PDA

View Full Version : LightWave CORE multi-platform development



evenflcw
02-04-2009, 09:41 PM
"By default, the plug-ins are multi-platform (mach-o bundle), enabling vendors to ship a single file to support Windows, Linux and Mac, in both 32 and 64 bit versions."

Could we get some more info on this. It just sound so wonderful. What will this mean practically? That it will be worryless to create plugins that work on all platforms? You won't need a mac to compile a mac plugin? You won't have to fiddle with the code for some OS specific **** (unless you used OS specific calls)?

evolross
02-05-2009, 12:01 AM
I was thinking the same thing. Maybe they're referring to an entire plugin being written in Python or another interpreted language that is ran by CORE and then abstracted to the OS.

GregMalick
02-05-2009, 12:59 AM
I was thinking the same thing. Maybe they're referring to an entire plugin being written in Python or another interpreted language that is ran by CORE and then abstracted to the OS.

To me, it's apparent that Python is the scripting language. Much like LScript. The video and Core site says that the SDK provides the same C++ environment that Core was written. And all the methods/functions that the LW development Team used to write Core are available to all plugin developers.

The site also mentions Qt being the multi platform library that is the basis of CORE. Don't know much about Qt - but I'm assuming it's a lot like the WxWidgets cross platform libraries.

My question is whether the first "delivered" incarnation (in a couple weeks) includes this full SDK. If so, I'm totally stoked. I'm not aware that anyone else has ever exposed their entire development library to programmers before. Basically that means that all the stupid limitations in the SDK won't exist. All the functionality in LWCore will be available to all us developers. That means all the limitations we encounter will be in the design of LWCore itself... and that can be improved with each release.

Awesome! :boogiedow

Mr Big
02-05-2009, 07:57 AM
Here's the link to QT http://www.qtsoftware.com/products Qt is a cross-platform application framework. Using Qt, you can develop applications and user interfaces once, and deploy them across many desktop and embedded operating systems without rewriting the source code.

Lightwolf
02-05-2009, 08:07 AM
Could we get some more info on this. It just sound so wonderful. What will this mean practically? That it will be worryless to create plugins that work on all platforms? You won't need a mac to compile a mac plugin? You won't have to fiddle with the code for some OS specific **** (unless you used OS specific calls)?
No, you still have to compile for the specific platforms (usually on that platform as well).
However, for the end user it will be a single file/folder to install (a bundle is basically a special folder structure, on OS X it is just displayed as a single file to the user).

The concept in more detail: http://openfx.sourceforge.net/Documentation/Guide/ch03s07.html

Cheers,
Mike

evenflcw
02-05-2009, 08:14 AM
Ah. Thanks Lightwolf. Doesn't sound as magical anymore but still sounds better than what we had. Although slightly awkward for those not yet introduced to this stuff.

Doctor49152
02-05-2009, 12:46 PM
I am really looking forward to getting in and playing around with the SDK. but I need to finish my regular LW plugin first though. I've been falling behind on my programming because I wasn't sure how CORE was going to effect LWs future.

kattkieru
02-08-2009, 12:42 PM
@Chuck or whomever is watching the dev forum: Will the SDK come with a compiler toolchain that automatically builds for the other platforms, or will we have to compile on each system and then make the bundle? I think using Mach-O bundles for this is brilliant, but it'd be nice to have Ming-W32/64 for cross-compiling from Mac instead of having to buy MSVC.

Lightwolf
02-09-2009, 01:44 AM
@Chuck or whomever is watching the dev forum: Will the SDK come with a compiler toolchain that automatically builds for the other platforms, or will we have to compile on each system and then make the bundle?
I doubt it. You'll still need the to link with binaries specific to the respective OSes, such as runtime libraries.

Cheers,
Mike

Mike Pauza
02-10-2009, 09:59 AM
Hi all,

This is either a stupid question or one of my best, but are what are the core "applications" NewTek mentioned?

Are these just LW plugins, or are they suggesting that standalone apps (for people wo LW licenses) can be developed and sold? Standalone app development may be too much to ask, but NT's wording has got me wondering. With solid IO, such apps developed for the entire 3D animation market and beyond could be VERY interesting.

-Mike

Mike Pauza
02-10-2009, 10:53 AM
What I mean is that if LW suddenly becomes easy to modify in any way we see fit, then enterprising folks could fairly quickly come out with niche 3D apps that compete quite nicely with established applications.

Apps like "animation for niche markets", "modeling for niche markets", scientific visualization, simulation, etc. could be developed using LW Core.

As an example:
My personal background is physics, and while I'm now fairly comfortable with LW as it is, I understand too well that a 3D system optimized for scientists would look much different than LW. With LW Core I can imagine that I'd be able to develop an animation system based on on how scientists think an animation system should work. While not too many scientists currently animate, I bet a well thought out product for them might eventually sell a few thousand copies if it was priced and marketed correctly. While this sort of scenario is small potatoes for NewTek, a single developer could probably make a niche product like this work.


-Mike

GregMalick
02-10-2009, 11:52 AM
Aloha Mike,

I hope I understand what you are asking.

I'm pretty confident that niche applications built upon the SDK for people without LW licenses would not happen. Basically, the SDK has functionality provided by NewTek. NewTek won't (and shouldn't) be giving away that functionality to be sold to people without recompense. That pretty much demands a licensing scheme under their control.

What developers can do, is create specific tools (and hopefully visualizations) that can be sold and used in niche markets by customers that have a LW license.

Pretty much the same plugin business model as we have now.

But maybe I don't understand what you are getting at.

Mike Pauza
02-10-2009, 12:44 PM
With Aloha Greg,

Yeah, I understand that NT wouldn't want to give away anything. But I wonder if they have thought of what I'm suggesting...stand alone apps with licensing fees largely based on how much native LW code is used. Wouldn't the added revenue of a lot of non competing niche apps make it worth their while? This is probably academic, but their wording got me thinking.

I guess a developer could always try to come up with an expensive high end specialty product, then sell that "on top of" a license of LightWave. That would eliminate the possibility of cool & affordable 3D apps based on LightWave though. A pared down animation system for scientists for example would sell a ton better at $300 than $300 + $995 for a copy of LightWave.

-Mike

GregMalick
02-10-2009, 01:51 PM
With Aloha Greg,

Yeah, I understand that NT wouldn't want to give away anything. But I wonder if they have thought of what I'm suggesting...stand alone apps with licensing fees largely based on how much native LW code is used. Wouldn't the added revenue of a lot of non competing niche apps make it worth their while? This is probably academic, but their wording got me thinking.

I guess a developer could always try to come up with an expensive high end specialty product, then sell that "on top of" a license of LightWave. That would eliminate the possibility of cool & affordable 3D apps based on LightWave though. A pared down animation system for scientists for example would sell a ton better at $300 than $300 + $995 for a copy of LightWave.

-Mike

That business model might work if NewTek was willing to do the selling and license administration. Otherwise, how would they know how pared down the lite-app is and how secured (licensed) their technology is. Personally, I think it would be better for NT to farm out the tech work and retain the rights to license and sell that "scientist-lite" version.

Lightwolf
02-10-2009, 02:12 PM
Personally, I think it would be better for NT to farm out the tech work and retain the rights to license and sell that "scientist-lite" version.
Look at VoluMedic as an example, which uses a customized version of LW (as far as LW allows) to make it easier to use.
And the scientific and engineering markets are more likely to shell out more anyhow, so any third party could customize CORE and bundle it with their additions at a higher price - without NT needing to be involved.

Cheers,
Mike

Mike Pauza
02-11-2009, 09:12 AM
Greg: Yeah, NT would need to control revenue somehow.

Mike:
Writing high end apps using core would be really cool, but I can't help but think that mainsteam apps based on technology like core are the way of the future too.

Everyone:
What do people think about some of us starting a list of what we might want to program core to do? Do you think NewTek would listen, or is this really what the beta program is for?


-Mike

GregMalick
02-11-2009, 10:29 AM
Greg: Yeah, NT would need to control revenue somehow.

Mike:
Writing high end apps using core would be really cool, but I can't help but think that mainsteam apps based on technology like core are the way of the future too.

Everyone:
What do people think about some of us starting a list of what we might want to program core to do? Do you think NewTek would listen, or is this really what the beta program is for?


-Mike
Aloha Mike,

I'm sure NT reads all reasonable posts and requests in this Forum. Nothing should stop that.

I joined the hardCORE beta team and I expect there will be especially close scrutiny in those forums. The cool thing is that we expect to have very close communications with the CORE developers and to have some serious influence in the CORE capabilities and UI being worked on.

Anyway, post a list of functionality you'd like to see and I'm sure people in the CoreBeta will echo some items (if not all) on your list.

Mike Pauza
02-11-2009, 11:00 AM
I joined the hardCORE beta team and I expect there will be especially close scrutiny in those forums.


I'd be right there with ya if I had more time or spare cash!
Maybe both of those limitations will change in a few months.

-Mike

Mike Pauza
02-12-2009, 08:46 AM
While I'm not about to commit to writing any plugins in the near future, my interests are in object deformations, particles, rigging, envelopes, hypervoxels, and interfaces.

If Core's Python allows control of most every aspect of these things then I'll be quite happy. BTW, Python can handle loops, multidimensional arrays, and high precision numbers right?

What would also really be sweet is if NewTek provided us some slick online saving, downloading, and installing plugins using thier new LW browser.

-Mike

evenflcw
02-13-2009, 10:10 PM
Everyone:
What do people think about some of us starting a list of what we might want to program core to do? Do you think NewTek would listen, or is this really what the beta program is for?

I can think of a few things I've missed access to in the old sdk. But nevermind those, I think most of those could be met by a "Full Access SDK".

What I'd really love to have is developer support and NT just answering questions. Mainly simple questions directly to do with the sdk such as "is this possible to do? can I get that in this situation? can I set that?". When those questions go unanswered, the developer has to find out himself by coding proof-of-concepts. With the old sdk/lscript these often go bust, with the developer banging his head at 4 o'clock in the morning, cursing the sdk/lscript and vouching never to write another plugin again, because it is such a waste of time and a futile effort, and he feels in the back of his head that he knew it would fail already from the start, nomatter what he tried, so why did he bother(?), and he hates himself for another week for not being sensible enough to know when to quit dreaming and just accept the fact that it can't be done (without wasting even more time trying to come up with more workarounds)! :stumped: :compbeati: :cursin: :bangwall: :cry:

So... just reasonable amount of support for developers, in forum or by mail.