Frodo Feature Freeze! - And the October Cycle
The rumors are true. As you may have noticed if you avidly watch XBMC development on Github, we have now entered Feature Freeze for the upcoming release of XBMC 12: Frodo.
For the average user, this means a couple things. First, you should begin to see XBMC Beta releases coming out very shortly. Second, given our fairly well documented open development cycle, there should be very few surprises from now until the release of Frodo. By definition, the features you see now are the same features you will see when Frodo is released with the possible exception of some smaller updates in the default skin Confluence. So if you’d like to see what XBMC 12 is going to look like, feel free to install the October alpha at the bottom of this post. As always, it’s important to remember that the alphas are definitely alpha software that you use at your own risk.
Release Manager
Spaceman Spiff spares no one!
Much as development for Frodo was dramatically different than development for Eden, so too will the upcoming Beta process be different. This year we’ve created the new position of Release Manager, who is in complete control of the Frodo branch from the time we entered feature freeze until the time XBMC has been released. The Release Manager will determine when a new Beta is released, when XBMC is ready to switch to release candidates and Gold status, and whether code changes are fixes (OK) or features (not OK). Needless to say, it is a thankless job and a job that requires a person who is both extremely well versed in the code of XBMC and not afraid to make people angry to get the job done.
Which is why we are excited to say that we’ve selected grumpy developer extraordinaire Spiff to be our Release Manager for XBMC 12. We are convinced that Spiff will do a truly fantastic job in the role, even if he does get a few white hairs in the process!
October Cycle
The current goal is to release the first Beta of XBMC 12 at some time mid-November. At that time, we will review all the many features that have been added since the release of Eden. Likewise, during the Beta cycle, we’d like to continue to highlight the work done by our Google Summer of Code students in one or two posts.
For now, we will simply review the new features submitted during the last dev cycle before feature freeze.
UPnP Update
Just as we hit the feature freeze wire, a significant portion of one of the remaining most frequently requested features was added to XBMC. alcoheca’s GSOC project, a massive update of UPnP to make XBMC a truly capable UPnP server is now part of mainline. While transcoding is not part of the update, all metadata, images, and library functions including watched and unwatched status are now available depending on your client – naturally, an XBMC client is best, though efforts are ongoing to work with the numerous alternative clients that all use slightly different specs from one another.
UPnP in action
Unfortunately, using XBMC as a UPnP server is still not the preferred method for an XBMC client. For one thing, Confluence and all current skins are designed to expect only a local database or a MySQL database masquerading as a local database, which means one would theoretically need to re-scrape the UPnP server into your local library for the TV shows link to show up in the home screen, which immediately cuts out much of the benefit of a dedicated UPnP server.* For another, at present while the server can tell the client what has and has not been watched, a major bug exists in that the client cannot tell the server the same thing. This means, if you mark Big Buck Bunny as watched on your XBMC client, the watched status on the server won’t be updated. For that matter, since you are using the server library, Big Buck Bunny won’t be marked as watched on your client either.
*For early testers, we highly recommend that you NOT Set Content, if that is still an option for you, as it will likely cause problems in the system.
As such, for now continuing to use SMB or NFS shares on your XBMC clients is still the recommended route, but we will keep you updated on the progress of UPnP, including any bug fixes throughout the Beta process. Additionally, skilled skinners, we are excited to see if and how you incorporate UPnP nodes into your design. To see an ongoing discussion of how UPnP will be improved now and in the future, feel free to visit alcoheca’s GSoC forum thread.
Advanced Filtering
If you have been reading the monthly cycle updates, you know that Montellese has been working on major improvements to the underlying XBMC library. A significant culmination of that work has been added this month in the form of the newly updated Filter.
The New Media Filter
Now those of you with an extensive library will be able to filter your library by any number of new, simple methods, including title, rating, year, actor, whether the series is ongoing, etc., and you can of course combine as many of those methods as you like for laser focus. Additionally, this filtering will occur in realtime, so the moment you narrow the span of years, your library will immediately narrow to meet the specification without any need of hitting the “OK” button.
Many More
Needless to say, because October was feature freeze month, numerous additional features have been added, including…
- numerous JSON-RPC updates, including PVR support, optimizations, better conflict handling, and better webserver support by Montellese
- PVR timeshift support (if the backend supports timeshift) – hat tip to non-team member Margo for this one
- improvements to the OMXplayer, the media player of Raspberry Pi by gimli
- Additionally, OMXplayer can now play back DVD ISOs if users have gotten the MPEG2 codec license thanks to user epideme
- the CD ripper now works in the background thanks to Spiff
- panoramic photo panning support by Montellese
- Dirty Region support is now on by default by TheUni
- Further improved image handling in the database, allowing one to select fanart, poster, banner, and thumb, rather than forcing the choice between downloading posters and downloading banners – kudos to Jonathan Marshall and Montellese for this work.
As always, this is a mere selection of numerous changes from this month. For a full list, feel free to check out the GitHub changelog. Or, if you are feeling a bit brave and a bit lucky, just start downloading.
- Windows and OSX
- Apple TV and iOS installation instructions
- Given the additional complexities of distribution, builds for Linux are not currently available at XBMC.org. A number of developers are currently hard at work on the problem however. With luck, we should finally have a Linux PPA available by the time the Betas begin rolling out.
@Bill
client->server watched status will not be in Frodo
Why the article says this if it’s not available?
“While transcoding is not part of the update, all metadata, images, and library functions including watched and unwatched status are now available”
I don’t think it’s a non issue…It’s quite a big issue, maybe you’ll be able to do the streaming well, but it’s still using twice the resources. It will be possible to find a workaround for the future or the people with separated SMB will always have to use MySQL?
Do you really think everyone has the luxury of 100Meg ethernet around their home? I do hope the UPnP implementation on XBMC can be made more efficient in the future, as doubling the network bandwidth consumed by media playback for a modicum of user convenience doesn’t seem like a great trade, or a smart solution.
@Charlie
I think having more than one dedicated HTPC in a single home is a luxury. I think having a TV is a luxury.
I really don’t see the point in bitching about something that has been apart of XBMC since the xbox days. It’s one of many options. If it doesn’t work for you, use something else. MySQL isn’t going away anytime soon. UPnP features in XBMC are currently evolving. Believe me, we’re not going to pull MySQL unless the replacement is reasonably on-par.
The point is this aticle and the accompanying video give the impression that MySQL will be making way to UPnP, that MySQL support may not continue in the future. I admit that UPnP looks like a nice idea, but not if it means doubling network bandwidth which is just a crazy solution.
I really don’t care if UPnP has been a part of XBMC since xbox days, if UPnP is now being pushed as the premier way to share content and metadata then it has to be said that UPnP on XBMC has a major drawback in terms of network usage efficiency that should be eliminated. With XBMC players costing as little as $35 a pop (Raspberry Pi) it’s obvious that sharing could become more common in future, and UPnP as it currently stands is a sub-optimal solution that may cause more problems than it solves.
@Charlie
1. If you want to share libraries and you want to install the server on box A but host your media on box B then that’s what’s going to happen. It’s not UPnP’s problem that you want to use a server and a NAS as two different machines. It’s not UPnP or XBMC’s problem that you have a horrible local network that makes 1995 want to cry. Complaining about a “Box A/Box B” set up is like complaining that 1080 doesn’t play well over wireless B. Just about every actual “media server” software works like this when you -choose- to set it up like this. —Even then, very few people will have an issue here.
2. Tons of people run an XBMC box on all the time as a file server. They do it for automated downloads, they do it to save a device, and they’ll probably do it even more because of PVR. Most people who install a MySQL set up for XBMC are doing so on actual desktop computers, not on stand-alone NAS boxes, and those computers have to be on all the time for MySQL to work. —Zero issues here for those people.
3. We typically don’t give promises, but a headless XBMC with UPnP that can be installed on many NAS drives (anything that runs linux, even a $20 PogoPlug) is almost a given. —Zero issues in this situation.
You got one little factoid about how two data streams are used in some situations and now you’re ranting about how UPnP is flawed. Even if it never improves from today, calling it a “sub-optimal solution that may cause more problems than it solves” is an uninformed and overblown statement.
Instead of jumping to conclusions maybe you should try learning more about what’s going on.
Support is going on.
Sorry; here’s the link:
https://github.com/pipelka/vdr-plugin-xvdr/issues/8#issuecomment-9438522
I don’t know if this PVR support requires changes to the skin to have a “Live TV” option on the main menu.
If it does I hope someone will update PM3 HD since it’s the only skin that suits my needs.
@nedscott
So remember to write this when you talk how great it is uPnP to replace mySQL. Because not everyone have bandwith to WASTE (yes, using twice the bandwith when you can avoid it it’s a huge waste of resources and not everyone has the possibility to wire all the house), and even if you have enough to do it it’s still a waste.
The problem i think it’s because with xbmc we’re trying to adapt a technology that it’s not think to be used to share a library but only to access remotely and easily to other sources.
The most important question is if there’s a way in the upnp standard to prevent this waste that they can implement, because if there’s not upnp it will never be a real replacement of mysql, just a different thing with different purpose
@eema
I need to apologies here for being a little worked up about this. Let me give you some background on the issue:
When XBMC needed a new MySQL guide because the lifehacker one was outdated, I pushed for a new guide on the wiki (and ended up making most of it), and advertised it all over the place and let everyone know. I love MySQL sharing, despite it being a pain in the ass to set up. When it was first internally suggested to have UPnP replace MySQL I was the first one to say “but UPnP sucks!”. However, the rest of the team showed me that UPnP has all sorts of abilities we haven’t even tapped yet. Then alcoheca did his GSoC project and improved the crap out of our existing UPnP server and told us all sorts of other things that it will be able to do (look forward to XBMC v13 being able to use UPnP controlling for XBMC boxes being able to control each other, or “Play to” your UPnP TV without having a dedicated XBMC box).
Given the current hard work that were done on UPnP in the here and now, I guess I got a little defensive about the feature in my replies here. It’s also a bit frustrating to see people diss something before it’s “done” (as far as a MySQL replacement goes). Rest assure, we want to do library sharing -right-.
To reply to eema specifically, you say “The problem i think it’s because with xbmc we’re trying to adapt a technology that it’s not think to be used to share a library but only to access remotely and easily to other sources.”, this isn’t actually true. UPnP is incredibly feature rich and can do “two-way” communication between clients and servers for some time now. It’s far more than a “read only” protocol. UPnP can even be used to sync just metadata, allowing for other protocols to be used for the actual video itself.
There’s a few ways that this “waste” (which, I still gotta nitpick here, is still only a problem in some setups, as many can have a UPnP server on the NAS/storage itself and thus not wasting any bandwidth) could be dealt with. How it will be is up to the developers working on it, so I don’t really have a technical explanation for people now (I’m still learning all about it myself), but I’m willing to bet we’ll be happy with the end results. These guys are far more critical of their own work than we are of them.
@nedscott
To start, i have a HTPC that is also my server with all the material in it, so the problem doesn’t affect me directly. I have a MySQL share at my home (over wifi-N because i can’t wire up all my house) and it works very well even with 1080p material.
When i discovered about the real potential of uPNP (i always considered it something not useful before reading about it here and in the forum) i thought it was a great technology…
But noone talked about this problem, you called a “non-problem” and noone told if in the future it will be possible to do something to fix it.
About the UPnP, the protocol is more than a read only protocol for sure, but it’s not even designed with this purpose. Probably can be adapted very well and produce a great result, but without solving this it’s not actually a real replacement.
About the problem itself it’s not everyone’s problem, sure… But the external storage is the theorically better solution for a shared material, i talked 2 weeks ago in openelec forum about adding mysql support and they consider every different solution than a network nas a blasphemy xD
@nedscott
I think we (users) need an updated wiki page, with pros and cons against mysql and the state, with the list of the implemented feature and the one it will be implemented in the future (for example, the view/not view part is already implemented or not? because the article says yes but someone in the comments says no)
@eema
I definitely plan on updating all the wiki pages on UPnP, and doing a comparison table for various sharing options.
As for the watch status sync feature, IIRC, the code is technically in XBMC v12 (Frodo), but is currently broken and won’t be fixed until afterwards. I’ll be sure to make that clear on the wiki as well.
Its long been talked about that XBMC should have a server component that can run headless, and true client server support and syncing, without the MySql install and config steps that are beyond the ability of many users. I don’t know much about Upnp either, but if this is the first step towards that goal, it is very welcome.
Whichever solution is the ‘right’ one should be chosen, it will benefit a LOT of users when XBMC gets this capability.
I already have a linux fileserver running, and I’d love to be able to run XBMC headless on that box…it would really make things easier. I hope it’ll be implemented in the future :)
In this thread they have built a headless xbmc server
http://forum.xbmc.org/showthread.php?tid=132919
I´m a bit too lazy to try it since I have a space rasperry pi running the library updates, but when some genius will build it for unraid I will jump on that :)
Don’t forget about those of us that have a Windows based server running. My utopia would be a lightweight application that runs as a service that will act as the head end for all the XBMC devices in my house – managing the library and watched status. The key is to run as a service in the background, which I think would benefit users with a server as well as users with a client that hosts the data (similar to Plex, but with all the additional benefits of XBMC).
Could we get a gold release for Christmas? That would be great, Santa!