Hope to pick up pace with some blogging on our plans for Fedora Workstation in 2016 soon
One of the things that makes me really happy in terms of the public reception to the Fedora Workstation is all the people calling out how stable and solid it is, as this was and is one of our big goals from the start of the Fedora Workstation effort.
From the start we wanted to bury the old idea of Fedora being only for people who didn’t mind risking a lot of instability in return for being on the so called bleeding edge. We also wanted to bury the related idea that by using Fedora you where basically alpha testing highly unstable and unfinished software for Red Hat Enterprise Linux. Yet at the same time we did want to preserve and build upon the idea that Fedora is a great operating system if you want to experience a lot of the latest and greatest new developments as they are happening. At first glance those two goals might seem a bit contradictory, but we decided that we should be able to do both by both adjusting our policies a bit and also by relying more on the Fedora retrace server as our bug fixing prioritization tool.
So in terms of policies the division of Fedora into a distinct server and workstation images and also the clearer separation of the spins, allowed us to start making decisions without worrying so much how they affected other usecases than our own. Because sometimes what from a user perspective seems like a bug or something being broken was non-workstation policy decisions getting in the way of the desktop behaving as expected, for instance firewall rules hindering basic desktop functions.
Secondly we incorporated a more careful approach into what and when we brought in new stuff, meaning we still try to keep on top of major upstream developments and be a leading edge system, but at the same time we do a little mental exercise for each decision to make sure its a decision that makes us ‘leading edge’ and not ‘bleeding edge’. And if we really want something in, but it isn’t 100% ready for prime time yet we do what we have done with Wayland or the GTK3 port of LibreOffice, we make it available as an option for early adopters, but we default to the safer choice while we work out the last wrinkles. (Btw, if you are interested in progress on Wayland, Kevin Martin, sent out an emailing with a link to a good Wayland development status just before the Holidays.
The final piece of the puzzle is regularly checking and identifying important bugs from the Fedora retrace server. Because like almost all developers we get way more bug reports than we realistically can ever address, so having the data from the retrace server allows us to easily identify the crashes that affect the most users, and just as importantly lets us filter out the bug reports that are likely caused by users installing weird stuff on their system. When we started using retrace various desktop modules tended to dominate the top 3 pages when sorting bugs based on count, but due to a continuous effort over the last few years desktop modules appearing in the top crashers list are few and far between and when they do appear we make sure to get fixes done quickly for them. So if you ever wonder if the data collected by these kind of systems are actually helping developers working on the software you use better, I can say that it is true for Fedora for sure.
That said I thought it could be interesting to explain a bit the challenges we have with tracking our progress in this area. So lets start by looking at a graph I pulled from the retrace server.
Looking at that graph one could say that it is clear that we have made great strides in improving system stability and I do believe that is the case, however the graphs doesn’t truly prove that inconclusively, they are just an indication. The reason they are not hard evidence is that there are a lot of things you need to take into consideration when reading them. First of all they are not adjusted based on total user population, which means that if you win or lose a lot of users between releases it can create an appearance of increased instability or decreased instability which is actually due to increase or decrease in user population, not in ‘how well does the system run on an individual users system’. So from what we see through other metrics our user population has been increasing since we launched the Fedora Workstation which means we shouldn’t be getting any ‘help’ in these graphs from a declining user population.
A second reason is that there are a lot of false positives being reported here, for instance we had an issue for a long while that the Intel graphics drivers generating a ton of this crash reports without it actually being crashes as such. So while they did represent bugs that should ideally be fixed they where not issues you might actually have noticed as a user of the system. So we spent some effort between Fedora Workstation 21 and Fedora Workstation 22 to reduce the amount of noise caused by this, which was an useful effort for us in terms of reducing noise in our retrace server, but from a user perspective it didn’t really make a tangible difference. And even with our efforts there are a still a lot of kernel issues showing up here which are not impacting users in a way that they are likely to perceive as the system being unstable.
A third item that might in a given release skewer the statistics is that we currently don’t differentiate between Fedora Workstation and spins in the statistics, which means that there might be issues caused by one of the spins generating a lot of bug reports against a module, but that might be a bug or an API usage issue that is not triggered by the Workstation edition and thus those items appearing or disappearing might affect the statistics, but as a user of the Fedora Workstation you would never experience it.
So keeping this is mind the retrace server is an important tool for us and one that at least gives us a decent indication of how we are doing with quality. But we can always do better so we will keep reviewing the reports we get through the ABRT and retrace systems and I also do strong recommend any application or library maintainers out there to look into what major issues are reported against their own modules.
So many of you have probably seen that RHEL 7.2 is out today. There are many important updates in this release, some of them detailed in the official RHEL 7.2 press release.
One thing however which you would only discover if you start digging into the 7.2 update is that its the first time in RHEL history that we are doing a full scale desktop update in a point release. We shipped RHEL 7.0 with GNOME 3.8 and in RHEL 7.2 we are updating it to GNOME 3.14. This brings in a lot of new major features into RHEL, like the work we did on improved HiDPI support, improved touch and gesture support, it brings GNOME Software to RHEL, the improved system status area and so on. We plan on updating the desktop further in later RHEL 7.x point releases.
This change of policy is of course important to the many RHEL Workstation customers we have, but I also hope it will make RHEL Workstation and also the CentOS Workstation more attractive options to those in the community who have been looking for a LTS version of Fedora. This policy change gives you the rock solid foundation of RHEL and the RHEL kernel and combines it with a very well tested yet fairly new desktop release. So if you feel Fedora is moving to quickly, yet have felt that RHEL on the other hand has been moving to slowly, we hope that with this change to RHEL we have found a sweet compromise.
We will of course also keep doing regular applications updates in RHEL 7.x, just like we started with in RHEL 6.x. Giving you up to date versions of things like LibreOffice, Firefox, Thunderbird and more.
So yesterday, the 10th of November, was the official launch day of the Steam Machines. The hardware are meant to be dedicated game machines for the living room taking advantage of the Steam ecosystem, to take on the Xbox One and PS4.
But for us in the Linux community these machines are more than that, they are an important part of helping us break into a broader market by paving the way for even more games and more big budget games coming to our platform. Playing computer games is not just a niche, it is a mainstream activity these days, and not having access to games on our platform has cost us quite a few users and potential contributors over the years. I have for instance met a lot of computer science students who ended up not using Linux as the main operating system during their studies simply due to the lack of games on the platform. Instead Linux got de-regulated to that thing in a VM only run when you needed it for an assignment.
Steam for Linux and SteamOS can and will be important pieces of breaking through that. SteamOS and the Steam Macines are also important for the Linux community for another reason. They can help funnel more resources from hardware companies into Linux drivers and support. I know for instance that all the 3 major GPU vendors have increased their Linux drivers investments due to SteamOS.
So I want to congratulate Valve on the launch of the first Steam Machines and strongly recommend everyone in the community to get a Steam machine for their home!
People who have had a good chance to test the hardware has recommended me to get one of the Alienware SteamOS systems, so I am passing that recommendation onwards.
As a sidenote we are also working on a few features in Fedora Workstation to make it a better host for Steam and Steam games. This includes our work on the GL Dispatch and Optimus support as covered in a previous blog and libratbag, our new library for handling gaming mice under Linux. And finally we are working on a few bug fixes in Fedora to make it an even better host for the Steam client related to C++ ABI issues.
Another major piece of engineering that I have covered that we did for Fedora Workstation 23 is the GTK3 port of LibreOffice. Those of you who follow Caolán McNamaras blog are probably aware of the details. The motivation for the port wasn’t improved look and feel integration, there was easier ways to achieve that, but to help us have LibreOffice deal well with a range of new technologies we are supporting in Fedora Workstation namely: Touch support, Wayland support and HiDPI.
That ongoing work is now available in Fedora Workstation 23 if you install the ‘libreoffice-gtk3’ package. You have to install this using a terminal and dnf as this is a early adopter technology, but we would love as many as possible for you to try and report any issues you have either to the upstream LibreOffice bugzilla or the Fedora bugzilla against the LibreOffice component. Testing of how it works under X and how it works under Wayland are both more than welcome. Be aware that it is ‘tech preview’ technology so you might want to remove the libreoffice-gtk3 package again if you find that it hinders your effective use of LibreOffice. For instance there is a quite bad titlebar bug you would exprience under Wayland that we hope to fix with an update.
If you specifically want to test out the touch support there are two features implemented so far, both in Impress. One is to allow you to switch slides in Impress by a swiping gesture and the second is long press, you can bring up the impress slide context menu with it and switch to e.g. drawing mode. We would love feedback on what gestures you would like to see supported in various LibreOffice applications, so don’t be shy about filing enhancement bug reports with your suggestions.
HiDPI it wasn’t a primary focus of the porting effort it has to be said, but we do expect that it should also make improving the HiDPI support in LibreOffice further easier. Another nice little bonus of the port is that the GTK Inspector can now be used with LibreOffice.
A big thanks to Caolán for this work.
So as we quickly approach the Fedora Workstation 23 release I been running Wayland exclusively for over a week now. Despite a few glitches it now works well enough for me to not have to switch back into the X11 session anymore.
There are some major new features coming in Fedora Workstation 23 that I am quite excited about. First and foremost it will be the first release shipping with our new firmware update system supported. This means that if your hardware supports it and your vendor uploads the needed firmware to lvfs you can update your system firmware through GNOME Software. So no more struggling with proprietary tools or bootable DVDs. So while this is a major step forward in my opinion it will also be one of those things ramping up slowly, as we need to bring more vendors onboard and also have them ship more UEFI 2.5 ready systems before the old ‘BIOS’ update problems are a thing of the past. A big thanks goes to Richard Hughes and Peter Jones for their work here.
Another major feature that we spent a lot of time to get right is the new Google Drive backend for the file manager. This means that as long as you have internet access you can manage your google drive files through Nautilus along with all your other local and remote file systems. I know a lot of Fedora users are wither using Google drive personally or as part of their work, so I think this is a major new feature we managed to land. A big thanks to Debarshi Ray for working on this item.
Thirdly we now have support for ambient light sensors support. This was a crucial step in our ongoing effort to improve battery life under Fedora and can have very significant impact on how long your battery lasts. It is very easy to keep running the screen with more backlight than you actually need and thus drain your battery quickly, so with this enabled you might often squeeze out some extra hours of your battery. A big thanks to Bastien Nocera and Benjamin Tissoires for their work on this feature.
And finally this is the first release where we are shipping our current xdg-app tech preview as part of Fedora, instead of just making it available in a COPR. So for those of you who don’t know, xdg-app is our new technology for packaging desktop applications. While still early stage it provides a way for software developers to package their software in a way that is both usable across multiple distributions and with improved security through the use of the LXC container technology. In fact as we are trying to make this technology as usable and widely deployed as possible Alexander Larsson is currently trying to work with the Open Container Initiative to make xdg-apps be OCI compliant.
This is important for a multitude of reasons, but mainly xdg-app fills an important gap in the container technology landscape, because while Docker and Rocket are great for packaging server software, there is no real equivalent for the desktop. The few similar efforts that has been launched are usually tied to a specific distribution or operating system, while xdg-app aims to provide the same level of cross system compatibility for desktop applications that Docker and Rocket offers for server applications.
Fedora Workstation 24
Of course with Fedora Workstation 23 almost out the door we have been reviewing our Fedora Workstation tasklist to make sure it reflects what we currently have developers working on and what we expect to be able to land in Fedora Workstation 24. And let me use this opportunity to remind community members that if you are working on a cool feature for Fedora Workstation 24, make sure to let the Workstation working group know on the Fedora Desktop mailing list, so that we can make sure your feature gets listed and tracked in the tasklist too.
Anyway, I sent out this email to the working group this week, to outline what I see on the horizon in terms of major Fedora Workstation features lined up for 24.
You can get the full details in the email, and the tasklist has also been updated with these items, but I want to go into a bit more details on some of them.
xdg-app for world domination
As some of you might be aware of, Christian Hergert, of GNOME Builder fame, recently joined our team. Christian will be doing a lot of cool stuff for us, but one thing he has already started on is working with Alexander Larsson to make sure we have a great developer story for xdg-app. If we want developers to adopt this technology we need to make it dead simple to create your own xdg-app packages and Christian will make sure that GNOME Builder supports this in a way that makes transitioning your application into an xdg-app becomes something you can do without needing to read a long howto. We hope to have the initial fruits of this labour ready for Fedora Workstation 24.
Another big part of this of course is the work that Richard Hughes and Kalev Lember are doing on GNOME Software to make sure we have the infrastructure in place to be able to install and upgrade xdg-apps. As we expect xdg-apps to come from a wide variety of sources as opposed to the current model of most things being in a central repository we need to develop good ways for new sources to be added and help users make more informed choices about the software they are installing. Related to this we are also looking at how we can improve labeling of the applications available,
to make it easier to make your decisions based on a variety of criteria. The current star system in GNOME Software is not very obvious in what it tries to convey so we will try looking at better ways to do this kind of labeling and what information we want to be able to provide through it.
Another major item is what I blogged about before is our effort to finally make dealing with the binary graphics drivers less of a pain. I wrote a longer blog post about this before, but to summarize we want to make sure that if you need to install the binary drivers this is a simple operation that doesn’t conflict with your installation of Mesa and also that if you have an Optimus enabled laptop, it is easy and pleasant to use.
Anyway, there are some further items in the email I sent, but I will go more into detail about some of them at a later stage.
So I realized I hadn’t posted a Wayland update in a while. So we are still making good progress on Wayland, but the old saying that the last 10% is 90% of the work is definitely true here. So there was a Wayland BOF at GUADEC this year which tried to create a TODO list for major items remaining before Wayland is ready to replace X.
- Proper menu positioning. Most visible user issue currently for people testing Wayland
- Relative pointer/confinement/locking. Important for games among other things.
- Kinetic scroll in GTK+
- More work needed to remove all X11 dependencies (so that desktop has no dependency on XWayland being available.
- Minimize main thread stalls (could be texture uploads for example)
- Tablet support. Includes gestures, on-screen keyboard and more.
A big thank you to Jonas Ådahl, Owen Taylor, Carlos Garnacho, Rui Matos, Marek Chalupa, Olivier Fourdan and more for their ongoing work on polishing
up the Wayland experience.
So as you can tell there is a still lot of details that needs working out when doing something as major as switching from one Display system to the other, but things are starting to looking really good now.
One new feature I am particularly excited about is what we call multi-DPI support ready for Wayland sessions in Fedora 23. What this means is that if you have a HiDPI laptop screen and a standard DPI external monitor you should be able to drag windows back and forth between the two screens and have them automatically rescale to work with the different DPIs. This is an example of an issue which was relatively straightforward to resolve under Wayland, but which would have been a lot of pain to get working under X.
We will not be defaulting to Wayland in Fedora Workstation 23 though, because as I have said in earlier blog posts about the subject, we will need to have a stable and feature complete Wayland in at least one release before we switch the default. We hope Fedora Workstation 23 will be that stable and feature complete release, which means Fedora Workstation 24 is the one where we can hope to make the default switchover.
Of course porting the desktop itself to Wayland is just part of the story. While we support running X applications using XWayland, to get full benefit from Wayland we need our applications to run on top of Wayland natively. So we spent effort on getting some major applications like LibreOffice and Firefox Wayland ready recently.
Caolan McNamara has been working hard on finishing up the GTK3 port of LibreOffice which is part of the bigger effort to bring LibreOffice nativly to Wayland. The GTK3 version of LibreOffice should be available in Fedora Workstation 23 (as non-default option) and all the necessary code will be included in LibreOffice 5 which will be released pretty soon. The GTK3 version should be default in F24, hopefully with complete Wayland support.
For Firefox Martin Stransky has been looking into ensuring Firefox runs on Wayland now that the basic GTK3 port is done. Martin just reported today that he got Firefox running natively under Wayland, although there are still a lot of glitches and issues he needs to figure out before it can be claimed to be ready for normal use.
Another major piece we are working on which is not directly Wayland related, but which has a Wayland component too is to try to move the state of Linux forward in the context of dealing with multiple OpenGL implementations, multi-GPU systems and the need to free our 3D stack from its close ties to GLX.
This work with is lead by Adam Jackson, but where also Dave Airlie is a major contributor, involves trying to decide and implement what is needed to have things like GL Dispatch, EGLstreams and EGL Device proposals used across the stack. Once this work is done the challenges around for instance using the NVidia binary driver on a linux system or using a discreet GPU like on Optimus laptops should be a thing of the past.
So the first step of this work is getting GL Dispatch implemented. GL Dispatch basically allows you to have multiple OpenGL implementations installed and then have your system pick the right one as needed. So for instance on a system with NVidia Optimus you can use Mesa with the integrated Intel graphics card, but NVidias binary OpenGL implementatioin with the discreet Optimus GPU. Currently that is a pain to do since you can only have one OpenGL implementation used. Bumblebee tries to hack around that requirement, but GL Dispatch will allow us to resolve this without having to ‘fight’ the assumptions of the system.
We plan to have easy to use support for both Optimus and Prime (the Nouveau equivalent of Optimus) in the desktop, allowing you to choose what GPU to use for your applications without needing to edit any text files or set environment variables.
The final step then is getting the EGL Device and EGLStreams proposal implemented so we can have something to run Wayland on top of. And while GL Dispatch are not directly related to those we do feel that having it in place should make the setup easier to handle as you don’t risk conflicts between the binary NVidia driver and the Mesa driver anymore at that point, which becomes even more crucial for Wayland since it runs on top of EGL.
A couple of weeks I visited my mother back home in Norway. She had gotten a new laptop some time ago that my brother-in-law had set up for her. As usual when I come for a visit I was asked to look at some technical issues my mother was experiencing with her computer. Anyway, one thing I discovered while looking at these issues was that my brother-in-law had installed OpenOffice on her computer. So knowing that the OpenOffice project is all but dead upstream since IBM pulled their developers of the project almost a year ago and has significantly fallen behind feature wise, I of course installed LibreOffice on the system instead, knowing it has a strong and vibrant community standing behind it and is going from strength to strength.
And this is why I am writing this open letter. Because while a lot of us who comes from technical backgrounds have already caught on to the need to migrate from OpenOffice to LibreOffice, there are still many non-technical users out there who are still defaulting to installing OpenOffice when they are looking for an open source Office Suite, because that is the one they came across 5 years ago. And I believe that the Apache Foundation, being an organization dedicated to open source software, care about the general quality and perception of open source software and thus would share my interest in making sure that all users of open source software gets the best experience possible, even if the project in question isn’t using their code license of preference.
So I realize that the Apache Foundation took a lot of pride in and has invested a lot of effort trying to create an Apache Licensed Office suite based on the old OpenOffice codebase, but I hope that now that it is clear that this effort has failed that you would be willing to re-direct people who go to the openoffice.org website to the LibreOffice website instead. Letting users believe that OpenOffice is still alive and evolving is only damaging the general reputation of open source Office software among non-technical users and thus I truly believe that it would be in everyones interest to help the remaining OpenOffice users over to LibreOffice.
And to be absolutely clear I am only suggesting this due to the stagnant state of the OpenOffice project, if OpenOffice had managed to build a large active community beyond the resources IBM used to provide then it would have been a very different story, but since that did not happen I don’t see any value to anyone involved to just let users keep downloading an aging releases of a stagnant codebase until the point where bit rot chases them away or they hear about LibreOffice true mainstream media or friends. And as we all know it is not about just needing a developer or two to volunteer here, maintaining and developing something as large as OpenOffice is a huge undertaking and needs a very sizeable and dedicated community to be able to succeed.
So dear Apache developers, for the sake of open source and free software, please recommend people to go and download LibreOffice, the free office suite that is being actively maintained and developed and which has the best chance of giving them a great experience using free software. OpenOffice is an important part of open source history, but that is also what it is at this point in time.
Wanted to let everyone know that the GStreamer Conference 2015 is happening for the 6th time this year. So if you want to attend the premier open source multimedia conference you can do so in Dublin, Ireland between the 8th and 9th of October. If you want to attend I suggest registering as early as possible using the GStreamer Conference registration webpage. Like earlier years the GStreamer Conference is co-located with other great conferences like the Embedded Linux Conference Europe so you have the chance to combine the attendance into one trip.
The GStreamer Conference has always been a great opportunity to not only learn about the latest developments in GStreamer, but about whats happeing in the general Linux multimedia stack, latest news from the world of codec development and other related topics. I strongly recommend setting aside the 8th and the 9th of October for a trip to Dublin and the GStreamer Conference.
Also a heads up for those interested in doing a talk. The formal deadline for submitting a proposal is this Sunday the 9th of August, so you need to hurry to send in a proposal. You find the details for how to submit a talk on the GStreamer Conference 2015 website. While talks submitted before the 9th will be prioritized I do recommend anyone seeing this after the deadline to still send in a proposal as there might be a chance to get on the program anyway if you get your proposal in during next week.
So this will be the first in a series of blogs talking about some major initiatives we are doing for Fedora Workstation. Today I want to present and talk about a thing we call Pinos.
So what is Pinos? One of the original goals of Pinos was to provide the same level of advanced hardware handling for Video that PulseAudio provides for Audio. For those of you who has been around for a while you might remember how you once upon a time could only have one application using the sound card at the same time until PulseAudio properly fixed that. Well Pinos will allow you to share your video camera between multiple applications and also provide an easy to use API to do so.
Video providers and consumers are implemented as separate processes communicating with DBUS and exchanging video frames using fd passing.
Some features of Pinos
- Easier switching of cameras in your applications
- Multiple types of video inputs
- GStreamer integration
- Pinos got some audio support
It will also allow you to more easily allow applications to switch between multiple cameras or mix the content from multiple sources.
Supports more than cameras. Pinos also supports other type of video sources, for instance it can support your desktop as a video source.
Pinos is built using GStreamer and also have GStreamer elements supporting it to make integrating it into GStreamer applications simple and straightforward.
Well it tries to solve some of the same issues for video that PulseAudio solves for audio. Namely letting you have multiple applications sharing the same camera hardware. Pinos does also include audio support in order to let you handle both.
What do we want to do with this in Fedora Workstation?
- One thing we know is of great use and importance for many of our users, including many developers who wants to make videos demonstrating their software, is to have better screen capture support. One of the test cases we are using for Pinos is to improve the built in screen casting capabilities of GNOME 3, the goal being to reducing overhead and to allow for easy setup of picture in picture capturing. So you can easily set it up so there will be a camera capturing your face and voice and mixing that into your screen recording.
- Video support for Desktop Sandboxes. We have been working for a while on providing technology for sandboxing your desktop applications and while we with a little work can use PulseAudio for giving the sandboxed applications audio access we needed something similar for video. Pinos provides us with such a solution.
Who is working on this?
Pinos is being designed and written by Wim Taymans who is the co-creator of the GStreamer multimedia framework and also a regular contributor to the PulseAudio project. Wim is also the working for Red Hat as a Principal Engineer, being in charge of a lot of our multimedia support in both Red Hat Enterprise Linux and Fedora. It is also worth nothing that it draws many of its ideas from an early prototype by William Manley called PulseVideo and builds upon some of the code that was merged into GStreamer due to that effort.
Where can I get the code?
The code is currently hosteed in Wim’s private repository on freedesktop. You can get it at cgit.freedesktop.org/~wtay/pinos.
How can I get involved or talk to the author
You can find Wim on Freenode IRC, he uses the name wtay and hangs out in both the #gstreamer and #pulseaudio IRC channels.
Once the project is a bit further along we will get some basic web presence set up and a mailing list created.
If Pinos contains Audio support will it eventually replace PulseAudio too?
Probably not, the usecases and goals for the two systems are somewhat different and it is not clear that trying to make Pinos accommodate all the PulseAudio usescases would be worth the effort or possible withour feature loss. So while there is always a temptation to think ‘hey, wouldn’t it be nice to have one system that can handle everything’ we are at this point unconvinced that the gain outweighs the pain.
Will Pinos offer re-directing kernel APIs for video devices like PulseAudio does for Audio? In order to handle legacy applications?
No, that was possible due to the way ALSA worked, but V4L2 doesn’t have such capabilities and thus we can not take advantage of them.
Why the name Pinos?
The code name for the project was PulseVideo, but to avoid confusion with the PulseAudio project and avoid people making to many assumptions based on the name we decided to follow in the tradition of Wayland and Weston and take inspiration from local place names related to the creator. So since Wim lives in Pinos de Alhaurin close to Malaga in Spain we decided to call the project Pinos. Pinos is the word for pines in Spanish