Forget about the other things. THIS is what needs to be done

[kurkosdr]

Pretty much everyone who is a frequent visitor of the video-related forums of the web and user of VLC, knows what really needs to be done to make it better.

Surprisingly however, none of these requests can be found in the "feature requests" secton of this forum, which seems to have become a wish dump (aka everyone dumps his wish to support the X feature of the Y proprietary hardware, ignoring how much work this needs and that is of no real benefit to the majority of users).

So, here you get it, a full list of all the things that really needs to be done. I bet there isn't a VLC user on planet earth that wouldn't despesrately want to see these features in VLC sometime in the future. The list:

1. Support for multicore CPUs.
There are people with pentium Ds and other quite powerfull CPUs, but they have problems playing 1080p -or even 800p- videos, just because VLC's codecs aren't threaded. High definition MKVs and MP4s are abundant on the net nowadays, so we need VLC to be threaded now more than ever. There is a threaded version of ffmpeg called ffmpeg-mt. So why not help the ffmpeg group perfect it and then put it on VLC? Is there any chance of this happening sometime in the future?

2. Support for DXVA.
Again, most modern (and not so modern) graphics cards have support for DXVA, a nice API that allows to offload some of the decoding to the GPU. So far I've found only this. Does this means DXVA is scheduled for intergation in the future?

3. Some basic support for unprotected Blurays.
Nothing like Menus or BD-J, just a basic access module that will allow playing the titles of the Bluray. Making BD-9s from MP4s and MKVs and other high def videos is an easy thing to do with software like TsMuxer or MultiAVCHD nowadays, so it would be nice to be able to play them on VLC, apart from our PS3. In the current version, you need to manually point to each m2ts file, instead of it being done automatically. There is already a "BD" access module, but it's practically hidden from most users. So why not just put it on the "Open Disc" menu where eveyone can find it? Something like this. It should be easy to do, so why not?

If these things are done, then VLC will arguably be the best opensource player, ever. Then, you might considering things like supporting keyboards or subtitle plugins and italic font on subtitles, and other less important stuff found in the feature requests. Most users would agree the above three things should be the highest priority.

[VLC_help]

1. There have been multiple discussions about this. You can already use ffmpeg-mt with VLC if you build VLC by yourself. But for official builds this transition might not happen for a while.

2. I don't know about current state. As I have stated earlier, OpenCL or similar is better option for DSP accelerated processing in multi-platform software project.

3. Adding BD access module to GUI makes more sense when it works better.

[Jean-Baptiste Kempf]

2. there is already some DxVA code in VLC, it might be activated for 1.1

3. we need more samples to activat it.

[kurkosdr]

2. there is already some DxVA code in VLC, it might be activated for 1.1

Please yes!

3. we need more samples to activat it.

Ok then...

1. There have been multiple discussions about this. You can already use ffmpeg-mt with VLC if you build VLC by yourself. But for official builds this transition might not happen for a while.

Yes, but it needs to be done. And the longer you leave it in closed beta and don't let people try it (most people don't know how to compile), the less feedback you will get, and the more the transition will take.

I strongly, strongly propose to make a "VLC 1.1 Multicore (beta)" edition, to be offered next to the official "VLC 1.1" version. It won't hurt the official release (and it's stability) and you will get precious feedback from early adopters like me for whatever bugs there might be on the multicore edition.

[Jean-Baptiste Kempf]

I strongly, strongly propose to make a "VLC 1.1 Multicore (beta)" edition, to be offered next to the official "VLC 1.1" version. It won't hurt the official release (and it's stability) and you will get precious feedback from early adopters like me for whatever bugs there might be on the multicore edition.

I strongly propose that you take 10+ hours of your time every week to help us...

[kurkosdr]

I strongly, strongly propose to make a "VLC 1.1 Multicore (beta)" edition, to be offered next to the official "VLC 1.1" version. It won't hurt the official release (and it's stability) and you will get precious feedback from early adopters like me for whatever bugs there might be on the multicore edition.

I strongly propose that you take 10+ hours of your time every week to help us...

I can get some friend of mine to compile the multicore version. Then it's up to the site admins to put it on the videolan.org/vlc site so people can find it, test it, and provide help either in feedback form or in code.