url: https://nlnet.nl/propose/
scope : Is it per task?
| Your name | Denis Carikli |
|---|---|
| Email address | PrivateContact + our contact at the FSF |
| Phone numbers | GNUtoo's phone number |
| Organisation | Replicant and the FSF |
| Country | France(Denis Carikli), USA (FSF) |
| Project name | <Depend on the task> |
|---|---|
| Website / wiki | <Depend on the task> |
| Abstract: Can you explain the whole project and its expected outcome(s).(you have 1200 characters) Please be short and to the point in your answers; focus primarily on the what and how, not so much on the why. Add longer descriptions as attachments (see below). If English isn't your first language, don't worry - our reviewers don't care about spelling errors, only about great ideas. On the up side, you can be as technical as you need to be (but you don't have to). Do stay concrete. |
<Depend on the task> |
| Have you been involved with projects or organizations relevant to this project before? And if so, can you tell us a bit about your contributions? |
Yes: I've been involved in Replicant since the beginning both as a developer and for managing the project: As a developer: * I did most/all the initial system work and made it work for the the HTC Dream, and the Google Nexus One. * I also worked on porting the Goldelico GTA04, Galaxy nexus, Galaxy Tab 2 7.1 along with other Replicant developers and did various bug fixes and improvements. * I am also doing code reviews for patches. And for managing the project I'm involved in: * public relations (blog post, etc) * fund usage decisions * infrastructure (system administration with other developers, etc) * documentation * project direction and strategic decisions |
| Requested Amount (Between 5000 and 50000 Euros) | <depends on the task> |
|---|---|
| * Explain what the requested budget will be used for? | <depends on the task> |
| * Does the project have other funding sources, both past and present? | The Replicant project has about 200000 dollars at disposition: * The Replicant project has a donation page https://crm.fsf.org/civicrm/contribute/transact?reset=1&id=19. Part of the donations were used for buying devices and reimburse conference attendances. We have about 20000 dollars remaining from the donation. * The Replicant project recently received 200000 dollars from Handshake: https://www.fsf.org/news/free-software-foundation-receives-1-million-from-handshake As the FSF takes 10% that leaves us 180000 dollars |
| Compare your own project with existing or historical efforts. | <Depend on the task?> |
| What are significant technical challenges you expect to solve during the project, if any? | <Depend on the task?> |
Describe the ecosystem of the project, and how you will engage with relevant actors and promote the outcomes?
The Replicant project contributors and the FSF will supervise contractors to do the work. I will write a blog post to anounce that the Replicant project has got some funding for this specific task, and that it is looking for a contractor to work on it. This is to make sure that every one has equals chances in the application process. Then the most suited contractor will be selected. Only contractors that already have worked on similar tasks as part of free and open source software projects will be chosen. This way we can look at their existing contributions and make sure that they are able to do the task before engaging with them. Some people already involved in the Replicant project with proven contributions, me included, are already interested about doing such contract work, so finding someone suitable to work on the task should not be an issue. The Replicant project will also make sure that the contractor has or gets the hardware required to work on the task, before starting to work on it.
| Attachments | None |
|---|
| What should we do in the other case, e.g. when your project is not immediately selected? |
I allow NLnet Foundation to keep the information I submit on record, should future funding opportunities arise |
|---|---|
| Send me a copy of this application. | check-box checked |
| PGP pubkey | None (if we use Replicant contact address, we can't encrypt to it) |
There is a thread about funding on the mailing list about that
Funding: We could apply to https://nlnet.nl/PET
It would also be useful to support devices using kernels that are based on upstream Linux with the least amount of kernel changes possible:
Currently, Replicant uses a dedicated Hardware Abstraction Layer per device, because device manufacturers implemented non-standard kernel interfaces. However, Android works with mainline kernels and supports plug-n-play hardware nowadays, so it makes sense to have generic Hardware Abstraction Layers for the standard interfaces of the Linux kernel (ALSA, V4L2, etc).
See also the wiki page on Upstream Linux for more details on why using upstream kernel is beneficial, and for what devices to choose to work on this task.
Hardware requirements :Difficulty: Medium
Requirements/Prerequisites: Knowledge of C, some C++, the ability to understand Java, kernel interfaces knowledge
Expected outcomes:The most recent smartphones that Replicant support are quite old (they were made around 2013). The goal here is to add support for more recent smartphones in Replicant.
Even if we think that it's at lot more important to support devices that are better for freedom (samsung devices usually have a nonfree bootloaders), adding supporting common (Samsung) phones and tablets is relatively easy and fast to do and could be a good way to get started in contributing to Replicant.
It's advised to pick a device that:Make sure to evaluate the device before starting to work on it. Some devices have been evaluated in the TargetsEvaluation wiki page. There is also a forum section for devices evaluation.
Hardware requirements:Difficulty: Medium
Expected outcomes:When porting Replicant to a new version, it's also a good idea to keep supporting all the devices we supported in the older versions, however this is not always possible or desirable.
In order not to require too much work, devices that were previously supported will have to meet the HardwareRequirements of the new Android version. Here many of the devices already supported by Replicant 6.0 already meet such requirements.
Hardware requirements and dependencies:Difficulty: Medium
It would be useful for a Replicant device to be able to update itself to a new version of Replicant without requiring being connected to a PC. LineageOS already supports this; we suspect that it should be possible to adapt this LineageOS functionality to Replicant.
Whenever possible, it would be useful to complete and submit some of the code written for Replicant to LineageOS.
Difficulty: Medium
Expected outcomes:| Project name | Finish porting Replicant to a newer Android version |
|---|---|
| Website / wiki | https://redmine.replicant.us/projects/replicant/wiki/Porting_Replicant_to_Android_9 |
Abstract: Can you explain the whole project and its expected outcome(s).in 1200 characters
Replicant is a fully free software Android distribution which is approved by the FSF (http://gnu.org/distros). The combination of Android Open Source Project source code with the Linux source code provided by the device vendor is not sufficient to produce a fully free Android distribution that works: A lot of the code that makes critical hardware component work (the modem, graphics, audio, GPS, etc) is in userspace. Because of that most device manufacturers don't release them as free software. To make such hardware work, the Replicant project manages to replace or avoid such nonfree software. Replicant is currently based on LineageOS 13.0 which in turn is based on Android 6.0.1 which are both not supported anymore. Replicant is based on LineageOS because it supports many more smartphones and tablets than the Android Open Source Project. The project consists in porting Replicant changes on top of the "Android 9" release of the Android Open Source project, and when LineageOS 16 will be ready, to backport our changes on top of LineageOS 16.
| Have you been involved with projects or organizations relevant to this project before? And if so, can you tell us a bit about your contributions? |
SEE TEMPLATE |
|---|
| Requested Amount (Between 5000 and 50000 Euros) | 50000 Euros |
|---|---|
| Does the project have other funding sources, both past and present? | SEE TEMPLATE |
Explain what the requested budget will be used for?
The budget will only be used to fund this task through contract work. We think it will take something between 3 and 6 month of work for one full time developer. However it is always difficult to evaluate precisely the amount of time that this kind of project would take as sometime it can be slown down a lot due to bugs needing to be fixed. For instance, when adding support for the Nexus One to Replicant, a lot of time was spent dealing with display issues that didn't affect the upstream projects, because they relied on the GPU which required nonfree software to work. If we take the cost of a Freelance developer in the USA (75$ to 150$ per hour) as a basis, to enable people living in Europe and the USA to apply, we can fund a developer to work on it for a period that is mostly equivalent to something between 2 to 4 months full time. So far we have at least one person interested on working on it as a contractor (me), and one volunteer that wants to work on it at the same time, but that cannot do it full time. We will make sure that everybody has a change to apply for doing contract work. If the work is not done when the 50000 run out, and that we cannot make sure that it will be completed by volunteers in a time that is not too long, the Replicant project will most probably use its existing funds to pay for contract work to make sure that this task is completed. The Replicant project will also take care of making sure that the people that will work on this task have the necessary hardware to do it, for instance by shipping or reimbursing the purchase of a compatible smartphone with the Replicant project money. Once we have the Samsung Galaxy SIII fully working with Replicant 9, we will then add support for most the smartphones we currently support in Replicant, and add support for more recent smartphones (the most recent one we currently support has been released in 2013). We also have a some very basic documentation on the Android 9 port here: https://redmine.replicant.us/projects/replicant/wiki/Porting_Replicant_to_Android_9
Compare your own project with existing or historical efforts.
Upgrading Replicant to a new Android version usually took about 2 or 3 months of full-time equivalent work for one person. Here, we already have a device (The Galaxy SIII 4G) booting under Android 9 with a kernel that is closely based on upstream Linux, but a lot still needs to be done (modem, audio, sensors, etc) and validated. The Android architecture also changed a lot more between Android 6.0.1 and Android 9 than it did when we ported Replicant to newer Android versions.
What are significant technical challenges you expect to solve during the project, if any?
We will also need to make sure that Replicant 9 can be built with a GNU/Linux distribution that is approved by the FSF. This could be challenging.
| Describe the ecosystem of the project, and how you will engage with relevant actors and promote the outcomes? | SEE TEMPLATE |
|---|---|
| Attachments | SEE TEMPLATE |
This task aims to fix all these severe issues by working in parallel with the Android 9 port and leveraging new graphics functionalities present in Android 9 like X, Y and Z. On Android 9 it should be easy to use SwiftShader, Google's current software renderer that is capable of full GLES 2.x.
Unfortunately Android 9 doesn't have a gralloc (graphics memory allocator) library for software rendering that is compatible with hwcomposer (Hardware Composer HAL). There are patches to work around that and have a gralloc library that uses the framebuffer interface, however this has very bad consequences on performance. This is also not an approach compatible with upstream projects as they have dropped support for the framebuffer interface.
The major and first goal of this task is thus to write a hwcomposer compatible gralloc library that implements the full Gralloc HAL and supports software rendering. If used in conjunction with SwiftShader, this new graphics' stack should enable full GLES 2.x support on Replicant, with a decent performance.
Hardware requirements: A computer that is able to build Replicant. A smartphone or tablet that is supported by Replicant to be able to test the result.
Difficulty: Medium / Hard
Requirements/Prerequisites: Knowledge of C++, kernel interfaces knowledge or the ability to learn them
Expected outcomes:The best way around this is to integrate Mesa into the graphics stack and use it as an alternative to SwiftShader. Mesa implements both GLES 3.x and Vulkan. It is a big community project, with hundreds of active contributors and great community support.
Replicant 6 already includes Mesa, albeit an old version (13.0.3), picked from Android-x86, that has long been deprecated.
The goal of this sub-task is to have the current mainline Mesa running on Replicant, in order to take advantage of it's performance improvements and new features to develop all following sub-tasks (llvmpipe optimizations and Lima driver).
Hardware requirements: A computer that is able to build Replicant. A smartphone or tablet that is supported by Replicant to be able to test the result.
Difficulty: Medium
Requirements/Prerequisites: Knowledge of C++, Makefiles and git. Android's graphics stack knowledge or the ability to learn them.
Expected outcomes:Mesa is a highly versatile library that can be extended with device drivers to allow it to be used in different environments ranging from software emulation to complete hardware acceleration. One such driver is the Gallium llvmpipe driver, which is a software rasterizer that uses LLVM to do runtime code generation. It only needs a CPU to run graphics computations and thus brings full GLES support to all Replicant devices.
llvmpipe has been integrated in Replicant 6 but it's not activated by default yet as it is very slow. It is also not fully complete.
To fix that, llvmpipe and/or the integration of it in Replicant should be optimized. We should first start by configuring llvmpipe and/or Mesa to not implement very expensive OpenGL operations. If that's not sufficient, or if that breaks application compatibility, various software or hardware features (ARM NEON, hardware 2D acceleration, etc) could be used to improve the speed.
Hardware requirements : A computer that is able to build Replicant. A smartphone or tablet that is supported by Replicant to be able to test the result.
Difficulty: Medium / Hard (depending on the amount of optimizations required)
Requirements/Prerequisites: See with Mesa project
Expected outcomes: faster llvmpipe on ARM devices, able to run apps such as Fennec F-Droid (Firefox).
Lima is a free software Mesa driver for ARM Mali-4xx (Utgard) GPUs. These GPUs are present in several Replicant supported devices such as Galaxy S2, S3, S3 4G, Note and Note 2.
Lima aims to full GLES support but it is still in development. However the current implementation status already allows the hardware acceleration of several tasks. GPU-based hardware acceleration is faster and less power hungry than software rendering, both by several orders of magnitude. It would allow Replicant devices to run applications with a performance close to that of non-free devices.
Hardware requirements : A computer that is able to build Replicant. A Replicant device with a Mali-4xx GPU that can run mainline Linux (e.g. Galaxy S3 or Note 2).
Difficulty: Medium
Requirements/Prerequisites: See with Lima project
Expected outcomes: Lima driver being used for GLES rendering on a supported device.
Samsung-RIL is the RIL (Radio Interface Layer) that many Replicant devices use to communicate with the modem. It is a free, reverse-engineered replacement for the proprietary RIL that the Samsung phones ship with by default (which has been found to have backdoors).
Right now, Samsung-RIL mostly implements only the protocol features that are absolutely necessary for the phone to be operable. As a result, many more rarely used protocol features are unimplemented, which decreases functionality compared to the proprietary RIL. You can help by implementing the missing features of Samsung-RIL.
It would also be nice to fix most the reported bugs involving samsung-ril and libsamsung-ipc that are impacting users very seriously. This includes the bugs about the SIM card not being detected, and the issue about having metallic sound quality when doing voice calls over 3G (bug #1773). It would also be nice to be able to recover from EFS (the modem filesystem) corruptions (Bug #1869).
Hardware requirements : A computer that is able to build Replicant. A smartphone or tablet supported by Samsung-RIL.
Difficulty: Medium to Hard
Requirements/Prerequisites: Knowledge of C.
Expected outcomes: Implement the missing features listed at Samsung-RIL. When all the features have been implemented, also ask usptream (LineageOS) if they want to use libsamsung-ipc and samsung-ril.
Dependencies: This task should be fairly independent as:Funding: We could apply to https://nlnet.nl/PET
| Project name | Complete libsamsung-ipc and libsamsung-ril |
|---|---|
| Website / wiki | https://redmine.replicant.us/projects/replicant/wiki/Samsung-RIL |
Abstract: Can you explain the whole project and its expected outcome(s).in 1200 characters
Replicant is a fully free Android distribution that is approved by the FSF (http://gnu.org/distros). It supports several Samsung smartphones tablets that have a modem. The modem can be thought as a separate computer in a chip that is dedicated for interfacing with the cellular network. Many use custom protocols that is implemented by nonfree software to communicate with the smartphone OS (Android). This has issues: https://redmine.replicant.us/projects/replicant/wiki/SamsungGalaxyBackdoor The Samsung IPC protocol is used by the modems of the devices currently supported by Replicant, and in a lot of other Samsung smartphones and Tablets. Replicant implemented it in: - libsamsung-ipc: the low level protocol implementation - libsamsung-ril: the interface between libsamsung-ipc and Android The project consists in implementing the missing features in samsung-ril and libsamsung-ipc, which are known and documented in the RIL API. They are things like "start a conference call". If the features are completed, we expect other Android and GNU/Linux distributions to start using and contributing to it. It will also improve Replicant usability and compatibility with networks.
| Have you been involved with projects or organizations relevant to this project before? And if so, can you tell us a bit about your contributions? |
SEE TEMPLATE |
|---|
| Requested Amount (Between 5000 and 50000 Euros) | 50000 Euros |
|---|---|
| Does the project have other funding sources, both past and present? | SEE TEMPLATE |
Explain what the requested budget will be used for?
The budget will only be used to fund this task through contract work. We think it will take something between 3 and 4 month of work for one full time developer. If we take the cost of a Freelance developer in the USA (75$ to 150$ per hour) as a basis, to enable people living in Europe and the USA to apply, we can fund a developer to work on it for a period that is mostly equivalent to something between 2 to 4 months full time. The Replicant project will take care of making sure that the people that will work on this task have the necessary hardware to do it, for instance by shipping or reimbursing the purchase of a compatible smartphone with the Replicant project money.
Compare your own project with existing or historical efforts.
Here, implementing the missing features will be done in the same way than before, which is running the proprietary implementation and understanding the data format of the data going from/to the modem that is gathered either with strace or by patching the kernel, and implementing the feature in libsamsung-ipc and libsamsung-ril.
What are significant technical challenges you expect to solve during the project, if any?
There is currently no CDMA support at all in Replicant and libsamsung-ril/libsamsung-ipc. A lot of areas in the world don't have any CMDA coverage, so testing the implementation could be challenging as it would either require the contractor to live in an area with CDMA coverage, or to be able to build a cheap CDMA infrastructure to be able to test the implementation. If we don't have good enough assurances that implementing CDMA is doable, that will not be attempted.
| Describe the ecosystem of the project, and how you will engage with relevant actors and promote the outcomes? | SEE TEMPLATE |
|---|---|
| Attachments | SEE TEMPLATE |
The LTE variants of the Samsung Galaxy S3 and Samsung Galaxy Note 2 use a different modem from the non-LTE variants that Replicant currently supports. You can help Replicant support those modems by implementing a QMI-RIL, which performs a similar role on the LTE variants as what Samsung-RIL performs on the currently-supported non-LTE variants. Wolfgang has done some preliminary work on this, so you'll probably be picking up where he left off.
Hardware requirements : A computer that is able to build Replicant. A smartphone or tablet supported by QMI-RIL like the Galaxy SIII 4G (i9305).
Difficulty: Hard
Requirements/Prerequisites: Knowledge of C.
Expected outcomes: A QMI-RIL that supports voice calls, SMS, and data, with as complete a protocol implementation as possible.
Dependencies: This task should be fairly independent as:Funding: We could apply to https://nlnet.nl/PET
The the Galaxy S 2 (I9100), Galaxy S 3 (I9300) and Galaxy Note 2 (N7100) currently use a kernel based on a vendor fork of Linux, which poses a maintainability and security issue. Forkbomb has done some initial work on porting these devices to use mainline Linux. You can help by continuing this work. This would also enable these devices to use generic hardware abstraction layers (HAL) when abstractions layers are ready, and to do some research on whether the TrustZone operating system can be removed from such devices.
Hardware requirements : A computer that is able to build Replicant. A Galaxy S 2 (I9100), Galaxy S 3 (I9300) or Galaxy Note 2 (N7100), and a serial port adapter to get the kernel boot logs.
Difficulty: Medium
Requirements/Prerequisites: C programming language, driver development
Expected outcomes: Audio working, modem working, and Replicant or LineageOS booting with mainline Linux.
Having an automated build and test infrastructure would be very beneficial for Replicant.
Issues:A lot of time is spent on the wiki documentation, and a lot of information is redundant (for instance the installation guide)
TODO:We might want to consider Android 9 HardwareRequirements before working on that
The Optimus Black from LG is an interesting device from the perspective of freedom and privacy/security. It has the ability to run a free bootloader and uses an OMAP3 SoC that is well-documented and supported in upstream U-Boot (bootloader) and Linux (kernel). Its modem is well-isolated from the rest of the device, ensuring a sane base for privacy/security. Currently, the device-specific parts of the mainline U-Boot and Linux ports are still at an early stage, where they are functional with a very limited set of supported hardware.
Advancing the Optimus Black U-Boot and Linux mainline ports would allow using the device with free, up-to-date and maintainable software and would pave the way for support in GNU/Linux systems as well as Replicant. A list of priorities in hardware support will be defined, with the objective of tackling as many as possible.
Hardware requirements : A computer that is able to build Replicant. An Optimus Black with u-boot and modified boot pins, a serial port adapter to get the kernel boot logs.
Difficulty: Medium to Hard
Requirements/Prerequisites: C programming language, driver development
Expected outcomes: Improved hardware support for the Optimus Black in U-Boot and Linux
The Kindle Fire (first generation) from Amazon is an interesting device from the perspective of freedom and privacy/security. It has the ability to run a free bootloader and uses an OMAP4 SoC that is well-documented and supported in upstream U-Boot (bootloader) and Linux (kernel). It does not embed a modem, ensuring a sane base for privacy/security. Currently, the device-specific parts of the mainline U-Boot and Linux ports are still at an early stage, where they are functional with a very limited set of supported hardware.
Advancing the Kindle Fire (first generation) U-Boot and Linux mainline ports would allow using the device with free, up-to-date and maintainable software and would pave the way for support in GNU/Linux systems as well as Replicant. A list of priorities in hardware support will be defined, with the objective of tackling as many as possible.
Hardware requirements : A computer that is able to build Replicant. A Kindle Fire first generation, a serial port adapter to get the kernel boot logs.
Difficulty: Medium
Requirements/Prerequisites: C programming language, driver development
Expected outcomes: Improved hardware support for the Kindle Fire (first generation) in U-Boot and Linux
Tablets with Allwinner SOCs are an interesting targets because they do not use signed bootloaders and the SOCs and various devices using them have good Linux and u-boot mainline support. If not much work is required for that, once the code is merged, the candidate is also required to work on the generic abstraction layer project which is also documented in this page.
The chosen tablet should have:It would be better if the chosen tablet doesn't use an AllWinner SOC with a PowerVR GPU, as MALI GPU have more probability to be usable with free software in the future.
Hardware requirements : A computer that is able to build Replicant. An Allwinner tablet, a serial port adapter to get the kernel boot logs.
Difficulty: Medium
Requirements/Prerequisites: C programming language, driver development
Expected outcomes: Replicant support for a tablet using an Allwinner SOC, with free software bootloader and mainline based Linux kernel.
Replicant is plagued by various security issues, that are mostly due to using a downstream codebase. One of the most crucial issues is that Replicant uses an old version of the Android WebView (from circa 2015), which is also a functionality drawback.
An initial evaluation of the security issues in Replicant should be conducted, followed by the integration or update of the concerned components of the system.
It would also be nice to do the same for privacy issues. Since Replicant indirectly depends on the "Android Open Source Project" and directly depends on LineageOS, not all privacy issues might have been found fixed by Replicant. Once security issues have been fixed, it would be nice to try to identify as many privacy issues as possible, and in a second time to fix them.
Hardware requirements : A computer that is able to build Replicant. A smartphone or tablet that is supported by Replicant to be able to test the result.
Difficulty: Medium-Hard
Requirements/Prerequisites: Android build system, knowledge of system security, advanced git
Expected outcomes: Integration or update of components of Replicant to tackle security issues
Replicant has some issues with FSDG compliance: F-droid repository is not FSDG compliant anymore (Bug #1629), and Replicant can't be built from an FSDG distribution (Bug #1861). This ought to be fixed. Replicant should also be fixed to build without issue.
It would also be nice to have the build system not depend on pre-built dependencies anymore, and to document which FSDG compliant F-droid applications crash because Replicant's incomplete EGL implementation (#705) and tag such applications as incompatible (so they are greyed out) until the EGL implemetation is fixed. Ideally Replicant builds should also be made reproducible if they are not already.
Hardware requirements : A computer that is able to build Replicant. A smartphone or tablet that is supported by Replicant to be able to test the result.
Difficulty: Easy
Requirements/Prerequisites: Knowledge of shell scripts and the ability to learn the Android build system
Expected outcomes: The ability to compile Replicant from an FSDG distribution, F-droid only showing FSDG compliant software.
All devices currently supported by Replicant have WiFi chips that requires a non-free firmware to work. So to have WiFi working with free software, users need to use external WiFi adapters. They typically use tiny ath9k_htc compatible USB WiFi adapter along with a tiny USB OTG Host adapter.
Such external USB WiFi adapters used with Replicant are originally intended for laptops, not phones. As a result, they tend to consume a lot of power. According to lsusb some ath9k_htc compatible devices can consume up to 500mA.
This poses several issues:Such USB WiFi adapters can also randomly stop working completely on some devices (e.g. needing to unplug and replug the adapter periodically to keep it operational).
You will need to investigate reliability issues such as the one mentioned above and look how power consumption can be improved in the adapter firmware and/or kernel driver.
You will also need to investigate how much miliampers USB devices can use, at the hardware level, on the smartphones and tablets Replicant supports.
Hardware requirements : An ath9k_htc compatible WiFi card, the ability to measure the current usage, the ability to build the ath9k_htc firmware and driver.
Difficulty: Medium/Hard
Requirements/Prerequisites: Knowledge of C
Expected outcomes: Reliable WiFi with external WiFi adapter