% |
30% |
Show
|
| manually |
NetSuite customization and scripting |
gitlab, notepad++ |
all code and code changes gets reviewed by at least one person. Code reviewers rotate. |
31-60 seconds |
30 |
40 |
30 |
Show
|
| Creating test procedures based on code features and validating against. |
Embedded devices, iot. |
ide, repository control, text editors, ai, debuggers. |
Have not had any yet with current company. |
11-30 seconds |
33 |
33 |
33 |
Show
|
| unit and integration tests |
arm, x64 |
jetbrains suite, vs code, sublime text |
3 reviewers on every change, 2 junior and 1 senior |
1-5 minutes |
5 |
3 |
2 |
Show
|
| Manual tests, custom python scripts, ci-fuzz, cpp-check |
ARM processors (STM32, nRF52) with Ethernet/Wi-Fi and various radios. RTOS or bare metal. |
makefiles, cmake, arm gcc, vim, github |
Each PR is reviewed by another developer. |
11-30 seconds |
30 |
40 |
30 |
Show
|
| n/a |
n/a |
n/a |
n/a |
Under 10 seconds |
n/a |
n/a |
n/a |
Show
|
| UT |
Yet to explore |
normal |
Team reviews |
30-60 minutes |
3 |
3 |
4 |
Show
|
| TDD |
n/a |
pycharm |
I do them |
2-4 hours |
50 |
10 |
40 |
Show
|
| in system testing |
various, mostly embedded MIPs |
various |
use codecollaborator |
1-5 minutes |
40 |
10 |
50 |
Show
|
| googletest |
embedded |
various |
pair programming, walkthroughs, formal reviews. |
31-60 seconds |
30 |
40 |
30 |
Show
|
| unittest module in python |
embedded linux devices |
vim
vscode
pycharm |
we dont do them |
31-60 seconds |
40 |
20 |
40 |
Show
|
| json based test input and manual verification of output |
optical embedded system |
visual studio
git
bitbucket |
good. |
2-4 hours |
5 |
5 |
3 |
Show
|
| Manual |
Just here to learn a better way to code. No current projects with target system. |
visual studio code |
currently limited in current job function |
11-30 seconds |
50 |
20 |
30 |
Show
|
| Trial and error. |
windows and linux |
visual studio code |
review with other engineers on the project. |
1-5 minutes |
50 |
25 |
25 |
Show
|
| manual testing in real products |
linux and freertos environment |
visual studio, logic analyzer, oscilloscope, jtag |
none |
31-60 seconds |
50 |
10 |
40 |
Show
|
| On-target |
Embedded MCUs |
vscode, eclipse |
I get okay feedback but we don't have the software review culture I would have expected. |
1-5 minutes |
50 |
20 |
30 |
Show
|
| Most of the times it is manual test. We build and replace the modified binary/library in target system and check. |
SW applications running on Debian |
vscode, tabnine |
Bitbucket PR review |
1-5 minutes |
2 |
5 |
8 |
Show
|
| unit tests, function tests, system tests, manual tests |
IoT devices, embedded product with connection to cloud |
zephyr, git, cmake |
Mine are usually pretty small and concise, |
5-30 minutes |
20 |
40 |
40 |
Show
|