RandR 1.3 Explained, Demonstrated

[phoronix]

Phoronix: RandR 1.3 Explained, Demonstrated

The first X.Org talk at FOSDEM 2009 was on version 1.3 of the Resize and Rotate extension. Matthias Hopf talked about RandR 1.3 and then Keith Packard demonstrated the transformations and panning operations using this soon-to-be-released version of RandR. Among the features for RandR 1.3 are querying state without output probing, multi-monitor panning, display transformations (translation, scaling, rotation, projection), and support for standard outputs.

http://www.phoronix.com/vr.php?view=13472

[zgold550]

Does anybody have any details on the problems with the planned GPU-Object support? Aka how far along is it, what major problems did they run into that made them not include it in 1.3. Do we need a whole new protocol and major release to include this (and hence many months) or is this something that may be available in a month or two.

[Michael]

Does anybody have any details on the problems with the planned GPU-Object support? Aka how far along is it, what major problems did they run into that made them not include it in 1.3. Do we need a whole new protocol and major release to include this (and hence many months) or is this something that may be available in a month or two.

It hasn't been coded yet.

[zgold550]

So what happened in between a couple months ago when it was still in the plan for 1.3 and now? Any major technical issue which made x devs realize that it would take longer than planned, or it was just lower in the priority list and they didn't get to it.

[bridgman]

My understanding (which could be wrong) was that in order to fit multiple GPUs into the RandR model you needed something like "shatter" to split drawing commands across GPUs. At XDC last April I think GPU objects were already pretty much out of 1.3.

[ssam]

It would be great to have a link to the video as an ogg file (or even a mpg or something)

[Kano]

Well you could convert that flv to mp3, basically you would only need the left channel as you see nothing on the video and sound is only from left ;)

[perpetualrabbit]

Hi there, I'm sorry if I offend anyone, that is not my intention. But I have to say that this video and some others I have seen on phoronix are not very good at all.:(

In this one the only thing I can see is a table and an unreadable projection screen. Partly because the table is in the way and partly because it is too bright and too little contrast.
Why do you not point the camera to the speaker at least?
Is it possible to choose a better position to shoot the video, like a seat at the front right before the prejection screen?

Also the audio is pretty noisy and I can hardly hear what is being said.

I love the fact that you make videos of these interesting talks, but with this level of quality there is not much point to it.

[mgc8]

Hi there, I'm sorry if I offend anyone, that is not my intention. But I have to say that this video and some others I have seen on phoronix are not very good at all.:(

I was just about to post the same, all (at least the ones I looked at) the latest FOSDEM videos seem like this: a table, noisy/poor sound, sometimes a hand waving, and a white place where supposedly something is being projected. Couple that with the terrible accent some of the speakers have and you get a useless waste of bandwidth!

Don't get me wrong, it's not that I don't appreciate the effort, however since the "video" is basically a static uninteresting picture, why not save the effort and simply present an audio recording together with some pics? Really now... just so you can say you've got "videos"? Dissapointing! Make it ogg/mp3 and that's it.

[CrystalCowboy]

I wasn't that thrilled with the video, or the audio track. To turn this into a positive suggestion: I would prefer a transcript over the poorly-miked audio, and if the speakers had any presentations (PowerPoint of OO Impress) that would be more useful than the current video.

[szczerb]

Whoever organised it should've given the press audio pulled from the audio console. It's as big of a problem as pluging any analog devices via a pretty standard cable. A f**cking tape recorder would suffice and made the audio understandable.

[bridgman]

I doubt that there was an audio console to pull from.

My recollection from last year is that the devroom sessions were held in classrooms, where the audio system is a sign saying "speak clearly".

[szczerb]

Oh....I assumed that they had a microphone, which had to be hooked up to at least an amp of some kind, most of which have other outputs then just for the speakers...

I mean, come on, every institution with a room for such events (or even just lectures) has audio equipment.

[bridgman]

Nope, this is just thousands of open source folks piling into a 60's-era university and taking over the classrooms. There are probably a couple of halls with real sound systems but since there are 20 or more sessions going on in parallel...

[szczerb]

ehh....nevermind. I just can't seriously believe that they couldn't get better audio with minimal effort...it bugs me because I really wanted to watch or at least here it...

[bridgman]

Michael covered the 2008 FOSDEM sessions, probably with the same equipment, and got pretty good results. It sounds like there was a bad mic cord or something which you wouldn't know about at the time the recordings were being made.

I imagine that having been burned by equipment this year Michael may arrange for some kind of monitoring mechanism for future sessions, unless we all bitch and complain so much that we discourage him from covering FOSDEM in the future (FAIL).

Even with the audio problems I found the recordings pretty useful. That said, some kind of "warning, don't play this with headphones" note on the web pages might still be a good idea ;)

[Michael]

Michael covered the 2008 FOSDEM sessions, probably with the same equipment, and got pretty good results. It sounds like there was a bad mic cord or something which you wouldn't know about at the time the recordings were being made.

I suspect that having been burned by equipment this year Michael may arrange for some kind of monitoring mechanism for future sessions, unless we all bitch and complain so much that he decides not to bother covering FOSDEM in the future (ie we successfully shoot ourselves in the foot as a community).

Actually, it was better equipment (both for audio and video) than last year at FOSDEM. However, I think I tracked down the culprit and it was an adapter for hooking up the quarter-inch microphone jack to the netbook. I was using the same microphone and adapter for recording audio at XDS last year, but at that time the adapter seemed to work fine and not cause feedback. I think that accounts for the poorer quality audio than anticipated. I was in fact monitoring the audio levels at the start of a few sessions via headphones and at that time it seemed fine, but that's maybe when the adapter was being friendly.

Though with Audacity the audio from this year can still be cleaned up, but I haven't got around to that yet as I just got back last night.

The audio board used during the talks is quite simplistic and I ended up checking but don't recall seeing any audio outputs but only inputs. Next year perhaps I'll just get a lapel microphone setup for the speakers to use and then first interface with that and then pass that feed onto their audio board. Or just use a different microphone/adapter :)