Hi, everyone – I’m Keith, and while I now find myself President of the Kodi Foundation, my initial involvement with Kodi goes way back, as a humble user. Since our 20th anniversary came and went, I’ve been meaning to write a bit of a look back on how we got here – how I got here. This is the story of my – and the project’s – journey.

In the Beginning…

In 2002, a couple guys saw the potential of being able to watch media on their own terms, through what was a new type of device: a “connected” TV device. They wanted to bring many things you’d traditionally need a computer for to the living room, operating it from the couch without needing a keyboard or mouse, controlling events through what they called a 10 foot interface because, well, it was an interface you could sit 10 feet away from.

While special-purpose devices already existed, they usually were things like DVD players, and they weren’t “connected” to anything else or even had the ability to hook up to a network. Indeed, even the TiVo, one of the most impressive devices of the time, lacked an Ethernet port. One of the creators of Samba, an open-source interoperability application to share files between Linux and Windows, eventually figured out how to add a network port to a TiVo; however, even though it was the first mass marketed consumer Linux-based appliance, it was still very much a closed, walled garden.

The Xbox changed all that. In 2001, Microsoft released their first console, leveraging their work on x86 platforms to build out a version of Windows specialized for running games. The familiarity allowed game developers to easily port to the new gaming console; the fact that Microsoft sold them at a loss, knowing they would make it back in game sales, ensured rapid consumer adoption. Luckily for the hackers out there, there were more similarities than differences when compared to something they were already very familiar with: the standard Windows personal computer. Very quickly, folks like Bunnie figured out ways around the security, achieving the holy grail in record time for a console: the ability to run untrusted code.

Long before we had the Raspberry Pi, OpenWrt routers and other tinkerer’s dream “open” environments, Xbox hacking quickly became the fastest route to general-purpose, connected computing. The availability of powerful, affordable, standardized hardware enabled an entire community of hackers wanting to break out of Microsoft’s walled garden and run whatever code they wanted on their Xbox. Emulation, or the ability to trick old games such as NES and Sega Genesis into believing the Xbox is the same hardware, was one of the first big use cases, since we now had something powerful enough that hooked up to a regular TV and offered a standard controller.

XBMC is Born

While many folk worked on getting Linux and various other things running, the project that attracted me the most was Xbox Media Player. This brought the ability to play compressed media to the Xbox: up until this point, only a computer could play these files, and if you were one of the those super early adopters who had network cables strung across your house, you could now even stream it over the network or the Internet (which was definitely not fast enough for video… yet!).

XBMP came out in 2002 as the convergence of two separate closed source apps natively written with the Microsoft Xbox Development Kit by d7o3g4q (a.k.a. duo) and RUNTiME. As many at the time pointed out, since it was leveraging GPL code, it likely needed to be open source. At the same time, another developer – Frodo – had created YAMP as “Yet Another Media Player”; when these projects also converged at the end of 2002, Xbox Media Center, or XBMC, was born. While the initial beta was closed source, the developers quickly open sourced it and it’s been open source ever since.

XBMC v2, released September 2006

(You can even have a read of the original manual if you like!)

I was a dedicated Xbox user at the time, for everything besides gaming. I diligently connected to #xbins on EFnet to find the latest binary release and watched in awe the quick iteration these folks were doing, adding so many incredible features, even supporting things like CD+G for Karaoke. You have to realise how groundbreaking a library mode was, allowing you to peruse your stored media collection while everyone else – even Netflix – was still renting DVDs by post or picking them up in strip mall stores.

The developers themselves deserve a ton of recognition: their tireless pursuit of the best media experience, one that was infinitely customizable and powerful and something they were proud of. While most have moved on to other things, a few, like spiff, remain around since those early days, and are still here to let others know why certain things are why they are and even to still occasionally contribute code.

Success Has Many Parents

But these folks did not do this in a vacuum: as the saying goes, “it takes a village”. FFmpeg, for example, being the Internet’s Swiss Army knife for video, handles so much of the core rendering and playback. In developer parlance, we call these ‘dependencies’ but I prefer to think of these types of libraries as the backbone. Without the ability to stand on the shoulder of these mammoth libraries, XBMC devs would never have been able to focus so much on the skinning and user experience. While XBMC was initially heavily based on mplayer, which is now known as mpv, it has evolved into being closely tied to FFmpeg.

XBMC v9 “Babylon”, released May 2009

There are also countless other contributors – to the code, artwork, forum, and beyond – who absolutely have to be acknowledged. It’s been a long journey from those early applications thru XBMP and XBMC to today’s Kodi, but it’s clear that the application wouldn’t be what it is without the add-on ecosystem, translators, community moderators, volunteer artists, system admins, database admins, and many more.

And while devs have came and gone, the project has lived on, continuing to expand to every major platform it possibly can, even to a couple of smart TVs now. Highly unusual – maybe unique – as an open-source application that started on a Microsoft product before spreading further to so many other, disparate platforms, Kodi has now even come full circle to be back in the Xbox Store (although, somewhat ironically, Microsoft platforms have become increasingly challenging for us because we lack current developer knowledge). We still try to bring all the functionality we can to every platform we can: you’ve got a weird mod file, from your demoscene days? Still got a pile of FLVs? We got you. What about wanting to stream Widevine-enabled Internet content with an open-source project? Yeah, we see you and we care.

So, Here We Are

One of the parts that some don’t realize is since our initial development, we’ve always been a monolithic application, meaning the entire codebase must be ported to the desired platform. I’ve said we’re the “largest open-source multimedia project” for years, half-jokingly, but I mean this by the literal lines of code that it takes to run it.

XBMC v10 “Dharma”, released December 2010

And we’ve stayed volunteer. As a consumer-oriented open-source project, no company has ever sponsored more than a port or a feature; indeed, since the nonprofit foundation behind it isn’t setup to be engaged, at worst companies have simply given some money to an individual to work on a feature. Even this is definitely not the norm: 99.9% of Kodi has been built out of passion and love for multimedia and the desire to have an open-source platform for which people can decide the look and feel and customize to their liking, in both hardware and software terms.

I do have to call out our longest and biggest sponsor and partner, Flirc: their support with this project has been unwavering.

This is truly software on your own terms, and to me, is the essence of open source.

Wrapping it All Up

Sorry this became so wordy, but I’ve been wanting to write a “20 year blog post” for so long, and it’s now been nearly 22 years since XBMP binaries first became available. Time flies. I hope this blog helped you stop for a moment and marvel at how far we’ve come, even if you’ve only been a user since we’ve been called Kodi: the developers see themselves in you and you’re the reason why we’ve kept hacking on this project for so long.

So, let’s raise our glasses for over 20 years, and, yes, we still have MANY lines of original, first generation code! We have to thank everyone who’s been involved and supported us: everyone who submitted a pull request or posted a message on the forum, all the devs who have tirelessly worked on the ‘dependency’ libraries that mean so much to us, and especially the official Team XBMC and Kodi Foundation developers. You folks, who continue to iterate, tweak and build the future of this platform, enable millions to consume media without having to care about how it works and why.

XBMC v13 “Gotham”, released May 2014

And while you’re here, we made a T-shirt to commemorate this moment! It (hopefully) has every team member on it. As we said, we’re volunteers, and the only money our non-profit receives is thru donations, merchandise and the only licensed product we’ve ever been a part of, Flirc’s Kodi Raspberry Pi case.

Take your pick from a dark T-shirt, a light T-shirt, a pillow or something else from the whole store.

Please, purchase something, and help us to keep thriving for the next 20 years; I personally plan our Devcons, and we might not be able to get together in 2025 without help from people like you, so anything you can do to help, we really appreciate it.

Oh, and while I’m asking: if anyone knows any non profit open source lawyers that are prepared to help us with things like trademark registration and maintenance of the Foundation, please send them our way! Legal-at-kodi-dot-tv.

Thanks again. I have no idea what the next 20 years will bring, or whether I will personally still be involved, but, for as long as developers and users are still interested, I know we’ll all try our best to support it.

Keith Herrington
President, Kodi Foundation

(There’s some further history here and here, if you’re interested).

Source link

The Team Kodi PPA has long been a staple for Ubuntu (and similar) users wishing to use more recent, and less adulterated, versions of Kodi.

It is with sadness that the team has come to a decision to officially retire the PPA due to the overhead of maintaining it.

We want to express our sincere thanks to wsnipex for maintaining it for so many years, tirelessly making it available to countless users in the pursuit of making Kodi more easily accessible for the masses on many of the more popular Linux distributions.

The Linux world is shifting slowly to packaged deployments and containers, and so, going forward, the team is looking to utilise Flatpak to hopefully give Linux users a similar, but more maintainable, “pure” Kodi application. We believe this will also allow us to provide greater access from more distributions than the PPA was ever able to.

We have a considerable amount of work to get the Flatpak builds up to the level the PPA provided (stable, nightlies, etc.), so things will develop over time. In the meantime, then, we will be looking to make this more obvious as an installation method, and update our documentation accordingly to allow users to install and use Flatpak builds more easily. You can, of course, still build from source if you’d prefer (just as people do on many other distros); whatever route you choose, we’d appreciate your feedback to help us improve the process, document the steps, and generally make things as easy as we can for our users.

Source link

Good morning, dear reader (okay, “readers” – I’ll be generous, and assume there’s more than one of you!). It’s a beautiful day, the skies are an endless sea of blue, the forecast is a congenial 20℃ spring day – and we’re back in a sealed room, and ready to get going with Day Two of this year’s DevCon.

First up, romanvm, with a talk on computer text encoding – not necessarily Kodi-specific, but something that affects so many applications, particularly when using non-Latin scripts. Fundamentally, all text is abstract, and we simply need an agreed way to represent some symbols in a binary form so they can be stored, manipulated, and retrieved as needed – from 7-bit ASCII to 8-bit national extensions to today’s 16- and 32-bit Unicode with UTF-8 providing variable-length encoding. This presents specific challenges for specific languages, as you need to know the size (number of bytes) for some languages and how e.g. concatenation would work in others.

Next, fuzzard led a discussion about our next release, version 22, Kodi P*: finalise a name, artwork, timescales, and similar. This segued into a broader conversation, co-led by martijn, about our approach to release management and versioning in general: when to merge stuff (do it earlier), the overall process (documentation, ownership, participants, responsibilities), schedules, impact on our users, what even constitutes a new version. Should we offer a stable release monthly, quarterly, annually? How do other projects do it, and what can we learn? Given that we’re not other projects, though, what are the implications for skins, databases, addons, users? And, anyway, could we require volunteers to work in their spare time to rigid deadlines: who wants to volunteer to be managed that tightly, when it’s supposed to be fun?

We then moved to some quick, impromptu, more free-format discussions: yol with infrastructure – what’s physical, what’s hosted, what’s cloud-based, AWS vs Azure vs OSUOSL vs stuff-under-a-desk; keithah with the ongoing implications of content protection; yol again about mirrors and approaches we’ve had from different potential non-standard hosts.

Continuing the quick-fire round, Prof Yaffle took a turn to stir a conversation about futures and how we respond to the ever-changing landscape – the art of the possible. What could we consider: streaming to Kodi, transcoding, content aggregation, Chromecast/Airplay support, headless Kodi (still!), thin client player, Ambilight…

Similarly, phunkyfish and fuzzard kicked off a discussion about increased automation of branching and building of binary add-ons, particularly when building a new major release (which links back to our approach to release management and versioning, above).

After a break for lunch (and a team photo), we continued with the short topics with multiple participants: release manager permissions; guidelines on team membership, permission, expiration, etc.; sponsorship levels. We then moved to a more in-depth conversation kicked off by Prof Yaffle about distributing Kodi on Linux, and the industry move more towards containers instead of repositories (e.g. PPAs): the implications for addons, documentation, user experience, configuration, and so on.

sundermann took the microphone next to talk about the current state of the webOS port: implications of having a limited set of addons built and bundled, webOS store status and submission challenges, remaining playback issues, SSL certification, installation process.

Next to take the floor, chewitt, with an update on LibreELEC. Overall user numbers have declined since the original separation from OpenELEC – an artefact of people moving to different platforms over time, perhaps, such as Chromecasts and Fire Sticks, which are good enough for very many uses – although there’s still a large, active installed base of mostly Pi (mostly 3, 4 and even 5) systems, with some platforms still growing. He explored the status of different platforms – Pi, generic, various legacy platforms, GPU variations (and thus driver challenges) that are out in the wild. Financially, LibreELEC remains solvent with donations via Open Collective, and he talked a little about their experience of that platform versus PayPal and similar. Finally, a quick update on hosting – what’s where, mirrors, status, plans – and some improvements that have been made to e.g. CI/CD tools.

Without pausing for breath, chewitt then moved on to Tvheadend, as a part of the LibreELEC/Kodi “ecosystem” that needed some assistance: it’s a very-widely used PVR backend for Kodi users, and it had fallen into disrepair as maintainers had come and gone. So, working alongside our special guest at this DevCon, Flole998 (who became the de facto maintainer of the project after a “field promotion”), Prof Yaffle (artisan blogger of this parish) and some genuinely-keen new contributors (you know who you are), he’s been working to refresh the forum and website and work towards getting a new release, updated documentation, and generally-invigorated life signs.

A quick dash for dinner, and we’re back to finish up for today, starting the last tranche of topics with phunkyfish to talk about catchup and timeshift for video add-ons. This is all about bringing consistency to how you can pause and navigate various video feeds: local, streaming/IPTV, or even OTA via a PVR add-on. In effect, we can bring all sources into a consistent “channel list”, with subtle differences based on what’s available (e.g. previous/next) but as much common look, feel and behaviour as possible (OSD, pause, stop, what else is on/EPG): local media is fully navigable, streaming might have a 7- or 21-day catchup, dependent on provider, and true live TV simply appears in the traditional linear fashion.

Bringing the day to a close, then, we moved to sports features in Kodi, led by zag (who, incidentally, runs TheAudioDB and TheSportsDB websites): Sports view, metadata for events from TheSportsDB site via a new Python scraper (thumbnail, artwork, description – just as you’d expect). Future concepts include timestamps for significant events (yellow cards/penalties, scoring events, and similar), team lineups (aka the “cast” of the programme), event databases; outside the core concept, we have the possibility of screensavers with scores or schedules, or perhaps Kodi notifications when, say, your team scores while you’re strong-armed into watching something else.

And that’s it for Day Two. A few more topics to roll over until tomorrow, along with some spontaneous coding while everyone is together – but, until then, that’s it for now.

PS it rained later, just as we went out. Boo.

Source link

Can you feel the excitement building in the air? Well, you should, as we are ready for our first Release Candidate for Kodi 21.0 “Omega”!

This is a major release, so, as you’d expect, it comes with many new features. However, with a core change to FFmpeg 6.0, there is a very good chance that we’ll get some regressions As such, please be aware that installing this will most likely break things, and we’d really like your help at that point to identify the problems and get them fixed. Please raise Github issues with full debug logs to help us resolve any issues that you wonderful testers come across.

First, we’d like to make a few targeted announcements.

Android users

Many users may still be on the Kodi testing track for their Android devices. It’s far easier to leave the testing track from a ‘phone-based device, but be aware that if your device is on the testing track, we will be continuing to deploy Beta 3 releases over the next week or so. If you wish to opt out of the testing group, one way you can is by visiting the following link Google Play – Kodi testing track. From here, you can opt out for your Google Play account. This is the best (and possibly only) way to opt out on many Android TV devices (e.g. Nvidia Shields).

With that said, for those adventurous types who would like to help us with testing, you can use the same link to opt in for a Google Play account.

Xbox users

As we do not have a testing track available for Xbox users, we will continue our customary deployment of RC1+ of our major releases to the Microsoft App store. This means that Xbox users will start to see their Nexus deployments updated to Omega RC1.

If you do not wish to receive this update, the only option we are aware of is for you to disable auto updates for all apps/games on your Xbox.

• You can disable auto updates to games and apps in the Microsoft Store through the Store’s settings.

  • Select the Start screen, then select Microsoft Store.
  • In Microsoft Store at the upper right, select the account menu (the three dots) and then select Settings.
  • Under App updates, set Update apps automatically to Off.

The UWP store deployment is scheduled to begin rolling out on the 16th March.

Skin Developers

An API bump has been made for Kodi 21’s testing cycle. See the following PRs for details:

https://github.com/xbmc/xbmc/pull/23926

https://github.com/xbmc/xbmc/pull/23862

This will require changes and submissions to the Omega branch for skins for those who submit to the Kodi Repository. Reach out on the forums if you have any questions/concerns.

Please note, https://github.com/xbmc/xbmc/pull/23927 was merged shortly after Beta 2, and all skins will need to update the GUI version number to be available in Omega repos for any builds (including nightlies) after Beta 2.

The following forum thread has the skinning change info for Omega: https://forum.kodi.tv/showthread.php?tid=372280&pid=3174566#pid3174566.

And now, for the list of changes for RC 1.

Estuary

• A consistency fix was made by @Hitcher to make OSD controls for some buttons to just be disabled, rather than hidden, so the buttons stay shown in a consistent manner.

Filesystem

• A partial revert has been made that fixes some binary addons being unable to correctly parse some XML files.

General

• @ksooo continues to clean up things related to the Video Version feature. This time he has fixed a media name label that would be blank under certain circumstances.
• A first time contributor, @hagaygo, has made a fix to allow users to set a subtitle save path rather than using a temp location under some circumstances.
• @CrystalP submitted two fixes that fix list locations after adding/removing versions/extras and their art related management.

Music

• audiobooks were showing their last chapter with no duration; this has been fixed by @the-black-eagle.

Platform Specifics

• Android:

  • Cleanups to build/package files have been made by @joseluismarti. This helps get the codebase nice and clean for the newer Android SDKs leading into the future, meaning we can more easily backport to Omega when we are making changes for v22 “P” in the future.
  • @thexai has made a change that allows subtitles to be tonemapped to avoid extreme brightness for HDR playback.

• Linux:

  • @fritsch, our resident audio expert, has submitted a change to allow Pipewire to properly identify HDMI for passthrough usage.

• macOS:

  • @enen92 has implemented Hotkeycontroller for media keys on macOS. This allows Kodi to more accurately capture and handle media-related keys and allows us to play more nicely with other running apps on a system expecting to use the same media keys.

• Windows:

  • @thexai has submitted some colour accuracy changes for 10bit SDR/HDR colour mapping.

PVR

• A fix for icons not being correctly refreshed was made by @ksooo.

VideoPlayer

• @popcornmix has tracked down a variable initialisation bug that helps to reduce video stalls under certain conditions.
• A fix that stops incorrect refresh rate changes occurring for interlaced content in some circumstances was submitted by @CrystalP.

Known Issues

You can see open issues on our tracker at https://github.com/xbmc/xbmc/issues. However, to call out a couple of specifics, please be aware of the following if you’re planning on upgrading:

• There are seek issues with multi-part MP4 files on Android (https://github.com/xbmc/xbmc/issues/23673).
• We’ve had reports that the new splash screen looks like an OpenGL corruption; we know, it’s not a problem, so don’t panic. We’d be happy to have your artwork contributions if you’d like to work on an alternative.
• If remote sources are unavailable at startup, users are seeing significant delays while waiting (https://github.com/xbmc/xbmc/issues/23958).

We continue to march forward to what we hope will be a successful v21 “Omega” release.

If you love to live on the bleeding edge, you can get RC 1 from here. Select your platform of choice, and look in the “Prerelease” section. Again, though, please: expect some breakage, don’t use this as your daily installation unless you know how to get things working again, and please share your experiences back with us so we can really get going on those bugs.

There is always a great deal more changes/fixes/updates than we can highlight in these reports, so, as usual, you can see a fuller list of what has been merged between v21 and v20 here. For a more targeted look at the changes made since v21 Beta 3, you can see those changes here.

Source link