Registered Member
|
This is really two different ideas: video database support in dragon player and better video database in kaffeine. You should submit the two separately and let users vote on which ones they like and which they don't. Also, the ideas are not mutually exclusive, both could be implemented. There are also a number of alternatives posted in the comments, the users might also want to post these as ideas.
Man is the lowest-cost, 150-pound, nonlinear, all-purpose computer system which can be mass-produced by unskilled labor.
-NASA in 1965 |
Registered Member
|
The database stuff should be in (Dolphin) the file-manager, with helper applets to extract/search/filter the preview/meta-data; or in a format translation service to a common presentation layer.
If you run a DNLA server, it allows you to browse your media collection by the meta-data (and folder structure). As said audio meta-data seems richer than video meta-data. As said above, do one thing well. There is also the bit about being forgiving with your (application's) input, but strict with your output - so the data quality improves. |
Registered Member
|
Actually a better version (UI, icons, fullscreen playback) of QMediaBibliothek, using Nepomuk for storing meta information would be awesome. I hope you will pick it up with its developer and move it to playground. Great idea! |
Registered Member
|
I also had a message from a fedora guy who has something like it too. He and the Guy from QMediaBibliothek will hopefully join me.
And if we have the luck we will bring some kaffeine in too so that we have the satelite tv set too. @BlackCat the Idea here much more like starting from scratch and developing something using a lot from others. Of course not really good though but time will by then show which of te applications that clash together here will survive. It's one of the core rules I might bend with the idea("Don't do everything twice. Reuse code") because it will use kaffeine, dragonplayer, QMediaBibliothek, mediaSpy and amarok but in the end it will pay out (i hope).
My blog :
http://adreasdevblog.wordpress.com I work with the head developer on this one: http://techbase.kde.org/Projects/Aki My code repository on github: https://github.com/xxtjaxx/xxtjaxxRepository/tree |
Registered Member
|
That is not really how the brainstorm forum works. You post an idea, and people vote on whether they like that idea or not. Ideas that are voted highly get submitted to the developers. You cannot vote on whether you like an idea when there are several different ideas in the same thread. What if you like the idea of better video management in kaffeine but don't like it in dragon player? Do you vote the idea up or down? If it gets a lot of votes, which idea do we submit to developers? We don't know which idea people were actually voting for. That is why the one idea per topic rule is in place. If you just want a discussion, perhaps we should move this to a different forum, and then based on the discussion there you can post one or more ideas here.
Man is the lowest-cost, 150-pound, nonlinear, all-purpose computer system which can be mass-produced by unskilled labor.
-NASA in 1965 |
Registered Member
|
The Matroska (mkv) container has quite extensive metadata (as well as chapters, menu, subtitle, Cover/attachment) support, though apparently there is not a single application out there that is able to read or write them (apart from their commandline muxer and an apparently rather broken Windows Haali splitter). MP4 seems to have similar capabilities, though if I remember correctly its license is not free. In my opinion these two container formats are the most important in the near future. They are definitely the most feature-rich. AVI seems to be obsolete as it does not support many video (x264) and audio (ogg) formats or subtitles etc. Personally I would love to see a native KDE programm capable of reading and writing (Matroska) tags, chapters and covers. Preferably this would be Kaffeine with its music support stripped off and internet video (youtube, podcast etc.) as well as extended DVB capabilities (e.g. DVB-C) added. |
Registered Member
|
I've never used Matroska, and don't know how widely supported it is, but it sounds like what I'm looking for. Ideally I should be able to put a dvd in my drive, and have the app rip the movie/series, just like a cd, getting all the metadata automatically from the Internet. This seems like quite an ambitious goal I know, but not unachievable. Along the same lines, is there a standard way to identify video files using a 'fingerprint', similar to the way it's done for music? Or maybe it would be easier to just take a fingerprint of the audio track and use that? |
Registered Member
|
I thought Dragon Player was supposed to be simple player for KDE. And this is great because not everyone wants fancy player with tons of functionalities witch eventually one doesn't need and would only slow down the performance. So there should be at least one phonon-based player that just plays media files stored on local media.
|
Registered Member
|
Well, Kaffeine, VLC, Dragonplayer, mplayer, Windows Media Player etc. all support it, but only a few hardware player seem to do, if any at all. They prefer MP4, which probably has something to do with the latter being proprietary since they both do not differ much feature-wise (MKV even seems to be better).
It's not unachievable at all as I can prove since I've written a bash script which does exactly that. The only thing I haven't been able to do is to directly change tags as for that one needs to know C or whatever and write the appropriate code. Right now the only way to change tags is with "mkvmerge" by writing a new xml tag sheet and remux the video with that, thus it is different from how id3 tags are written for example.
Opensubtitles.org calculates a hash to identify videos and associate the proper subtitles. There is plenty of code for uploading and downloading of subtitles, a feature which of course shouldn't be missing in the ideal movie manager/player .
Good idea, though it seems to be better to stay with an already existing standard, unless it has some major shortcomings you need to overcome. And apparently there already is a quite large database of movie "fingerprints". |
Registered Member
|
I didn't use one of these before can you point out one that is using free software?
Don't worry I won't do a thing to your precious dragonplayer just using some of it intestence for my own project.
And by using dragonplayer technology I will be able to do that.
Ok mybe this for a future but for now the task is a program like amarok just for movies. And remember unless you use plugin written in qtscript you aren't able to rip music either.
Opensubtitles.org calculates a hash to identify videos and associate the proper subtitles. There is plenty of code for uploading and downloading of subtitles, a feature which of course shouldn't be missing in the ideal movie manager/player . [/quote] okey I will see what is possible.
Good idea, though it seems to be better to stay with an already existing standard, unless it has some major shortcomings you need to overcome. And apparently there already is a quite large database of movie "fingerprints".[/quote] Lets please keep it low for now amarok isn't quite at its goals by now either. So we will see.
My blog :
http://adreasdevblog.wordpress.com I work with the head developer on this one: http://techbase.kde.org/Projects/Aki My code repository on github: https://github.com/xxtjaxx/xxtjaxxRepository/tree |
Registered Member
|
Agreed, using an existing standard would be better, and subtitle support should definitely be on the roadmap. However opensubtitles doesn't cover the metadata completely (things such as titles, authors, series/episode info for tv show dvds) and I'm struggling to find a database that does have this info (although I haven't looked for that long...), maybe one needs starting? Let's go register dvddb.com
Yes I'm forward thinking a lot... but I'm just throwing ideas in to gauge reactions, to see what people expect from this kind of app. Personally ripping is important to me, and it's the biggest feature I miss from amarok.
I agree we shouldn't get ahead of ourselves, but I also don't think we should be bound by making an amarok clone (although i love amarok very much). A video manager will most probably have different requirements (or video management would probably be in amarok...) Also I should add I can program... and I'd quite like to be involved in this project. |
Registered Member
|
Everythings setup and ready you me and another guy from mediaspy will work on that i got it in my git repository you should find it either on my blog mention in the first message in this ideathread. i called it komedy I doing the basics now please consider talking alot on irc in #kde-devel on irc.freenode.net . My handle is the same as here ping me when you got time.
My blog :
http://adreasdevblog.wordpress.com I work with the head developer on this one: http://techbase.kde.org/Projects/Aki My code repository on github: https://github.com/xxtjaxx/xxtjaxxRepository/tree |
Registered Member
|
i prefer amarok only for music, maybe use some plasmoid but just like a extra.
keep dragon simple and fast, just to play. kaffeine3 was really good, but the kde4 version need work. My idea: work togheter with the amarok-devs and use code from amarok to build a powerfull Video magnament App. my two cents.
El Barto, proud to be a member of KDE forums since 2008-Dec.
|
Registered Member
|
Right, Opensubtitles doesn't cover metadata, but they've associated video hashes to imdb numbers, which can be used to get all the metadata you like. Though in my opinion free software shouldn't use the unfree IMDb. Firstly out of principle and secondly because it would inevitably lead to trouble (license etc.). Fortunately there is a free alternative, Open Movie Database (aka The Movie Database, which, contrary to IMDb, even has an API that allows to search for a movie based on its IMDb ID. Unfortunately this database is not yet as complete as IMDb's, but given some time and the power of many, this shouldn't be a problem. |
Registered Member
|
I think I agree with you on that in principle... but IMDb is very good. I think both need consideration.
This looks good (ignoring the fact that it's incomplete) but it would be even better if OMD saved the hashes so we could query them directly, rather than doing everything using the IMDb ID. I don't know how open they would be to this. Evidently the hard bit is finding a fingerprint that they all agree on. P.s. IMDb doesn't seem *too* restrictive, just some kind of attribution-stly licence: http://www.imdb.com/help/show_leaf?usedatasoftware Edit: I just emailed TMDb concerning hashes, I'll let you know what they say |
Registered users: Bing [Bot], Evergrowing, Google [Bot]