Skip to main content

Projects to Work on

In this section, some projects, issues, improvements or ideas are presented for the larger Ada community to work together and make Ada better. However, if you are a newcomer, this section is also for you! If you are looking for practical topics to train your skill and help the community, this is a great starting point.

How is this list structured?​

The proposals are ordered by topic. Within each topic, they are ordered in terms of difficulty, that way, regardless of your experience with Ada or with the topic at hand, there should always be something for everybody.

Notice that the selection of projects here is not exhaustive and will never be. The criteria for adding new points to this list is that the projects or topics should have a noticeable impact on the community. Individual ideas or personal projects will not be promoted here, but feel free to work on those too!

List of projects

Ada Advocacy​

Ada does not have a lot going in the marketing department, let's fix that! All these proposals apply regardless of your Ada experience, each person will have their own take and point of view.

  • Helping newcomers: Ada has a constant influx of new and curious people wanting to try the language and learn it. One of the most important parts of the learning experience is the community and how it welcomes and aids newbies. The Ada community could use a few more people willing to help and guide new programmers with their first few steps into the language. You will find links to various sites where learners frequently ask questions in the Community column of the ada-lang.io footer.
  • Blog posts: writing blogs is always a nice way to share knowledge, experience, tips and anything in between. The Ada community could have a little more of that. And don't forget to share the blog posts with the community afterwards!
  • Forums, social media and language discussions: there are a lot of general discussions going on in the Internet, we should make Ada have a bit more presence there (if it is suitable!).
    • Example: the Awesome WASM Lang repository lists all the languages that support targetting WASM. Ada supports WASM! but it is not on the list. It would be great if someone did a Pull Request. Thanks to mgrojo for his pull request. It didn't even require writing a single line of Ada code, it is that simple!
  • Conferences: if you feel confident with your speaking and teaching skill, going to a conference and showing off Ada is a great way to attract the curious eyes of the audience. The topics here are also quite broad, ranging from simple and short tutorial all the way to niche and advanced features of the language.

Documentation​

Documentation can easily make or break projects, luckily, Ada is one of the best documented languages out there, thanks to it being an ISO standard and having extensive compiler and tooling documentation. Nonetheless, things could be much better, specially for individual projects and libraries. Documentation for individual tools will be listed in their respective categories. Here are the main places which could use help with their documentation:

  • Ada-Lang.io: you can help grow the documentation, guides, experience of this webpage, which tends to be one of the first resources that new people to the language take. The Learn section could use some extra hands, specially with the addition completion of How-Tos and Tips and Tricks. The addition of SPARK documentation would also be greatly welcomed.
  • Ada Wikibook: the Ada Wikibook has been a great source of knowledge for many years and many people. Nonetheless, it could use some clean-ups, completion and updates related to Ada 2022.
  • SPARK: SPARK does not have as much documentation as Ada does. This is just natural as it is a newer system and it is also a much more complex and advanced one. Therefore, SPARK could use some extra examples, guides and help from the community.
  • Improve the Standard Library documentation. Ada has an advanced Standard Library that is carefully documented in the Ada standard and in the GNAT Reference Manual. However, there are not many examples regarding its use. It is also not easy to read/digest documents describing the Standard Library. Better documentation would be useful for the larger Ada community and newcomers. Writing examples and documentation for the Standatd Library would also be very useful for new Ada programmers, as it makes for a wonderful practical learning experience.
  • Translations: the Ada community is all over the world and there are many people who are not fluent in English. If you speak a different language and you feel like translating content to it would help other people, go ahead! Luckily, nowadays, most people are somewhat fluent in English to be able to read it and understand it, so translation work is not as necessary as it used to be.

Teaching​

While Ada is a very readable and easy to learn language, its community does not shine when it comes to diverse and omnipresent documentation. If you would like to help fix this, here are a few ideas.

For beginners and new Ada programmers:

  • Rosetta Code: Rosetta provides examples and solutions to common problems and tasks in many programming languages. It serves as a learning resource as well as a comparison tool between languages. Ada already has a lot of examples, but still, some are missing compared to other languages. And some of the examples are quite old and they could be updated and cleaned up. These tasks should be fairly easy to carry out and they will help you learn Ada along the way!
  • GeeksforGeeks and Tutorials Point: these well-known websites have resources on different languages and topics about them, ranging between simple syntax elements to solving specific tasks in them. Ada has little to no presence in them. It would be nice if we could document Ada in them or point people to use Learn Ada for more information.

For intermediate and seasoned Ada and SPARK users:

  • Alice is a work-in-progress platform that is focused on teaching Ada and SPARK by focusing on high quality lessons and solutions (excellence as it is indicated in Alice's documentation). One could help develop its infrastructure, testing and adding lessons and tasks to it.
  • Exercism is a learning platform where students solve problems (with varying degrees of complexity) and then a tutor or reviewer can give them feedback if necessary. Ada used to have an entry in the website see deprecated repo. However, due to the lack of exercises and people willing to carry out the work maintaining it and helping students, it is no longer available. It would be great if it could be brought back. Nonetheless, Alice, see above, is trying to achive a similar goal while being focused on Ada/SPARK.

Toolchains​

  • Alire is one of the most important tools for the Ada community, regardless of the experience that one may have with the language. Alire is always looking for testers for new releases, feedback, improvements and fixes. If you are looking for a project to help with, Alire would be great!
    • Add new crates: one of the easiest ways to help Alire is by adding Ada applications and libraries (called crates) to the index and making them available for the entire community.
    • Help with the documentation.
    • Help build Alire in new architectures (for example RISC-V).
    • Improve the compiler packaging of Alire by adding a new architecture and testing it more.
    • Package GNAT-LLVM.
  • Operating System's Packages: Ada is not widely packaged in a lot of distributions' default package manager. A lot of people initially try to use the system's provided package only to find out that Ada is not packaged or poorly packaged. It would be very benefitial to a lot of new (and seasoned) Ada users to have a nice out-of-the-box experience with their operating system.
    • Guix: help package a working toolchain for the supported architectures. Package Alire.
    • OpenSUSE: extend the gcc-ada packaging to all the supported architectures. Create cross-compilation toolchains. Package Alire.
    • RedHat and Fedora: improve the support of the Ada toolchain. Package Alire.
    • FreeBSD: improve the current port and extend support to other architectures. Update the packaged version.
    • NetBSD: update the compiler and patches. Check support for NetBSD 10.
  • Ravenports is a package manager similar to pkgsrc. Help test it, package programs in it and help grow its OS support.

Libraries and Tools​

Bindings​

  • Create bindings to well-known tools and libraries: not a lot of tools are implemented in Ada and not all need to be! But it is always more convenient if Ada users could just start using a wider variety of tools and libraries. For that reason, it would be greatly benefitial to have more Ada bindings to well-known and widely-used projects so that both, new and seasoned Ada programers, can easily pick them up and start using their functionality.
    • Ideas:
      • ImGUI is a widely used GUI library. There is already an Ada-binding, but it could be updated and added to Alire.
      • stb is a collection of C header-only utilities in the public domain.
      • raylib is a collection of simple commonly-used utilities and libraries geared towards creating videogames.
      • ccv and OpenCV are libraries geared towards computer vision and processing.
      • And many more!
  • SWIG4Ada: Ada has outstanding support for creating bindings to other languages, specially C, Fortran and Cobol. GCC also has a method for creating bindings to C++ projects, however, it is quite limited with what it can do. SWIG4Ada implements support for Ada into the SWIG binding generator. It is still incomplete and it could use some more hands.

Graphics and Games​

  • SDLAda: the bindings to SDL still have some missing pieces/functions that need to be finalised, which can be found in the link's README page. This could be a great start for people wanting to get into game/graphics programming with Ada.
  • GTKAda is the GTK binding to Ada. It is poorly documented, there are not many tutorials nor blogposts about using it. Therefore, documentation would be a great starting point for someone wanting to get into GTK and Ada. Moreover, the bindings still target GTK3. If you are feeling adventurous, you may want to try and help update it to GTK4!

Web development​

  • AWS (Ada Web Services) is the standard web framework for Ada. It has some rough edges regarding TLS certificates, supporting newer web standards and documentation. All of these points could make use of a few willing hands.
  • AdaWebpack (Ada-WASM): WASM is the new cool kid in the web world and not without reason. It allows to run traditionally compiled languages in the browser, and Ada is no exception. AdaWebpack is a project that supports compiling Ada to WASM. Its support could be improved and more apps could be created with it. Here is a cool example of a spinning cube (running on WebGL) and the Ada app compiled to WASM in Android!
  • AWA (Ada Web Applications) is a collection of tools in order to build Web applications using Ada. Help by using it, testing and documenting it!

Cryptography and safety​

  • WolfSSL is a complete and certifiable, among others, TLS library. It recently added Ada binding support. It would be nice for the community to test it, improve it if necessary, potentially create SPARK code around it and package it into more places. Promoting it would also be a benefit.
  • SPARKNaCl is a cryptographic library fully proven in SPARK with no runtime nor external dependencies. It would be nice seeing it receive a bit more marketing in non-Ada environments and projects implemented with it!

Miscellaneous​

  • Ressource-Embedder is a resource embedder for Ada, Go and C programs. It allows the user to load data as part of the executable! You can help by utilising and promoting it.

Operating Systems​

The Ada ecosystem sports some of the most interesting kernels/OSes out there. Operating System development tends to be labeled as an advanced topic. However, if you are learning, thanks to Ada's clear syntax, it becomes much easier to understand them, use them and even contribute to an OS. Here is a list of kernels/OSes written in pure Ada/SPARK that may greatly benefit from the community:

  • Ironclad is a fairly new kernel that strives to be POSIX compatible. It partially written in SPARK, making it much more robust and safe by default. It aims to support x86, ARM and RISC-V architectures. You can contribute to it or you can help grow the Gloire OS distribution, which uses Ironclad as its kernel.
  • HiRTOS is a high integrity RTOS kernel as well as a separation kernel written in SPARK. Go and give it a try! One can even install it using Alire! It currently targets the ARM Cortex-R52 CPU. One could try and port it to a different CPU.
  • Muen is a separation kernel fully written in SPARK. It has been certified by some large military bodies for their internal use. It also sports some drivers fully written in SPARK!
  • M2OS is a small RTOS that was created from the same people behind the Ada OS Marte.

There are some well known Operating Systems that allow users to run Ada programs in them directly even if they are written in a different language. You may want to use them, improve the Ada support they have or write examples or blog posts about their use:

  • RTEMS is a POSIX RTOS that is widely used in the aerospace industry. It has official Ada support. However, its use is not clearly documented nor widely known about; this could be fixed.

Embedded Programming​

  • Ada Drivers Library is a collection of drivers for microcontollers and embedded boards. You can help by utilising it, testing the built-in drivers and specially by helping improve drivers and boards that are not yet present.

  • Create more HALs: HAL (Hardware Abstraction Layer) is a piece of code that abstracts the device away from the code. This is what generally drivers do. By implementing a HAL the low level aspects of the hardware are not needed in order to program the device, which makes the experience of embedded programming a lot more pleasant and comprehensible. Ada has quite a few HALs (see the Ada Drivers Library), but we are nowhere near to other communities, even though we exceed at embedded programming.

  • Improve svd2ada and startup-gen. These two tools allow for the quick creation of a initial HAL layer. There are a few issues and limitations in their Github page which could be improved.

  • Help the SweetAda development framework. SweetAda is a terrific framework designed to bring Ada support to as many architectures as possible. It would greatly benefit from more testing and specially documentation.

  • Improve support for tasking profiles in embedded OSes (Jorvik, Ravenscar). This is a very advanced topic, but an important one. Embedded systems generally do not support tasking due to the limited support that the runtime implementation has for them. Therefore, specialised profiles that allow for some basic, real-time and safe tasking abilities need to be created. It would be great to see more CPUs/Boards with support for the Jorvik and Ravenscar profiles.

  • AVR-Ada is an embedded development library that brings support for a large number of AVR devices to the Ada programming language. You can help by creating more examples, promoting it and expanding its capabilities.

Performance and benchmarks​

Ada tends to be shown in a lot of benchmarks, and that should be no surprise, as it is a highly optimizable, compiled programming language that does not use garbage collection and leverages some of the best compiler infrastructure in the world (GCC and LLVM). However, we do not tend to rank at the top. "Why is that?" one may ask... Well... Most Ada code used for benchmarking purposes has not been receiving the same amount of effort and care that other programming languages have. Let's fix this!

The following proposals are meant to be undertaken by people familiar with Ada and HPC (High Performance Computing):

  • The Computer Language Benchmarks Game: Ada programs could be optimised to utilise newer language constructs, algorithmic design and, once available and if useful, the new Ada 2022 parallel keyword. This should yield better results than the current, already performant programs. I believe we can improve our times quite a bit!
  • Programming Language and Compiler Benchmarks: Ada is not even listed in this competition! The task would require providing performant Ada examples for this language benchmark comparison.

Compilers​

The proposals here are quite advanced, therefore, they are not recommended for people unfamiliar with the Ada language.

  • Mainlining more patches to upstream GNAT: GNAT/GCC-Ada is a wonderful piece of technology, and thanks to it being based on the GCC compiler infrastructure, it can target a large number of Operating Systems and architectures. However, GNAT/GCC-Ada requires a bit of setup and additional files that are needed in order to compile Ada code for these targets. The most common architecture/OS pairs are already present inside GCC, but some communities maintain their own set of patches in order to support GNAT. It would be ideal if those patches could become part of the GCC source tree. Here are some resources and patches that would need mainlining:
    • Android: in the past there were some versions of GNAT that had support for Android and there is still Android code in the GCC/GNAT source code. However, it has not been tested nor used in a long while and it seems to have fallen in disrepair. Some patches for Android support can be found in this repository. Nonetheless, Ada also supports Android with WASM, see this demo.
    • OpenBSD has supported GNAT for a few architectures for a long while, however, the patches to add support have never been mainlined. The patches can be found in their ports tree or in this Github mirror
    • NetBSD: updated GCC patches to add support for NetBSD (amd64) can be found here. Adding support for other architectures can somewhat easily be done. Thanks to J. Marino for his work.
    • FreeBSD support within GNAT could use quite a bit of help. This is the current patch used by FreeBSD.
  • Adding support for more OSes: support could be added to the following "full-fledged" Operating Systems:
    • Ironclad is a POSIX compatible RTOS completely written in Ada. It currently has no RTS (RunTime System) support for compiling Ada programs targetting it.
    • Illumos is the continuation of OpenSolaris as an open source project. It currently has no Ada support within GNAT.
    • Gnode is an OS framework that is based on components and separation of concerns and security.
    • Haiku is a reemplementation of BeOS.
    • Microkernel and RTOS Systems: there are quite a few other OSes that we could support. They are either widely used or best in class for their target audience.
      • Zephyr is an RTOS that has been gaining a lot of traction and support. Initial support for the Zephyr RTOS was started here.
      • se4L is a secure, verified RTOS microkernel for x86, ARM and RISC-V devices.
      • FreeRTOS is a widely used RTOS for microcontrollers. There are some Ada ports that have support for it. However, mainline support would be preferred.
      • Azure ThreadX is a certified RTOS for embedded applications.
      • Apache Nuttx is a RTOS for embedded devices.
      • Minix 3 is one of the most well-known microkernels. It has enough POSIX support that is able to run the NetBSD userland on top of it.
      • RT-Thread is a new RTOS for embedded devices.
      • RIOT is an RTOS for small embedded systems.
    • Unikernels: these kernels have been created to run a single application and be the most secure possible as they have the smallest surface. The Ada community should ensure that Ada applications can run on these kernels.
      • Mirage OS is a well-known unikernel written in OCaml. See also Solo5. For example, the Muen website runs on it.
      • Unikraft is a unikernel focused on the cloud and security.
      • Toro is a unikernel written for microservices.
  • GNAT-LLVM: the GNAT-LLVM project is fairly new but very promising. It would be nice to test it, help with the documentation and create cool demos, see this Ada-WASM example.
  • Bootstrapping GNAT: GNAT currently requires a previous version to compile itself. Sadly, there was never an original public version of GNAT that did not have such requirement. This means, that GNAT cannot be compiled without already having a working copy of a GNAT installation. The GNAT Bootstrapping project aims to create a small, minimal compiler together with the live-bootstrap project in order to build an initial GNAT compiler.
    • Requirements: Scheme and compiler programming knowledge.
  • Improve Ada 2022 support in GCC/GNAT: there are still some Ada 2022 features that are yet to be implemented in GCC/GNAT, mainly the parallel construct. You can help develop new features, tests them and improve the current ones. For more information about the implementation of parallel in GCC/GNAT, see this forum thread
  • HAC is a small and limited Ada compiler, but it could make use of some help for those interested in compilers and Ada.