Lively Kernel as plugin without requiring a full webbrowser?
Locked
Lively Kernel as plugin without requiring a full webbrowser?
 
|
How difficult would it be to divorce LK from the webbrowser? If some
hypothetical LK plugin were given a handle to a buffer of some kind, how much work would it take to make LK a plugin without dragging all of html with it? The reason why I ask is that LK would make a great game engine GUI in many contexts. Lawson |
Re: Lively Kernel as plugin without requiring a full webbrowser?
|
|
Hi Lawson,
Wouldn't that describe Google's O3D effort?
But they ditched the plug-in earlier this year for a pure JS to Web Browser solution, now relying on WebGL and Canvas. - Darius On Thu, Sep 2, 2010 at 1:26 PM, Lawson English <[hidden email]> wrote: How difficult would it be to divorce LK from the webbrowser? If some |
Re: Lively Kernel as plugin without requiring a full webbrowser?
|
|
In reply to this post by LawsonEnglish
Peter Fraser had been dabbling with Lively running in Batik. When Lively was at Sun one of the first implementations I think was not browser-based. But that was probably in 2007 and Dan, Robert, or Jens can probably better answer this. Otherwise, here are some links:
http://xmlgraphics.apache.org/batik/ On Thu, Sep 2, 2010 at 3:26 PM, Lawson English <[hidden email]> wrote: How difficult would it be to divorce LK from the webbrowser? If some To plug into what? I think it would be nice maybe if an application existed which implemented solely JavaScript and the W3C canvas api to run Lively.
Yes, I think it would.
|
Re: Lively Kernel as plugin without requiring afull webbrowser?
|
|
In reply to this post by Darius Clarke
Somewhat related to this topic:
Another comment:
Philip Weaver wrote:
> When Lively was at Sun one of the first implementations
> I think was not browser-based. But that was probably in
> 2007 and Dan, Robert, or Jens can probably better answer this.
That is correct. The first version of Lively
(then known as "Flair"
or "ScriptBrowser" was implemented on top of the
SpiderMonkey
JavaScript VM and ran on Windows and Nokia
770/800/810 web tablets.
There's a summary of the ScriptBrowser work in
this technical report:
This was a very different beast compared to the
Rhino-based
(and later SVG-based) LK implementation that was started a bit later.
-- Antero
|
WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Great to hear from you, Antero -
...and remarkable timing, as we were just
discussing the possibility of working with WebGL as part of an intern
project this year.
Is there a chance that we could
collaborate?
Is that work open and, if so, under what
license?
Is it dependent on Qt?
What's the best way to follow the
work?
All the best
PS: Could you pass along a suggestion
that the Lively@TUT home page be changed from
to...
and that the link be changed
from...
to...
Thanks!
Somewhat related to this topic: http://livelygoes3d.blogspot.com/ Another comment: Philip Weaver wrote: > When Lively was at Sun one of the first implementations > I think was not browser-based. But that was probably in > 2007 and Dan, Robert, or Jens can probably better answer this. That is correct. The first version of Lively (then known as "Flair" or "ScriptBrowser" was implemented on top of the SpiderMonkey JavaScript VM and ran on Windows and Nokia 770/800/810 web tablets. There's a summary of the ScriptBrowser work in this technical report: http://research.sun.com/techrep/2008/abstract-177.html This was a very different beast compared to the Rhino-based (and later SVG-based) LK implementation that was started a bit later. -- Antero |
Re: WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Hi Dan,
great to hear that you have interest in WebGL as well.
I'm not directly involved in the Lively3D effort
myself.
Tommi has students working on this project
at the
university.
There is no dependency on Qt. The work will
be
fully open, although I don't think there has
been
any discussion about the legalities yet.
So far
that hasn't been necessary.
Tommi should be able to answer these
questions
once he's back from
France (SEAA conference trip).
I'll make sure the web links get fixed. We
are not
updating the web site very frequently, so
this may
take a while.
Best regards,
-- Antero
|
Re: WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Hi Dan, great to hear that you are looking at the same stuff! I was planning to write this email earlier, but wanted to wait for some time so that we would have our new demo (3D minigolf) online. We're still extending it with some new features (mashing up the appearance of the tracks from other web sites + implementing multiple players), so it may take a while before it will be online. Anyways, back to the questions: We've been looking at WebGL for some time, with the idea of building 3D mashup applications in our mind. So far we have looked into a number of different libraries that ease the use of WebGL API, although some of the experiments are built on top of the API directly. Some of the results are online at http://livelygoes3d.blogspot.com/ as Antero emailed you earlier. License is MIT, so that should be very compatible with LK, and as we go, I hope that all the new stuff will be put online as well. The stuff runs directly inside a browser, assuming that the browser is WebGL enabled (number of browsers are not). Qt dependency does not exist, apart from the fact that Qt WebKit will hopefully in the end run also our WebGL apps. As for cooperation, we would of course love it. As our main goal is to build apps, we would be thrilled to use LK capabilities for this. So my proposal would be that we build a 3D version of Lively, the current LK team managing the platform part and we would then implement apps on top of that. tjm Antero Taivalsaari wrote: > Hi Dan, > > great to hear that you have interest in WebGL as well. > > I'm not directly involved in the Lively3D effort myself. > Tommi has students working on this project at the > university. > > There is no dependency on Qt. The work will be > fully open, although I don't think there has been > any discussion about the legalities yet. So far > that hasn't been necessary. > > Tommi should be able to answer these questions > once he's back from France (SEAA conference trip). > > I'll make sure the web links get fixed. We are not > updating the web site very frequently, so this may > take a while. > > Best regards, > > -- Antero > > > ----- Original Message ----- > *From:* Dan Ingalls <mailto:[hidden email]> > *To:* Antero Taivalsaari <mailto:[hidden email]> > *Cc:* [hidden email] > <mailto:[hidden email]> > *Sent:* Saturday, September 04, 2010 10:22 PM > *Subject:* WebGL projects [was: Re: [lively-kernel] Lively Kernel > as plugin...] > > Great to hear from you, Antero - > > ...and remarkable timing, as we were just discussing the > possibility of working with WebGL as part of an intern project > this year. > > Is there a chance that we could collaborate? > > Is that work open and, if so, under what license? > > Is it dependent on Qt? > > What's the best way to follow the work? > > All the best > > - Dan > > PS: Could you pass along a suggestion that the Lively@TUT > <mailto:Lively@TUT> home page be changed from > _...Lively Kernel_ research team at Sun Microsystems > Laboratories. > to... > _...Lively Kernel_ research team at the Hasso Plattner > Institute. > > and that the link be changed from... > http://labs.oracle.com/projects/lively/ > to... > http://www.lively-kernel.org/ > Thanks! > > >> Somewhat related to this topic: >> >> http://livelygoes3d.blogspot.com/ >> >> Another comment: >> >> Philip Weaver wrote: >> > When Lively was at Sun one of the first implementations >> > I think was not browser-based. But that was probably in >> > 2007 and Dan, Robert, or Jens can probably better answer this. >> >> That is correct. The first version of Lively (then known as "Flair" >> or "ScriptBrowser" was implemented on top of the SpiderMonkey >> JavaScript VM and ran on Windows and Nokia 770/800/810 web tablets. >> There's a summary of the ScriptBrowser work in this technical report: >> >> http://research.sun.com/techrep/2008/abstract-177.html >> >> This was a very different beast compared to the Rhino-based >> (and later SVG-based) LK implementation that was started a bit later. >> >> -- Antero > > ------------------------------------------------------------------------ > > _______________________________________________ > lively-kernel mailing list > [hidden email] > http://lists.hpi.uni-potsdam.de/listinfo/lively-kernel > |
WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Here is something interesting, done with Lively Kernel and WebGL:
http://www.cs.tut.fi/~salmin39/webgl/GLGE/canvasexperiments_2/expt.xhtml To try this out you will need the latest Google Chrome started with command line option "--enable-webgl". There is a YouTube video available too: http://www.youtube.com/watch?v=ggLjIJDh2MQ BR, Arto Salminen |
Re: WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Arto, this is great! Glad to see you making progress.
I haven't had a chance to look at the code (is there an easy way?), but I'm curious how you render to the faces of the cube. Is there some way you are "pasting" part so the world image onto the cube, or are you actually using Lively's Canvas display mode to render onto each face? I am hoping the latter, as it would mean that you have essentially all the canvas display protocol (fonts and all) working to the WebGL canvas. Is this true? Either way, it's great to see the beginnings ;-) - Dan ---------------------------------- >Here is something interesting, done with Lively Kernel and WebGL: > >http://www.cs.tut.fi/~salmin39/webgl/GLGE/canvasexperiments_2/expt.xhtml > >To try this out you will need the latest Google Chrome started with >command line option "--enable-webgl". > >There is a YouTube video available too: >http://www.youtube.com/watch?v=ggLjIJDh2MQ > > BR, > Arto Salminen > > >_______________________________________________ >lively-kernel mailing list >[hidden email] >http://lists.hpi.uni-potsdam.de/listinfo/lively-kernel |
WebGL projects [was: Re: Lively Kernel as plugin...]
|
|
Hi Dan,
I think the easiest way to see the code is to press Ctrl+U (see the source) or Ctrl+Shift+J (developer tools) in Chrome. On the left HTML5 canvas there is the experimental canvas version of the Lively Kernel. On the right side is another HTML5 canvas with WebGL (which is used with a library called GLGE). There is a xml file called mylevel.xml that is used to define the GLGE scene. In this xml, the cube and the canvas used as a texture for the cube are defined. In my demo code (demo.js) I use HTML5 canvas API (i.e. function called drawImage(image, dx, dy, dw, dh)) to draw the Lively Kernel canvas on certain places on to the texture canvas of the cube. This drawing is done in every 33 milliseconds. I believe that, the left side canvas could be used as a texture for the cube directly, but because textures are "wrapped" around objects, it would look weird. Haven't test this though. BR, Arto Salminen Dan Ingalls wrote: > Arto, this is great! Glad to see you making progress. > > I haven't had a chance to look at the code (is there an easy way?), but I'm curious how you render to the faces of the cube. Is there some way you are "pasting" part so the world image onto the cube, or are you actually using Lively's Canvas display mode to render onto each face? I am hoping the latter, as it would mean that you have essentially all the canvas display protocol (fonts and all) working to the WebGL canvas. Is this true? > > Either way, it's great to see the beginnings ;-) > > - Dan |
«
Return to Lively Kernel
|
1 view|%1 views
| Free forum by Nabble | Edit this page |