INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: cross-platform ready solutions  (Read 1357 times)

Hendrik

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 10772
cross-platform ready solutions
« on: June 21, 2012, 01:11:25 am »

Several people have mentioned that there is a risk of neglecting work on our Windows version if we do another OS.  I don't think it will now slow us down much. 

While i applaud such plans to be independent of Microsoft specific frameworks, there are still many parts in MC, especially in the video pipeline, which leave you tied to Windows.
Replacing those will not be easy at all. I'm mostly thinking about DirectShow for the video pipeline here, and Direct3D for TheaterView.

I could imagine replacing those solutions with cross-platform ready solutions, but that would either mean a severe functionality loss on the Windows side, or that you have to maintain two video pipelines (one for Windows based on DirectShow, and one for any other OS).

Anyway, my main point is that the underlying Framework is not everything needed for a successfull player. If you want to keep the same quality you have now, you will have to create specialized audio and video pipelines on every OS you support.
Logged
~ nevcairiel
~ Author of LAV Filters

MrHaugen

  • Regular Member
  • Citizen of the Universe
  • *****
  • Posts: 3774
cross-platform ready solutions
« Reply #1 on: June 21, 2012, 02:28:57 am »

Anyway, my main point is that the underlying Framework is not everything needed for a successfull player. If you want to keep the same quality you have now, you will have to create specialized audio and video pipelines on every OS you support.

And that have to be rather time consuming right? I have no idea what it REALLY involves, but I've seen my share of codecs and all their options. I've tweaked it until I've almost passed out on the couch. And I've spent hours and tried to learn about how it really works, without even getting close of understanding anything but the most basics. That's video. There are probably also a complicated codec setups for audio as well, even though I think it's a tad simpler.

Then you have Direct3D vs OpenGL, which I guess is the two 2D and 3D renderers you'll end up with? Sorry if I'm confused on technologies and terms here. API's is not my field. Getting this to work on the initial port on platforms is ONE thing. It probably will be lot's of work to set it up for the first time. But how will this work when pushing new builds? Will you end up with a system where you can just do changes on one platform, and the new features (taken advantage of new API functions for example) will automatically be compiled for all the different platforms magically? Example of Theater View features being added, and they have to work with both Direct3D and OpenGL? Will that be so simple as Jim makes it out to be?

I really have no idea, and the reason I'm so skeptical is because of this questions. If this port and development there after is as grim as my fears indicate, I'm looking at 0 new builds for 4-6 months until the initial OSx build is out. After that, the builds and new features will perhaps be developed 20% slower than today, because it have to be adjusted to several platforms. It's not something I'm looking forward to. Of course, you might compensate by hiring additional programmers, but I still can't help but to think that it might have been better spent elsewhere?

Again, this is just my theory! Nothing here is facts. I have NO idea of what's really involved, and how you would deal with this. How much time will be spent. If I'm exaggerating wildly, please explain why this will not be a problem. If you can. To calm me and others down :) Because I'm rather nervous for such a platform adaptation.
Logged
- I may not always believe what I'm saying

glynor

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 19608
Re: cross-platform ready solutions
« Reply #2 on: June 21, 2012, 10:22:48 am »

I agree that porting Theater View over won't be a simple task, with the lack of Direct3D.

However, MC can already use the Quicktime Engine for video playback on Windows, and the MacOS version of the Quicktime Engine is much more capable.  It might not be as challenging as you suspect.  Now, will we have as nice of a solution as MadVR+LAV?  Certainly not.

However, I don't think it'll be impossible if their architecture is really as flexible as they say.  OSX has a lot built into their APIs.

One thing I thought I'd mention briefly...  To replace Red October on MacOS.  Perian has long-been the go-to "codec pack" for Quicktime on OSX that allows the Quicktime engine to support XviD and MKVs and all of that good stuff.  It works very well.  Unfortunately, the Perian developers for their own reasons, have decided to move on to other projects.  The last release of Perian is coming out soon, and then they are going to release the source on GitHub (or Google Code), and stop development.

I actually think this could be even more of an opportunity for JRiver, if they choose to pounce on it.  LOTS of Mac users (including the "internet famous" high profile ones) are going to be looking for something to continue on in Perian's name.  Sure, there's VLC, but that isn't the same thing at all (and the OSX version of VLC, while improved from days of yore, is not the same as the Windows one, and is decidedly clunky on OSX).

All they'd have to do is take the Perian source, modify it to work with whatever video engine they develop for the Mac version of MC (and follow the license terms as they have with the other open source projects they've used), and we can have Red Perian, the Mac OSX corollary to Red October.  I'm sure, if they try, they could even make it much, much better.  Especially with suggestions from wizards like you, Nev, and Madshi.

I do agree, and I think Jim made clear, that the project would not be a simple one.  Their back-end is apparently ready to be ported, but it'll still take a substantial amount of work.  I think that much is obvious.

However, I also think it could generate substantial rewards.  Apple owns the high-end market for PCs and Laptops.  Us old-school geeks might not like it, and I certainly don't like some of what they do myself, but the fact remains that if you limit it to PCs over $1k, they are crushing all the other manufacturers.  These are GOOD customers.  While their overall market share numbers remain relatively low, that's now primarily overseas markets (which are shifting, just look at China), and corporate PC sales.  But, are those corporate PCs the "market" for MC?  Plus, they're now sucking all of the profit out of the market at the low-end too with the iPad.

That's why Microsoft reacted.
Logged
"Some cultures are defined by their relationship to cheese."

Visit me on the Interweb Thingie: http://glynor.com/

Daydream

  • Citizen of the Universe
  • *****
  • Posts: 770
Re: cross-platform ready solutions
« Reply #3 on: June 21, 2012, 04:50:34 pm »

I'm thinking about 2 things right now.

One, if JRiver worked for the last 5 years to make a (big) part of their code platform agnostic, they will have to use that aspect of their code, or else why was that endeavor started in the first place. It's just a matter of where and when, not if.

Two, this matter of when and where is going to be about market share, business and other stuff like that.

I'm sure that along the way there will be lots of technological challenges and solutions to port MC to any new platform. But for the life of me I do not see any kind of (technical) innovation, me, the Windows user will enjoy from JRiver attempting to expand their market share over other platforms. This already looks like a PR battling arena; we have most of the code platform agnostic; but it's going to be a big undertaking. Well, which is it? The previous concerns about dividing resources to something that the Windows users will see no benefit from, still stands.

And, on a total unrelated note - and maybe unfair to everybody, and not really comparable - there is this other thing. I know a team that works by this craze, that whatever they develop has to work on all platforms they support or it doesn't get implemented. What was the end result of that? They got HD bitstreaming working officially only last month. That's XBMC.

I understand there is no comparison, there are different teams, with different goals and different paths to attain those goals. But I'm an end-user. We don't live in a vacuum, we look around at software in general and extrapolate expectations. They are rather grim if JRiver chooses to keep its team the same but take over other platforms too. You guys would need a PR machine just to alleviate fears, let alone make this Windows user enthusiastic about it.
Logged
Pages: [1]   Go Up