Thanks for the update tbc++. I am anxious to see a good working FOSS tool in this arena. I’ve got a huge project hanging over my head that could use Positron. Please keep us posted.
one way to look at this is that Blender is not an application, but a platform for applications - Positron, as anything else, is free to be on it, but it is still Blender.
e.g. Node based compositing and video sequence editing functionality is at least so far AFAIK implemented in Blender, and not in Positron – but more contributions and efforts are of course always welcome.
if there is something, like for example mesh editing, that video editing does not need, Blender can be made more modular so parts of it can be easier used by other apps. but do also note that Blender is so small and light, that there is often no reason to drop unused parts (even GE is included in the official build, which is still not huge) – and especially for small studios and individual producing it is nice to have it all integrated.
just my 2cents,
without reading all of the thread (just the remarks on this page),
hoping not to offend anyone.
~Toni
TD / Orange
[EDIT: i read the full thread now and saw tbc++ telling about the new version being really a diffrent codebase, so i guess we all are now just interested to see the first peek into it.]
Don’t try to hamper someone’s creativity and innovation. tbc++ has stated that the project currently does not share any code with blender and that the objective is also completely different. What exactly is the gripe you have with this?
There’s no comparing them to Positron either, because the only thing that’s available to actually see on the Positron site is an old version of Blender with a different splash screen. Anything else is hypothetical.[/quote]
Albeit it is hypothetical, there is no reason to doubt that tbc++ has put substantial work into his project. I think it really is uncalled for to dismiss his efforts merely based on the fact that the original verion of Positron was based on the Blender source code. I know there is no proof of what he has done, and you as a contributor to Blender might feel that the credit given is a little unfair/exaggerated, however I think that a little time should be taken to see how Positron evolves before passing judgement. Just a heads up.
tbc++, Im with you 100% of the way, and can see 100% where you are comming from. I was so annoyed with the lack of any good nle’s available, that I was nearly ready to learn to program just to try to get a decent project off the ground - till I found yours (plus I could never find the time).
I agree that the 2 programs should be seperate for now, maybe an idea would be that down the track, if the same interface was used, there could be a pluggin written so that if you installed both blender and positron, positron could appear in blender where the sequencer is now, as well as a stand alone app, but I suppose that depends on how different the code will be by then. However, as you said, pro 3d apps never try to put both in the one project, and in this case, first priority should be for a decent nle, cause who cares if u have to render and use the nle sperately if its going have better useability.
I think the main thing I would like to see (as has been mentioned) is some screenshots. However, I imagin that you would be flat out trying to get the cvs up and running, and I wouldnt blame you if you didnt have time until it is running.
Tell ya what, if I could program, this would be a program I would love to help out, as there is a huge demand for nle’s, yet nothing decent in opensource. The other area I was wanting to see some opensource products is in camera tracking/mo-cap 
You never know, I may yet find the time and learn how to program and help you out down the track.
btw, what language does it look like positron will use? base with c++ and modules in python? ie. what languages do I need to know to input to the project?
Thanks, and keep up the good work,
MicWit
P.S.At the moment, I am doing a digital media cource, and I think we will be using cinellerra as the nle 
if you are interested in getting the name of positron out there, once there is a stable, fully functioning version, Im sure I could con them into changing over 
except pro apps do have both in the same product - you will tend to have 3d particle systems and 3d objects in the same system - you tend not to have full modeling and animation systems inside them for two reasons. The first, and primary reason is that the compositing tools were generally designed in house for movie production companies when there weren’t any adequate tools on the market. These companies weren’t using 3d elements or at least were not using their own tools to create them. Secondly because the software companies who have bought th compositing software don’t want to do the integration work and especially doesn’t want to canabalize its market for its own competing software projects. It has absolutely nothing to do with the usefulness of having full integration. For instance for the low low price of 20 grand you can get Houdini with everything integrated (compositing, NLE, animation, modeling, texturing, particles, UV unwrapping, etc.). XSI and Cinema 4D also have fully integrated compositing.
Also Boris FX and Boris Red are extremely popular plugins which are specifically for integration of 3d elements into compositors such as Adobe After Effects and Final Cut.
No offense, but it looks like you don’t understand the market or the reason for why it has developed the way it has.
LetterRip
No offense taken, I don’t understand the market 100%, as Im not part of the market for a living yet, and are still relativly new. However, not all pro apps do have it, its only a few. Most of them have the nle, not as part of the app, but as part of the package, which may consist of the 3d program, nle, and a texture painter (ie bodypaint like).
Why cant the nle be seperate for now, and then added back into blender later? There are a lot of people out there who want to do video editing without any 3d, so not only will they probably not want to download and install a 3d app to sequence/greenscreen video etc, but they probably wont hear of blender as an nle. You mentioned that there were pluggins for after effects etc, but I imagine that down the track positron will have these features built in, with special support for blender. When the project gets that far, then maybe it would be time to add it back into blender, but in the meantime, I think it will be a fine app for basic video editing.
Please note that these are just my opinions, not from a professional point of view.
MicWit
it is perfectly fine if he wants to develop however he develops it, I was just pointing out that his reasoning wasn’t justified based on the actual market. If it ends up superior to blenders internal then I’m sure some coder in the future might decide to integrate them if there is demand/interest.
I can understand that he was annoyed at the complaints that he apparently recieved, but on the other hand I could understand peoples confusion since not much in the way of original code had been done at that point.
I look forward in seeing how the effort progresses, I’m sure it will ultimately be a useful contribution to opensource.
LetterRip
Is this a NLE or Image compositor software?
About NLE in opensource, there is cinelerra 2.0 it’s an independend branch of cinelerra from herroineWarrior but they keep close relations.
The latest version of cinelerra are more stable and less “heavy”, take a look.
About Positron, well it’s look like Jahsaka, a lot of words, but at least, Jahshaka had its own base code.
TBC++ good luck
My 2 cents:
I can’t really see the point of this project. As far as I can tell there is no benefit in forking the blender code and seperating the sequencer out into a seperate product. The only thing this will accomplish is splintering development efforts, and in the end I see the positron project suffering because of it, not blender. So heres the rub; it seems to me that the only logical reason for forking in this case is to lend some false sense of legitimacy to ‘positron’ as being a product unto itself, apart from Blender. This is, in my mind at least, kind of creepy.
I might feel differently if there was something on the positron webpage that showed any actual development had taken place that would signifigantly differentiate it from blenders current sequencer module. Unfortunatley I have not seen anything like that yet, just a lot of hype.
Regards
Xarf
LetterRip, I’m pretty sure he’s developing an NLE rather than a compositing suite. Node-based compositing, sure, but probably limited.
In reality the industry tends to be split into three separate types of apps, with limited overlap of capabilities: 3D, compositing, and NLE. So if anything there is more division than we’re discussing (the simple 3D/NLE separation).
The fact that XSI and Houdini are rarely ever used for anything but the most trivial compositing work (even in productions that use them extensively for other things) is evidence that, if anything, the industry prefers the three tasks to have separate specialized applications. (If you have sufficient examples of professional productions doing otherwise, I guess I’ll have some nasty egg on my face. :-P)
And I don’t know of any professional production that used a 3D app for NLE. I mean, honestly, what possible benefits could that present? Even in project Orange it doesn’t sound like they’re doing anything with the integration that wouldn’t be perfectly easy if they were separate.
Zarf:
I added the bold. 
I added the bold. :-P[/quote]
Perhaps you can also add a link to the source code for this version that ‘shares no code’ with blender… (you can even put it in bold if you like).
Heres the only linke I could find (a modified version of an old copy of the blender source)
http://cvs.sourceforge.net/viewcvs.py/positron/
(I added the bold. :-P)
If it looks like vapour, smells like vapour and walks like vapour, why it must be…
“the first fully-functioned, feature-comparable NLE that is totally, 100% FREE!”
Cheers,
Xarf
From an earlier post of mine, that you also seemed to have missed:
The website reeks of vaporware and empty promises.
So we agree on that point.
But if any further development happens at all, it will apparently be independant of Blender’s source, making your comment out-dated.
I imagine you replied to the original post without reading any further. And there’s nothing wrong with that–I do it all the time (and have made similar mistakes due to that). I was just trying to give you a friendly update.
Okay, I’m going to be blunt. I really don’t enjoy working on Positron. In fact if I could I’d let the project die. However, everytime I try every other project out there I get totally irritated. Blender is slower than a snail, Jahshaka is buggy as all get out. Cenilerra(sp?) is worse, and is for Linux only (and Mac if you want to hack a few things), and as far as their PR goes, go to their website (where on earth did they come up with “render like a dentist” or “do you want some alsa with your taco”?).
But the fact is, there is no Open Source NLE app on the market that can do what after effects/smoke/avid can do! If there was this project would be done a long time ago. I don’t enjoy getting back from work at 6:00pm and working on Positron till 1:30am just to get up and be back at work at 8:00 the next day. But I’m doing it because I’m sick and tired of Linux getting a bad wrap fot not having a NLE. And I’m tired of using $500 programs that hang every 3 hours.
As for the website, I’ve said it before and I will say it again, I agree, and as soon as I find the time I’m going to change it. Again, I agree, if you have to say that your product is “Professional” than you have problems.
If Jashaka 2.0 is as good as they claim it is, it might (might) just kill Positron off, and I’m fine with that. All I want is a program that will do what I want, and won’t make me want to shoot my computer with a shotgun in the process.
Blenders NLE is nowhere near the quality of Final Cut Pro, but it’s good considering it is bolted on to a 3D app. I think most people would agree there.
I think the Positron sounds interesting. Come on people, why be so mean? Nobody knows if this project will be fruitful, if the app will ever compete with the big boys or whatever, but this guy deserves a pat on the back for at least trying to come up with a great product. Why not just encourage a good idea?
Regarding the development of the built in NLE in Blender I absolutely agree with those who have said that Blender cannot and should not do everything. Blender is a 3D modeling/animation/rendering suite and should not IMO become a huge uber app that replaces everything else. It is beyond the scope of Blender. One app shouldn’t do eveything. Ofcourse, we could also get started on that Blender OS
With that said, I don’t think the Sequencer should be taken out of Blender. There are certain smart things that can be achieved when you integrate 3d data with 2d compsiting and editing, and you need to export and import a lot less.
The more I write in this post I don’t know what to think. I suppose the convergence between 3d and 2d could get really interesting. Maybe making Blender into more of a full blown NLE wouldn’t be too silly. All depends how it’s done though.
tbc++
I support Positron and I respect what you are doing.
I’ll leave it at that.
Koba
Here is the FAQ for NUKE, one of the major compositing tools,
Does NUKE support true 3D compositing?
Yes, NUKE offers perhaps the most feature rich 3D workspace in its class. It lets you build up environments composed of objects (OBJ models, bicubic meshes, bilinear meshes, cards, and various primitives), cameras (to which you can apply imported 3D tracks), and images. You can map images to objects using a variety of techniques including projectoin mapping. You can feed the rendered result of a 3D scene to any 2D portion of your script.
Fusion 5, also has significant 3d support.
Regarding compositing with Houdini - hope you like your hats taste 
“Jimmy Neutron: Boy Genius” and the upcomming “The Ant Bully”, also I’ve seen it frequently refered to as ‘Shake for 3d’.
William,
Blenders NLE is nowhere near the quality of Final Cut Pro, but it’s good considering it is bolted on to a 3D app. I think most people would agree there.
I think Blender has had the sequencer since its original design.
LetterRip
Tbc++:AE isn’t a NLE but a compositor.
And about cinelerra-2.0, give a try, really, there was a good improvement, and it costs nothing to build up a linux system.
But I’m doing it because I’m sick and tired of Linux getting a bad wrap fot not having a NLE. And I’m tired of using $500 programs that hang every 3 hours.
And you hope to do better alone? As somebody said in this thread, you should present a detailed paper about your architecture/design and give some infos about your background.
Why not start from scratch? It isn’t the Blender’s GUI which will make the difference.
Good luck really but avoid the Jahshaka’s style.
Regarding compositing with Houdini - hope you like your hats taste
“Jimmy Neutron: Boy Genius” and the upcomming “The Ant Bully”, also I’ve seen it frequently refered to as ‘Shake for 3d’.
Links, please? I know a guy who’s working on Bardyard (by the same people as Jimmy Neutron). I’ll ask him about their pipeline. I imagine any compositing they do with Houdini is pretty trivial (combining render passes, for instance, which is nearly the only compositing that happens in entirely 3D productions). I could be wrong, of course, but I just can’t see Houdini being used for anything more in the compositing realm.
Insofar as it being called Shake for 3d… have you used Houdini? Or Shake? Both are heavily node based. In fact, their node-based nature is pretty nearly what defines them. So I strongly suspect those comments you’ve heard refer to that shared defining attribute rather than to similar purpose or uses.
As for Nuke, Fusion, etc…
I’ll add another one to your list: Inferno. It also has some 3D capabilities. And I wrote that previous post knowing about all three apps (although, admittedly, I haven’t used those particular compositing apps myself).
The things is, all of this (your links included) actually proves my point even further: the industry prefers separate apps for 3D, compositing, and NLE–even if that means there has to be some overlap of functionality. The industry likes specialized apps for different tasks. Mega-apps are generally frowned upon.
Of course, that’s assuming that the market represents what the industry wants.
Alternatively, it could be the software developers recognize that trying to maintain a gargantuan mega-app that does everything would be a pain in the ass.
Or perhaps everyone recognizes that it’s just generally better to have different, optimized-for-their-purpose interfaces for different tasks.
I would much rather do compositing in Shake or Combustion, for instance, than Houdini or XSI. Shake and Combustion’s interfaces are optimized for the task, whereas Houdini’s and XSI’s… well… aren’t.
Or perhaps everyone recognizes that it’s just generally better to have different, optimized-for-their-purpose interfaces for different tasks.
I would much rather do compositing in Shake or Combustion, for instance, than Houdini or XSI. Shake and Combustion’s interfaces are optimized for the task, whereas Houdini’s and XSI’s… well… aren’t.
If you have a flexible interface, then it can be ‘optimized’ for whatever task you want. Of course it is most likely that the dedicated apps will have the best designed interfaces.
Insofar as it being called Shake for 3d… have you used Houdini?
Nope was just repeating what others have claimed.
The things is, all of this (your links included) actually proves my point even further: the industry prefers separate apps for 3D, compositing, and NLE–even if that means there has to be some overlap of functionality. The industry likes specialized apps for different tasks. Mega-apps are generally frowned upon.
How does that prove your point? They may not be fully developed capabilities but the arguement was against having those capabilities at all. I suspect it is less ‘prefers different apps’ than ‘best tool for the job’. If one tool were ‘the best’ at all of the tasks needed it would be used for all of the tasks.
People adopt Silo and Modo, not because they prefer their modeler and animation tools seperate, but because the integrated subd modeling of most integrated packages is much worse than that of Silo or Modo. People use the renderman renderer, not because they prefer an external renderer, but because renderman is superior on speed, quality, and flexibility in comparison to the internal renderers.
You might have a reasonable arguement that the only way to become ‘the best’ for a particular niche is to be dedicated to a particular niche, but I don’t see you having a case that lack of integration is a selling point.
Another important aspect and probably the most important for a major development pipeline is not just ‘is it good’ but if their are enough people skilled in its usage that if they need to ‘throw more bodies’ at a project to meet a deadline, that there are enough talented people available to meet the positions.
Alternatively, it could be the software developers recognize that trying to maintain a gargantuan mega-app that does everything would be a pain in the ass.
I just don’t think that software compositors have been around long enough that grand unification has been much of an option. It isn’t very far in the past that everyone was using high end dedicated hardware.
LetterRip