• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

PS3 Web Browser Discussion - big upgrade rumoured for long time, but no concrete news

Beam said:
Since Sony and Google are close now. Is there any possibility, that Sony can use Crome as a browser?

They apparently are not using Chrome as the javascript engine ported to the PS3 is generic webkit, it's a standard with JIT performance similar to Chrome! http://www.webkit.org/projects/javascript/index.html

When will we see it? Who knows. This is the reason for trying to determine where they are in the porting process with GNU license disclosures of Webkit or supporting libraries.

This is not set in stone and I am only an outsider looking in trying to determine what Sony is doing by looking at various releases and GUESSING what they mean. I would appreciate others looking too and commenting on this.
 
SolidusDave said:
but not only pointer support but the whole Minority Report tech demo thing:
2nvezuv.jpg

wjc7jn.gif

Looks like this can be done, may have been done using Gstreamer plugins with PSGL or possibly cairo (Pixman) support (at least the video warping and video keying). Gstreamer was available in 2005.

http://www.collabora.co.uk/projects/

Telepathy
GStreamer
D-Bus
Webkit
Farstream
libnice
PiTiVi
PulseAudio
 

itxaka

Defeatist
How would the licenses affect whatever modifications that Sony does to the source code? They would have to share them wouldn't them?*

*QT and Webkit are under different licenses so those would be free to use, not sure about any other libs.
 
http://blog.mfabrik.com/2011/02/25/webgl-on-android-webkit-browser-demoed-by-sony/

WebGL demoed on Android webkit browser

3D accelerated effects are no longer the monopoly of desktop beta browsers (Firefox 4, Google Chrome 9… I am looking at you). Sony demoes Javascript WebGL on its Xperia handset. WebGL would be major turning point for mobile gaming, as it would lower the barrier of enty for building cross-platform mobile games.


"So, Chrome has WebGL on by default ... if you have hardware OpenGL 2.0 support"
, which platforms that can play games like the above has. As do all Playstation certified Android platforms. The NGP and PS3 have PSGL and custom support for WebGL would be needed.

I believe Sony is going to support WebGL games and will want support for it in all platforms. That we were told a year and one month ago that the PS3 browser would be able to display 3-D inside the browser in a year meant that the Sony employee was told we would have a HTML5 WebGL browser. 3-D and games are the primary use for WebGL and none of the handhelds (xperia play) can display 3-D.

http://www.gamedev.net/page/news/index.html/_/gdc-news/khronos-annouces-webgl-10-at-gdc-2011-r11899

3rd March, 2011 – Game Developers Conference, San Francisco – The Khronos™ Group today released the final WebGL™ 1.0 specification to enable hardware-accelerated 3D graphics in HTML5 Web browsers without the need for plug-ins. WebGL defines a JavaScript binding to OpenGL® ES 2.0 to allow rich 3D graphics within a browser on any platform supporting the industry-standard OpenGL or OpenGL ES graphics APIs. WebGL has the support of major silicon and browser vendors including Apple, Google, Mozilla and Opera with multiple browsers already shipping with WebGL implementations including the beta releases for Mozilla Firefox 4.0, all channels of Google Chrome 9.0, an Opera preview build, and Apple Mac OS Safari nightly builds.

WebGL leverages the pervasive availability of OpenGL ES 2.0 graphics on almost all browser-capable desktop, mobile and embedded platforms and the recent developments in Web technology including the massive increases in JavaScript performance. The ability for Web developers to directly access OpenGL-class graphics directly from JavaScript, and freely mix 3D with other HTML content, will enable a new wave of innovation in Web gaming, educational and training applications and graphically rich user interfaces to make the Web more enjoyable, productive and intuitive.

There is already a thriving middleware ecosystem around WebGL to provide a wide diversity of Web developers the ability to easily create compelling 3D content for WebGL-enabled browsers. These tools include: C3DL, CopperLicht , EnergizeGL, GammaJS, GLGE, GTW, O3D, OSG.JS, SceneJS, SpiderGL, TDL, Three.js and X3DOM. Links to these authoring tools and WebGL demos can be found at www.khronos.org/webgl/wiki/Main_Page.

"WebGL enables an entire new class of applications on the web. Being able to take advantage of first-class 3D hardware acceleration in a browser on both desktop and mobile allows web developers to create compelling and immersive experiences for their users," said Vlad Vukićević of Mozilla and chair of the WebGL Working Group.

WebGL Standard released March
Sony Xperia Play released March..................(Supports WebGL)
PS3 Portal 2 released April...........................(uses Webkit in game, does not need a completed active webkit, just parts in firmware)
New PS3 HTML5 browser soon.......................(Supports WebGL from 1 comment)

The within a year 3-D from inside a browser on PS3 puts the projected browser release about this time which appears to tie into multiple projected industry timetables for a March WEBGL browser completion. PS3 WebGL projected for this time a year ago does not mean it will be done on time.

The following are hardware issue release dates.
NGP release Nov 11, 2010............................(Supports WebGL not confirmed)
S1 tablet September 2011...........................(Supports WebGL)
 
DrXym said:
A couple of random thoughts:

a) Webkit doesn't need QT and it would be mad to port QT to the PS3 just for some browser. There are impls of WebKit that go through QT, GTK, Cocoa etc. I imagine the best approach for the PS3 would be to make some thin widgets staight over OpenGL ES
b) It's not the disk foot print size that matters, but the runtime footprint. i.e. how much it consumes when running standalone or in a game. I expect Webkit has reasonable performances thanks to it's use in Android / iOS.

I didn't give thought or answer your very valid points above in my posts.

a) answered for the most part but several others appear to be under the impression that my explaining how Netflix created their application was how Sony was porting webkit. This is not necessarily true.

It is not known what low level support libraries are going to be used beyond Cairo. Sony will want WebGL support for Cairo so a OpenGL backend for that. My best guess and this only Google researching a few key words (Gstreamer, Cairo, Webkit) brings up this:

The GTK port of the WebKit HTML rendering engine has gained support for the HTML5 video element. The media backend, which uses GStreamer, was implemented by Pierre-Luc Beaudoin of Collabora. Developer Alp Toker integrated the backend with GTk/WebKit's Cairo graphics pipeline, making it possible for the video content to be embedded in SVG and manipulated with CSS and JavaScript.

Plans for GStreamer-based HTML5 video support in GTK/WebKit were initially revealed by Toker in September, when he published slides from his presentation at LinuxConf Europe.

As some of you may recall, support for the HTML5 video element was implemented experimentally for Firefox back in August.

The GTK/WebKit port appears to be maturing rapidly and offers some unique advantages over Firefox's Gecko rendering engine in certain contexts. GTK/WebKit is lightweight and less resource intensive than Gecko, which makes it a particularly good choice for mobile and embedded environments. GTK/WebKit will also eventually be a very good solution for GTK and GNOME applications that want lightweight embedded HTML rendering

And b) is possibly answered in the quote above. There are two lectures in the Gstreamer link about tree shaped kernels and only loading and dynamically linking those libraries being used to reduce footprint and decrease loading times. (the process description from memory) The reasoning stuck with me...applies to embedded systems with limited resources like the PS3 and XMB.

So possibly a GTK/Cairo/Gstreamer/Soup/Pango port. (guess) This still requires some form of windows support; so libraries for that? Big job! Back to what is the NGP using?

There is also now a GTK/Cairo/Gstreamer/Clutter webkit port by Collabora. Which might lend it'self to a game platform more than any other Webkit port. http://blog.kov.eti.br/?p=139

Clutter: http://en.wikipedia.org/wiki/Clutter_(toolkit)

Clutter allows implicit animations of every item on the canvas using special objects called behaviours: each behaviour can be applied to multiple actors, and multiple behaviours can be composed on the same actor. Behaviours handle animations implicitly: the developer specifies the initial and final states, the time (or number of frames) needed to complete the animation, the function of time to be used (linear, sine wave, exponential, etc.), and the behaviour will take care of the tweening. Clutter provides a generic base class for developers to implement custom behaviours, and various simple classes handling simple properties, like opacity, position on the Z axis (depth), position along a path, rotation, etc.

Supporting my statement that Sony is going more open source:

http://news.ycombinator.com/item?id=1938652

http://snap.sonydeveloper.com/about/

Sony’s Networked Application Platform is a project designed to leverage the open source community to build and evolve the next generation application framework for consumer electronic devices.

The developer program gives access to a developer community and resources like SDK, tools, documentation and other developers.

The foundation upon which this project is base comes from the GNUstep community, whose origin dates back to the OpenStep standard developed by NeXT Computer Inc (now Apple Computer Inc.). While Apple has continued to update their specification in the form of Cocoa and Mac OS X, the GNUstep branch of the tree has diverged considerably.

The GNUstep core libraries strictly adhere to the OpenStep standard and OPENSTEP implementation. They consider changes and additions to their API only under the following circumstances:

SNAP has a re-architected display model and backend based on Cairo evolving toward COLLADA over time. Our look and feel also includes navigation metaphors derived from standard Sony UX constructs such as our XrossMediaBar.

"SNAP has a re-architected display model and backend based on Cairo evolving toward COLLADA over time. Our look and feel also includes navigation metaphors derived from standard Sony UX constructs such as our XrossMediaBar."

So the XMB will be rewritten to use Cairo and Cairo will be supported with PSGL. (guess) It also appears that Sony is expecting Cairo to evolve to support the Collada file format and that would mean a feature set that would support game assets (Another guess). Out on a limb here but it fits.

People are already making Cairo/SVG driven desktops.

http://avisoadee.blog.com/2010/10/01/cairo-an-alternative-desktop-environment/

So Cairo will be used to support a custom PS3 desktop with windows like support. I guess this fills out the requirements for a webkit port.

How The Google Android OS does the same thing: http://www.wiredvanity.com/articles/34/on-the-x-windowing-system-and-why-google-isn-t-using-it-for-chrome-os

On the X Windowing System and why Google isn't using it for Chrome OS
14 July 2009

Now, why then does Google not use an X windowing system in Chrome OS? The technical explanation for this is “they simply don’t need it for what they’re doing.”

Another acquisition Google made in 2005 was a company called Skia Inc. that marketed a 2D graphics library also called Skia. The Skia library, written in C, does vector graphics, text, and images. It was open-sourced in 2008. Same features as Cairo

In writing Android, Google used Skia as the graphics backend for its windowing system and GUI toolkit. They also wrote a Skia backend for WebKit, which is used in the Android web browser. This same Skia Webkit backend is also used by Chrome on all of the platforms it supports. This is what I think Sony will do with Cairo

What Google has done in Android, and will do in Chrome OS, is write a simplistic windowing system using Skia, and let that output directly to the device framebuffer, while I’m no expert on Skia, I assume it also has a backend implementing its painting operations via OpenGL, which would then make it easily hardware-acceleratable on a variety of different hardware offerings.

So to sum it up:

Google already has Skia-on-kernel-without-X working in Android, and they already have Chrome/WebKit using Skia. Chrome OS is unlikely to have the sort of drag-windows-around-on-a-desktop user experience that OS X or KDE offer (I expect it will be something more iPhone/Android-like), so while they could certainly use X, the small subset of its functionality they actually require is easily written anew using Skia, and doing so means they don’t have to import a significant outside codebase and get involved with its development.

I believe the above also applies to the PS3 and NGP.


Youtube video of a Cairo Dock on a Linux desktop. The Cairo part are the animated icons at the bottom of the screen. http://www.youtube.com/watch?v=VMe3oag7Mqs

COLLADA is a COLLAborative Design Activity for establishing an interchange file format for interactive 3D applications. COLLADA is managed by the not-for-profit technology consortium, the Khronos Group.

COLLADA defines an open standard XML schema for exchanging digital assets among various graphics software applications that might otherwise store their assets in incompatible file formats. COLLADA documents that describe digital assets are XML files, usually identified with a .dae (digital asset exchange) filename extension.

Originally created at Sony Computer Entertainment
 
magicaltrevor said:
Bundled with Cross-game Voice Chat?

:(

I don't know and perhaps Sony doesn't know. Depends I guess on the final size of the XMB. IF they rewrite it to include the new features and the final size is smaller, enough to not impact current games,....possibly.

Increasing the size of the C library (GlibC) for OS calls beyond the limited support in the PS3 now is going to impact the XMB size (larger). Implementing a new dynamic kernel process might answer that issue.

Currently I think the PS3 reboots with a smaller kernel/XMB when a game is loaded and reboots and loads the XMB OS we associate with the PS3 when we exit a game. This in an effort to have a clean machine with the MAX free memory for games or applications.

Sony uses a feature they call Snapshot Kernel which is similar to resume on PCs.

Calling Webkit inside a game might have a game developer able to call various subsets of webkit. For instance there has to be support for bitmap and vector graphics fonts in webkit but only one could be specified and used inside a game. Having Webkit use Cairo with PSGL support rather than OpenGL would result in a resource savings as the game uses the same low level programmable GPU calls. Lots of possibles.
 

androvsky

Member
A minor update. I checked the page where Sony keeps their webkit source, and noticed there's an update; a new snapshot of the webkit port as of early February, uploaded March 31. Now, the nice thing about this is there's a file that shows only the changes being made to the project to make it run on the PS3. The update from the fall only showed updates to the javascript core.

http://downloads.snei-opensource.com/pub/webkit/

This update has a ton of changes, including additions to the webkit interface core, like virtual keyboard support, mouse support, various strings ("Save as...", "missing plugin", etc) to be localized, etc. We're definitely getting a webkit browser, and it's probably mostly ready.

fake edit: They added a test directory two days ago. One of the pages tests the "video" tag (even though the page is supposedly html 4 "transitional", whatever that means).
http://downloads.snei-opensource.com/pub/webkit-tests/

Note that the Silverlight test page is apparently to make sure it handles the lack of a plugin correctly. The flash page is amusing though (it embeds a 640 x 380 youtube clip). If they're testing it on sample pages, I'm going to make a wild guess that a wide release is still a couple months out. At least it's in a form suitable for testing...


And yes, there's some Cairo patches listed in the new directory, but they're just a couple of quick hacks to help make them build on the PS3. There's no source changes to Cairo. Is PS3 support in the main Cairo branch, or is there still work to be done there?



real edit: Additional information from the distant future for quick and easy referencing:
As of April 22, Sony added a job offer for a webkit engineer. The juicy bit:
Job Summary: The Sony Network Entertainment Senior Software Engineer-WebKit will join a dynamic team of experienced engineers developing WebKit and related open source software for the Playstation3, Sony Android devices, and other Sony consumer electronics products.

http://downloads.snei-opensource.com/jobs/webkit-engineer.txt
 
I'll be so happy if we get a new browser. The current one is garbage. And why only watch youtube at such low resolutions on it? Hope they get that fixed.
 

androvsky

Member
I'm going through ps3-webkit-support.diff because I'm bored. Looks like we're actually getting Chrome, since Geoff's added a couple of files for ChromeClientPOSIX.
Source/WebKit/posix/WebCoreSupport/ChromeClientPOSIX.cpp | 664 +++++
Source/WebKit/posix/WebCoreSupport/ChromeClientPOSIX.h | 177 +
Naturally the full source of his changes are in that .diff file also. From what I can tell, the posix platform only refers to the PS3 in webkit, it doesn't show up at all in the mainline source code (unix versions are typically under their widget library, like Qt or gtk). There's a lot of functions that aren't implemented, but they're mainly ones that involve windowing that you wouldn't expect to be supported on the PS3.

This might not mean much since I don't see the cool stuff from Chrome in these sources, like bookmark syncing, but it's a huge freaking project., and the diff only covers what needed to be changed so it works on the PS3. It might not even get Chrome branding visible to the user, but I think if he's calling it ChromeClient, we can expect good things for the future.

edit: okay, so Sony probably is NOT bringing a Google Chrome branded browser to the PS3. Apparently ChromeClient is a generic term in webkit. Sorry folks, but at least we're getting a new browser, right?
 

Jtrizzy

Member
What was the trick to using the current browser on gaf? I'd like some improved web browsing, mostly for watching game videos etc.
 

Lesiroth

Member
Jtrizzy said:
What was the trick to using the current browser on gaf? I'd like some improved web browsing, mostly for watching game videos etc.
I think secondapps.com/neogaf works, but I'm not sure.
 
http://www.neogaf.com/forum/showpost.php?p=27012121&postcount=260
androvsky said:
I'm going through ps3-webkit-support.diff because I'm bored. Edit: Incorrect assumption Looks like we're actually getting Chrome, since Geoff's added a couple of files for ChromeClientPOSIX. See below

Naturally the full source of his changes are in that .diff file also. From what I can tell, the posix platform only refers to the PS3 in webkit, it doesn't show up at all in the mainline source code (unix versions are typically under their widget library, like Qt or gtk). There's a lot of functions that aren't implemented, but they're mainly ones that involve windowing that you wouldn't expect to be supported on the PS3.

This might not mean much since I don't see the cool stuff from Chrome in these sources, like bookmark syncing, but it's a huge freaking project., and the diff only covers what needed to be changed so it works on the PS3. It might not even get Chrome branding visible to the user, but I think if he's calling it ChromeClient, we can expect good things for the future.

Good find and you've gone into the files.... A couple of points that only impact your guess about Chrome.

Chromeclient is a term coined because the "Chrome" in a browser is that front end interface with the user (client); thus the ChromeClient name. It does not mean Chrome is coming to the PS3.

The ChromeClientPOSIX mentioned above is the opening and drawing of a window (and more) using the Sony POSIX theme Cairo SVG script. If you look at the .diff file many of the ChromeClient routines return not enabled.

ChromeClientGTK using the GTK toolkit=ChromeClientPOSIX with Sony custom Cairo scripts


The First LGPL required webkit disclosure was for a HTML5 javascript engine (partial webkit) that was released in PS3 Firmware 3.5 (Sept 28, 2010) and had information also confirmed in the second disclosure (March 2011). http://downloads.snei-opensource.com/pub/webkit/
Subject: Add PS3 target to Webkit build configure

Signed-off-by: Geoff Levand <geoff.levand@am.sony.com>
---

configure.ac | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)

--- a/configure.ac
+++ b/configure.ac
@@ -170,13 +170,13 @@ AC_SUBST([PNG_LIBS])
# determine the GDK/GTK+ target.................................This is the key to understanding what's coming *GTK+ webkit* http://www.webkitgtk.org/
AC_MSG_CHECKING([the target windowing system])
AC_ARG_WITH(target,
- AC_HELP_STRING([--with-target=@<:mad:x11/win32/quartz/directfb@:>@],
+ AC_HELP_STRING([--with-target=@<:mad:x11/win32/quartz/directfb/ps3@:>@],
[Select webkit target [default=x11]]),
[],[with_target="x11"])

case "$with_target" in
- x11|win32|quartz|directfb) ;;
- *) AC_MSG_ERROR([Invalid target: must be x11, quartz, win32, or directfb.]) ;;
+ x11|win32|quartz|directfb|ps3) ;;
+ *) AC_MSG_ERROR([Invalid target: must be x11, quartz, win32, directfb, or ps3.]) ;;
esac

AC_MSG_RESULT([$with_target])
@@ -196,6 +196,8 @@ if test "$with_hildon" = "yes"; then
AC_SUBST([HILDON_LIBS])
fi

+if test "$with_target" != "ps3";
+
# minimum base dependencies
LIBSOUP_REQUIRED_VERSION=2.28.2
CAIRO_REQUIRED_VERSION=1.6
@@ -281,6 +283,8 @@ AC_SUBST(GTK_LIBS)
AC_SUBST(CAIRO_CFLAGS)
AC_SUBST(CAIRO_LIBS)
All the AC_SUBST in the diff file above are changing the git (Internet link-able source code libraries) to Sony internal libraries


gtk_webkit.png


The GTK+ webkit port uses Cairo. All GTK+ webkit ports use Cairo but the version of Cairo being used is different for the different backends being used (x11/win32/quartz/directfb/ps3). For the PS3 I believe OpenGL will be used which is not x11/win32/quartz/directfb. (PS3 is in the position normally occupied by OpenGL)

EDIT: GDK/GTK+ in the above disclosure took me a while to figure out. GDK is a wrapper for Xwindows in that it uses Xwindows to provide windows support for GTK+ or other drawing tools. GDK/GTK+ is a reference to newer versions of GTK+ that have the GDK Xwindows style calls as part of GTK+ without needing Xwindows or GDK libraries. SO newer versions of GTK+ using for instance OpenGL do not need Xwindows or GDK which makes for a smaller cleaner/leaner GUI toolkit.

The Javascript engine currently in the PS3 is not Chrome's V8. From everything I have been able to gather, the PS3 has the same as a Linux kernel version 2.6 same as Android (POSIX standard) Wiki is speculating FreeBSD Unix, still POSIX.

Cairo is incompatible as a backend with Chrome (uses Skia). Also the PS3 uses Gstreamer for Audio Video streams and Gstreamer with cario bindings for HTML5 <video> in the coming webkit. Gstreamer is not supported by Chrome either.
AC_SUBST([LIBSOUP_CFLAGS])
@@ -772,6 +778,8 @@ if test "$enable_video" = "yes"; then
AC_SUBST([GSTREAMER_LIBS])

Several hardware developers are porting Cairo and Gstreamer to Android platforms because the Open Core AV library and Skia the SVG library don't properly support newer more powerful platforms. Google has even gone Cairo-Clutter for the Chrome OS.

Testing for Flash? Disappointing as I thought that Google might release an open source DRM for HTML5 video and Flash would not be needed. Wrong again Jeff. Edit: It's just a Flash video codec test (Gstreamer plugin) without DRM, Ultraviolet DRM still coming from Google?

Perhaps your latest find added to mine will finally have the press take notice that a webkit port to the PS3 is coming and is in the testing phase. Really really good find.

Thanks.

Edit: to correct errors
PS target did not mean OpenGL
GTK APIs are used, GTK tookit chrome changed to SONY custom called POSIX and it only changes the look and feel.
Gstreamer APIs are used not gstreamer for disclosure reasons related to the Gstreamer license.
 
androvsky said:
A minor update. I checked the page where Sony keeps their webkit source, and noticed there's an update; a new snapshot of the webkit port as of early February, uploaded March 31. Now, the nice thing about this is there's a file that shows only the changes being made to the project to make it run on the PS3. The update from the fall only showed updates to the javascript core.

http://downloads.snei-opensource.com/pub/webkit/

This update has a ton of changes, including additions to the webkit interface core, like virtual keyboard support, mouse support, various strings ("Save as...", "missing plugin", etc) to be localized, etc. We're definitely getting a webkit browser, and it's probably mostly ready.

fake edit: They added a test directory two days ago. One of the pages tests the "video" tag (even though the page is supposedly html 4 "transitional", whatever that means).
http://downloads.snei-opensource.com/pub/webkit-tests/

Note that the Silverlight test page is apparently to make sure it handles the lack of a plugin correctly. The flash page is amusing though (it embeds a 640 x 380 youtube clip). If they're testing it on sample pages, I'm going to make a wild guess that a wide release is still a couple months out. At least it's in a form suitable for testing...

And yes, there's some Cairo patches listed in the new directory, but they're just a couple of quick hacks to help make them build on the PS3. There's no source changes to Cairo. Is PS3 support in the main Cairo branch, or is there still work to be done there?

A minor update? Too modest, it's a major update to the information we have. It's now, with three links, a chain of logic pointing to "We're definitely getting a webkit browser".

Everything above good logic. I'd only add the following.

What about the NGP, is the test page for that and to be used for the PS3 also? IF this is true it might impact your statement that the PS3 is at a stage to be tested. I think you are correct but I had to mention another use.

Re: Cairo, if you have been following my posts, is supposed to be the back end for ALL new networked Sony platforms (2011 TV, Blu-ray confirmed) and PS3 (guess) based on the webkit port using cairo, Snap developer language RE: Cairo and both Android and iOS using SVG for their display. Add to that the Sony infinite Zoom picture database which uses Cairo SVG and more examples I give in the link below and I think Sony is moving to SVG on all their platforms that can support Cairo SVG; the PS3 Cell SPU is superb at vector calculations. Edit: Cairo SVG includes Pixman code which is picture/pixel manipulation also using OpenGL, most of the Playview/infinite picture database viewing is using Pixman routines for zooming.

"There's no source changes to Cairo. Is PS3 support in the main Cairo branch, or is there still work to be done there?" Cairo is written to be cross platform so few if any changes (read as NO changes) would have to be made to a PS3 Cairo port. The openGL support needed for cairo would be custom provided by Sony using their PSGL. I would guess that would not need to be disclosed.

There appears to be an issue with Sony not disclosing Cairo-Pango use in Sony TVs and blu-ray players

http://old.nabble.com/Pango-License-td30466635.html

That said, doesn't mean that all companies follow that section correctly. For example, the latest versions of Sony TVs and some other consumer devices (Bluray players, etc) use Pango and cairo to provide i18n text. They make the sources to pango and cairo available on their website, but no way for me to build a new firmware using an updated pango. Reminds me that I wanted to contact the SFLC about this.

So either my guess is correct in some way (no source changes needed), Sony has found a loophole or they are in violation.

I believe the entire PS3 XMB and XMB applications will be rewritten to use Cairo. That is still a guess and needs to be confirmed.

http://www.neogaf.com/forum/showpost.php?p=26945282&postcount=436

"I'm going to make a wild guess that a wide release is still a couple months out." I'd agree and add that my guess as to the XMB rewrite to use Cairo making the XMB a cairo surface that webkit code can use to open windows, display widgets and animation would put off a full release even longer. That's a guess based on a guess.

I would guess an announcement soon, E3 or sooner with a major update (4.0) following sometime later but certainly before the S1 Sony Android or NGP is released. Perhaps some announcement at E3 about both webkit and PS3 changes to allow NGP-PS3 ecosystem features.

If this was done using Gstreamer with cairo bindings (last E3):

2nvezuv.jpg


Then development PS3 XMBs might be further along and a timetable would be E3. That's a supportable argument as the SNAP program has had the Cairo notice since April 2010 and 2011 Sony TVs and BLu-ray players are using Cairo-pango. Webkit just got the announcement of a Stable release Feb 2011. Cairo except for some optimizations (AA and Shader) has been done for months, enough that Sony used Cairo in their TV's released months ago.
 
androvsky said:
I'm going through ps3-webkit-support.diff because I'm bored.
Naturally the full source of his changes are in that .diff file also. From what I can tell, the posix platform only refers to the PS3 in webkit, it doesn't show up at all in the mainline source code (unix versions are typically under their widget library, like Qt or gtk). There's a lot of functions that aren't implemented, but they're mainly ones that involve windowing that you wouldn't expect to be supported on the PS3.

This might not mean much since I don't see the cool stuff from Chrome in these sources, like bookmark syncing, but it's a huge freaking project., and the diff only covers what needed to be changed so it works on the PS3. It might not even get Chrome branding visible to the user, but I think if he's calling it ChromeClient, we can expect good things for the future.
I've been going through the Diff file too. Naming conventions for the different styles of webkit branches like QT or GTK look like it's based on the GUI toolkit used; Androvsky mentions that above. In the PS3 webkit port they apparently take a GTK port and convert it to what they call a POSIX port similar to what Google does with Android. This would be confusing/will be confusing to just about everyone because POSIX is normally thought of as a extended set of OS standards for a bunch of things starting with File I/O and ending up with mouse events. We don't normally associate it with a webkit port.

Posted below is a simple explanation of what Google did with Android but the .DIFF file is actually much clearer. It's not just the GUI that has to be handled, POSIX IO too.

On the X Windowing System and why Google isn't using it for Chrome OS
14 July 2009

Now, why then does Google not use an X windowing system in Chrome OS? The technical explanation for this is “they simply don’t need it for what they’re doing.”

Another acquisition Google made in 2005 was a company called Skia Inc. that marketed a 2D graphics library also called Skia. The Skia library, written in C, does vector graphics, text, and images. It was open-sourced in 2008. Same features as Cairo

In writing Android, Google used Skia as the graphics backend for its windowing system and GUI toolkit. They also wrote a Skia backend for WebKit, which is used in the Android web browser. This same Skia Webkit backend is also used by Chrome on all of the platforms it supports. This is what I think Sony will do with Cairo

What Google has done in Android, and will do in Chrome OS, is write a simplistic windowing system using Skia, and let that output directly to the device framebuffer, while I’m no expert on Skia, I assume it also has a backend implementing its painting operations via OpenGL, which would then make it easily hardware-acceleratable on a variety of different hardware offerings.

So to sum it up:

Google already has Skia-on-kernel-without-X working in Android, and they already have Chrome/WebKit using Skia. Chrome OS is unlikely to have the sort of drag-windows-around-on-a-desktop user experience that OS X or KDE offer (I expect it will be something more iPhone/Android-like), so while they could certainly use X, the small subset of its functionality they actually require is easily written anew using Skia, and doing so means they don’t have to import a significant outside codebase and get involved with its development.

So for every GUI toolkit call based on GTK, Sony provides a custom vector drawing based on Cairo, they replace the GTK GUI with one of their own. There are other differences and PS3-webkit-support.diff file will probably give us enough information to tell what will be coming. It's 425K of text so as androvsky said alot of work by Sony.

So Sony is taking a GTK/cairo/gstreamer/pango webkit port and making it a GTK-POSIX/cairo/gstreamer/pango port.

It's premature for any of this to be posted until several people have gone through the .DIFF file. Already we have a serious difference in our takes on the DIFF file. 425K of language that is not native to my thinking processes can have me make serious errors in my conclusions as to what it's saying.

As you can see below, many of the POSIX labeled files have descriptors for a toolkit.

Source/WebCore/platform/posix/SearchPopupMenuPOSIX.cpp | 61
Source/WebCore/platform/posix/SearchPopupMenuPOSIX.h | 50

The following is a Cairo SVG draw routine for a Popup menu. Notice the RendererThemePOSIX POSIX is the Sony name for the toolkit theme

+ // draw popup menu bg
+ cairo_set_source_rgba(cr, RGBA_BG);
+ cairo_rectangle(cr,
+ m_x,
+ m_y + m_h,
+ m_w,
+ m_itemH * m_visibleCount + m_outlineW * 2);
+ cairo_fill(cr);
+
+ // draw popup menu outline
+ cairo_set_source_rgba(cr, RGBA_OUTLINE);
+ cairo_rectangle(cr,
+ m_x,
+ m_y + m_h,
+ m_w,
+ m_itemH * m_visibleCount + m_outlineW * 2);
+ cairo_set_line_width(cr, m_outlineW);
+ cairo_set_line_join(cr, CAIRO_LINE_JOIN_MITER);
+ cairo_stroke(cr);
+#endif
+
+ // set clip area
+ cairo_rectangle(cr, m_x, m_y + m_h + m_outlineW, m_w, m_itemH * m_visibleCount);
+ cairo_clip(cr);
+ cairo_new_path(cr);
+
+ // draw text
+ cairo_select_font_face(cr,
+ "",
+ m_fontIsItalic ? CAIRO_FONT_SLANT_ITALIC : CAIRO_FONT_SLANT_NORMAL,
+ m_fontIsBold ? CAIRO_FONT_WEIGHT_BOLD : CAIRO_FONT_WEIGHT_NORMAL);
+ cairo_set_font_size(cr, m_fontSize);
+
+ for (int i = 0; i < m_itemCount; i++) {
+ if (i == m_selected) {
+#if USE_THEME_IMAGES
+ WebCore::RenderThemePOSIX::paintThemeImage3x3(&ctx, highlight,
+ WebCore::IntRect((int)(m_x + m_outlineW),
+ (int)(m_visibleY + m_itemH * i),
+ (int)(m_w - m_outlineW * 2),
+ (int)(m_itemH)),
+ 1, 1, 1, 1);
+#else
+ // draw hovered highlight
+ cairo_set_source_rgba(cr, RGBA_HIGHLIGHT);
+ cairo_rectangle(cr,
+ m_x + m_outlineW,
+ m_visibleY + m_itemH * i,
+ m_w - m_outlineW * 2,
+ m_itemH);
+ cairo_fill(cr);
+#endif
+
+ // draw hovered text
+ cairo_set_source_rgba(cr,RGBA_TEXT_HOVERED);
+ cairo_move_to(cr,
+ m_x + m_outlineW + m_paddingW,
+ m_visibleY + m_itemH * (i + 1) - m_paddingH - m_fontDescent);
+ cairo_show_text(cr, m_items);
+ } else {
+ // draw non-selected text
+ cairo_set_source_rgba(cr, RGBA_TEXT_NORMAL);
+ cairo_move_to(cr,
+ m_x + m_outlineW + m_paddingW,
+ m_visibleY + m_itemH * (i + 1) - m_paddingH - m_fontDescent);
+ cairo_show_text(cr, m_items);
+ }
+ }
+
+ cairo_restore(cr);
+

GTK and Cairo mentioned below

+if TARGET_POSIX
+WebKit := $(srcdir)/Source/WebKit/posix
+else
WebKit := $(srcdir)/Source/WebKit/gtk
+endif
WebKit2 := $(srcdir)/Source/WebKit2

# Libraries and support components
@@ -106,8 +110,11 @@ global_cppflags += \
-Wformat -Wformat-security -Wno-format-y2k -Wundef \
-Wmissing-format-attribute -Wpointer-arith -Wwrite-strings \
-Wno-unused-parameter -Wno-parentheses \
- -fno-exceptions -DENABLE_GLIB_SUPPORT=1
+ -fno-exceptions

+if !TARGET_POSIX
+global_cppflags += -DENABLE_GLIB_SUPPORT=1 This is interesting.
+endif

global_cxxflags += \
-fno-rtti
@@ -163,6 +170,7 @@ libJavaScriptCore_la_CPPFLAGS = \
$(javascriptcore_cppflags)

# Extra checks and flags
+if !TARGET_POSIX
global_cppflags += \
-DBUILDING_CAIRO__=1 \
-DBUILDING_GTK__=1 \
 

Elios83

Member
Considering that PS3 firmware updates have been minor for a looooong time, it would be nice if they had in store a big UI upgrade, Chrome as the new web browser and finally cross game chat implemented for 4.0 to be released around E3.
 

Truespeed

Member
H_Prestige said:
How would Chrome even run on just 256mb RAM? That browser is a notorious RAM hog.

Wait, what? Is there a reason why it can't use the available video ram? As for Chrome being a memory hog, well that's debatable as it opens a new process for each tab - which is good. There's nothing preventing them from not doing this in the PS3 implementation or just using virtual RAM to handle the extra tabs.
 
Truespeed said:
Wait, what? Is there a reason why it can't use the available video ram? As for Chrome being a memory hog, well that's debatable as it opens a new process for each tab - which is good. There's nothing preventing them from not doing this in the PS3 implementation or just using virtual RAM to handle the extra tabs.

One last time, it's not Chrome it's Android Chrome-like. Webkit isn't even the big news, it's Cairo SVG.

And for those of you who question how much memory will be needed, just like with Chrome on Android platforms, it's all Scalable Vector Graphics which requires less ram.
 

Truespeed

Member
jeff_rigby said:
One last time, it's not Chrome it's Android Chrome-like. Webkit isn't even the big news, it's Cairo SVG.

And for those of you who question how much memory will be needed, just like with Chrome on Android platforms, it's all Scalable Vector Graphics which requires less ram.

Of course it's not going to be Chrome. Google wouldn't waste their time with a PS3 Chrome build.
 

androvsky

Member
Oh boy, there's a lot here... I thought jeff_rigby would enjoy this. :)

Truespeed said:
Wait, what? Is there a reason why it can't use the available video ram? As for Chrome being a memory hog, well that's debatable as it opens a new process for each tab - which is good. There's nothing preventing them from not doing this in the PS3 implementation or just using virtual RAM to handle the extra tabs.
At least in Linux, the PS3 can use the video ram as a swap file. It's slow, but still better than going to disk. The problem is that the Cell can't access video memory directly at anything close to a reasonable speed, so you have to tell the RSX to move data in and out of system ram. The PS3 has much leaner (I hope) GUI libraries, which will help, in addition to the OS not taking a lot of ram. Might not be able to have as many tabs open as you can with the current browser, but I'll take two good tabs over eight poorly-rendered ones.

Considering that PS3 firmware updates have been minor for a looooong time, it would be nice if they had in store a big UI upgrade, Chrome as the new web browser and finally cross game chat implemented for 4.0 to be released around E3.
Well, Steam did just get a new codec for cross-game chat, and we are supposed to have cross-platform play... heh, would be nice. I'm still half-expecting a firmware update before Portal 2 just due to how cagey Sony and Valve are about the details even now.

Lol, wow. Someone even asked Google. Reminds me of a person I know who works at Google who was surprisingly upset at the rumor that Firefox was coming to the PS3 that was going around a couple years ago. :)


cairo: If the PS3 supports normal opengl, that's great news. A real webkit browser is probably going to run slow enough on the PS3, last thing it needs is unaccelerated drawing routines.

Edit: Chromeclient is a term coined because Google Chrome developed the library and pushed it back into the main webkit branch. It's related to the opening of windows in webkit. It does not mean Chrome is coming to the PS3.

The ChromeClientPOSIX mentioned above is the opening and drawing of a window using the Sony POSIX theme Cairo SVG script. If you look at the .diff file many of the ChromeClient routines return not enabled.
Ooops, good catch. Oh well, at least we're getting something close enough to Chrome (although I'd love bookmark syncing on the PS3). I did notice all the not enabled ChromeClient functions, but I assume they had to do with how the PS3 handles windows.
 

androvsky

Member
Kingsora said:
Link doesn't work..

this one does: http://translate.google.com/transla...jgt-de-ps3-google-chrome-als-internetbrowser/

Btw when is the soonest you guys think this could come out? Months, a year, years?
wild optimism:
If they tested it in private and only just now put the testing pages out for people with debug units to check out, it could happen in a few weeks. Even though the code was recently posted, it was dated from early February. Since they're making their GPL-compliance post now, one could assume that a debug firmware is going out soon for wide testing.

realism:
Summer is much more likely though. It looks like Geoff Levand is the only person working on it, so if any issues crop up, it could get delayed a lot. And those testing pages are pretty basic... I have no clue how far along the port actually is. I'd like to say it's done and ready to be distributed, otherwise they wouldn't have posted the snapshot, but I have no way of knowing. They still haven't uploaded the video clip for the video tag test page. :/

I can't imagine it not being done this year, the fall update at the latest I'd say.
 

sharath_kumar

Neo Member
since the browsers are susceptible to buffer overflow attacks, wont this be a problem for ps3 security? the browser will be running in gameos. won't it? unless they can make it to run in some non-priviliged mode, i wonder how sony will handle this problem.
 

theBishop

Banned
Just friendly head's up. "POSIX" is just a generic term for unix-like platforms. Maybe somebody has super insider info that I missed and a Sony coder is being cute, but it's highly unlikely a Playstation Theme would be named that.

http://en.wikipedia.org/wiki/Posix

I haven't read this thread very carefully, but jeff_rigby seems to be stabbing wildly in the dark.
 

Raistlin

Post Count: 9999
jeff_rigby said:
One last time, it's not Chrome it's Android Chrome-like. Webkit isn't even the big news, it's Cairo SVG.

And for those of you who question how much memory will be needed, just like with Chrome on Android platforms, it's all Scalable Vector Graphics which requires less ram.
Not to mention you won't have extensions running, and they could easily put a tab limit like most portable browsers.





Truespeed said:
Of course it's not going to be Chrome. Google wouldn't waste their time with a PS3 Chrome build.
No, but given their relationship ... it'll be curious to see how things pan out for the future. Could PS4 see an actual Chrome build? Or how about GoogleTV, or an Android-like browser? For an Android-like browser, I wouldn't be shocked if NGP sees it. GoogleTV? Not be surprising if it hits their TV's, BD-players, receivers, and PS4.
 

Raistlin

Post Count: 9999
sharath_kumar said:
since the browsers are susceptible to buffer overflow attacks, wont this be a problem for ps3 security? the browser will be running in gameos. won't it? unless they can make it to run in some non-priviliged mode, i wonder how sony will handle this problem.
Admittedly I'm a newb for haxx0ring ... but is there any reason to assume the current PS3 browser is any less susceptible to such an attack?
 

Truespeed

Member
Raistlin said:
No, but given their relationship ... it'll be curious to see how things pan out for the future. Could PS4 see an actual Chrome build? Or how about GoogleTV, or an Android-like browser? For an Android-like browser, I wouldn't be shocked if NGP sees it. GoogleTV? Not be surprising if it hits their TV's, BD-players, receivers, and PS4.

That's a good point. There's no question Google wants Chrome everywhere, but I think it's important for Google to control the Chrome experience if it were to ever appear on the PS4. Sony can always take the Chromium source and run with it, but then it wouldn't be Google Chrome. Providing Google is willing to do this, it would be silly not to hand over the browser duties to Google.

Android on the PS4 is interesting because it's platform agnostic so we could theoretically see an Android layer / UI for the PS4 and be able to run a majority of apps that don't make use of the native client SDK - which is part of the brilliance of Android. Since Android is already part of the NGP I don't see why they wouldn't unify that experience with the PS4.
 

MThanded

I Was There! Official L Receiver 2/12/2016
sharath_kumar said:
since the browsers are susceptible to buffer overflow attacks, wont this be a problem for ps3 security? the browser will be running in gameos. won't it? unless they can make it to run in some non-priviliged mode, i wonder how sony will handle this problem.
assuming the browser runs inside the hypervisor.
 

Truespeed

Member
sharath_kumar said:
since the browsers are susceptible to buffer overflow attacks, wont this be a problem for ps3 security? the browser will be running in gameos. won't it? unless they can make it to run in some non-priviliged mode, i wonder how sony will handle this problem.

The PS3 has a few things going for it and one of them is security through obscurity and perhaps some type of sandboxing. When you attack a traditional OS, like Windows, the exploit usually gains control of the OS providing it get's past any OS safe guards like ASLR, DEP, etc. The PS3 doesn't exactly work that way. If you were to exploit the browser you would then need a way to inject your PowerPC based payload into it, bypass any OS safe guards and then try to run unsigned code on a CPU that only will execute signed code.
 
theBishop said:
Just friendly head's up. "POSIX" is just a generic term for unix-like platforms. Maybe somebody has super insider info that I missed and a Sony coder is being cute, but it's highly unlikely a Playstation Theme would be named that.

http://en.wikipedia.org/wiki/Posix

I haven't read this thread very carefully, but jeff_rigby seems to be stabbing wildly in the dark.

Message 271 above from jeff_rigby
I've been going through the Diff file too. Naming conventions for the different styles of webkit branches like QT or GTK look like it's based on the GUI toolkit used; Androvsky mentions that above. In the PS3 webkit port they apparently take a GTK port and convert it to what they call a POSIX port similar to what Google does with Android.

This would be confusing/will be confusing to just about everyone because POSIX is normally thought of as a extended set of OS standards for a bunch of things starting with File I/O and ending up with mouse events. We don't normally associate it with a webkit port.

Got you too <grin>. I spent about an hour confirming.

"I haven't read this thread very carefully," Please do so, from previous posts I feel that you could confirm what I've posted. So far it looks accurate.

I'm 100% sure that Sony is calling it a POSIX port. Edit: And you have a valid point, outside of Sony programmer circles it will be called something else.

Chrome like, well it's SVG based like ANDROID Chrome but so was the latest GTK port. It doesn't use Xwindows and manages windows similar to the way Andorid does.

androvsky; There are at least 3 other programmer I saw in the posts that are involved in the webkit port. They all have a North America email address.

And guys, the big news is Cairo SVG & Gstreamer!

Second, Pango for international Fonts. IF you have frequented the Share Blog, Arabic support has been widely requested and Sir Howard just met with the king of Saudi-Arabia. Cairo-pango international fonts can be used for the PS3 OS features/XMB and I believe this is likely, a XMB rewrite will happen if only to internationalize the PS3.

Libraries included: EDIT: On closer examination many of the libraries below can't be for the PS3, Android or NGP; there are Linux desktop (Gnome) and Windows support libraries. The library names don't match repository names so they are Sony internal libraries. I'm at a loss for them to be included in the PS3 diff file. Extreme oversight or are they going to be used for something? (in browser remote desktop???)

$(CAIRO_LIBS) \ Cairo SVG library
+ $(COVERAGE_LDFLAGS) \
+ $(ENCHANT_LIBS) \...............Front end API for a spell checker and more
+ $(FREETYPE_LIBS) \...............Font Library
+ $(GAIL_LIBS) \......................GNOME Accessibility Implementation Library
+ $(GEOCLUE_LIBS) \...............No need for this on the PS3, NGP will use!
+ $(GLIB_LIBS) \......................low level C cross platform lib (needed for just about everything webkit & cairo & Gstreamer)
+ $(GSTREAMER_LIBS) \...........Gstreamer AV library *
+ $(GTK_LIBS) \.......................GTK toolkit library
+ $(HILDON_LIBS) \..................See below*
+ $(JPEG_LIBS) \......................Jpeg compression picture library (Video too?)
+ $(LIBSOUP_LIBS) \................HTTP library
+ $(LIBXML_LIBS) \
+ $(LIBXSLT_LIBS) \
+ $(PANGO_LIBS) \...................International Fonts Cairo-pango = SVG international fonts
+ $(PNG_LIBS) \.......................PNG picture library
+ $(SQLITE3_LIBS) \.................Data Base Library
+ $(UNICODE_LIBS) \

+ $(XT_LIBS) \.........................
+ $(WINMM_LIBS) \..................
+ $(SHLWAPI_LIBS) \ wrapper functions convert the Unicode input string parameters to ANSI and call ANSI versions of functions
+ $(OLE32_LIBS)

Firefox is supporting multi-touch in their browser, this could be included in the PS3 for use by accessories or ecosystem platforms or for the control panel and home applets. Or it is included in error and was meant for the NGP like geoclue (GPS).
Hildon is an application framework originally developed for mobile devices (PDAs, mobile phones, etc.) running the Linux operating system. It was developed by Nokia for the Maemo operating system and is now a part of GNOME. It focuses on providing a finger-friendly interface.[1] It is primarily a set of GTK+ extensions that provide mobile-device–oriented functionality, but also provides a desktop environment that includes a task navigator for opening and switching between programs, a control panel for user settings, and status bar, task bar and home applets.
 

patsu

Member
You should check if Cairo SVG and POSIX are part of PS Suite specs. If so, it may mean that every PS Suite device has a WebKit browser. And perhaps the POSIX theme is just a temp name for PS Suite theme.
 
patsu said:
You should check if Cairo SVG and POSIX are part of PS Suite specs. If so, it may mean that every PS Suite device has a WebKit browser. And perhaps the POSIX theme is just a temp name for PS Suite theme.
Edit: You could be on to something. It looks like Sony is planning to port webkit to multiple platforms. This to probably support webkit applications. Chrome for instance is included in Sony Vaio laptops sold by Sony for I would guess coming webGL games.

POSIX "C" libraries like GlibC and others provide for common POSIX functions allowing cross compatibility but WEBKIT features in newer handhelds are becoming STANDARD tools used in applications and GAMES. So I expect besides LibC allowing for POSIX cross compatibility, Sony might recognize the need for common webkit tools as part of "cross platform" compatibility.

Your seeing POSIX as in a cross platform standard for "C" compatibility and Sony using the POSIX theme for webkit meaning cross platform webkit for their platforms...thus the name.

Sony programmers are again being cute <grin>. Kinda reminds me of the "3.5" Flash stream and HTML"5" references in number for Firmware 3.5 in the PS3 and 6.35 for the PSP<grin>.

Most I have been able to find is NGP and PS3 are going to share libraries. I'm confident that the NGP is going to use Cairo. Sony is apparently going SVG on ALL networked platforms that can support SVG.

I know you already know this; NGP is not based on Android, it's a CE/Embedded Linux (2.6) kernel just like the PS3.

Cairo & Gstreamer are now in the PS3 (I believe Gstreamer since 2008) and from the NGP demos are already in the NGP.
 

androvsky

Member
jeff_rigby said:
Kinda reminds me of the "3.5" Flash stream and HTML"5" references in number for Firmware 3.5 in the PS3 and 6.35 for the PSP<grin>.
It's best if we don't go there...

Most I have been able to find is NGP and PS3 are going to share libraries. I'm confident that the NGP is going to use Cairo. Sony is apparently going SVG on ALL networked platforms that can support SVG.

I know you already know this; NGP is not based on Android, it's a CE/Embedded Linux (2.6) kernel just like the PS3.
Isn't Android based on a 2.6 linux kernel? And how do you know NGP and the PS3 run on linux? I remember seeing some evidence that the PS3 OS uses a commerical embedded unix variant.
Cairo & Gstreamer are now in the PS3 (Gstreamer since 2005) and from the NGP demos are already in the NGP.
Okay, why do you think Gstreamer is in the PS3? Video codecs on the PS3 are mostly useless without SPU support, and I think it's safe to say that there wasn't any SPU support in Gstreamer until after linux was removed. The only way it could've been is as a wrapper for Sony's codecs, at which point I have to wonder why bother with Gstreamer?
 
androvsky said:
It's best if we don't go there...


Isn't Android based on a 2.6 linux kernel? And how do you know NGP and the PS3 run on linux Linux 2.6 kernel? I remember seeing some evidence that the PS3 OS uses a commerical embedded unix LINUX variant.

Okay, why do you think Gstreamer is in the PS3? Video codecs on the PS3 are mostly useless without SPU support, and I think it's safe to say that there wasn't any SPU support in Gstreamer until after linux was removed. The only way it could've been is as a wrapper for Sony's codecs, at which point I have to wonder why bother with Gstreamer?

EDIT: It appears from one post that Collabora was given a PS3 DEV kit in late 2007 to "see what they could do." This was after the GTK+webkit with Gstreamer and Cairo lecture two months earlier at a Linux conference. This might mean Gstreamer was in the PS3 after this date.

1) Gstreamer, Cairo and Linux 2.6 are mentioned in the Sony Snap developer site
2) The libraries I listed pulled from the .Diff file are Linux (glib for instance) and the list includes Gstreamer and Cairo.
3) Wiki states that Power PC support for Gstreamer was available in 2005 with it's primary use for game consoles. Too soon for the Wii and MS wouldn't use it in the Xbox360 as Gstreamer is a AV standard for Linux platforms.

Sometime about 2008 Sony updated the individual codec plugins for Gstreamer to use SPUs. Gstreamer is a process for linking individual modules from source to sync with standards that allow passing of parameters as well as the data stream. There are 200 plugins that allow an amazing range of features, some of which we are now seeing in demos and games like Start the party and Eyepet. Gstreamer plugins

Look at the Gstreamer picture and check out the plugins. Notice how modular. IF you want to introduce a filter into the video stream to make it look like old film, there is a module (plugin) to do that. Want to automatically switch from WebM to H.264 when using HTML5 <video>, Gstreamer will do that. Gstreamer and Cairo are used by Firefox for just such a purpose and Gstreamer with cairo bindings for WebGL.

I've gone into this; for instance, Gstreamer with Cairo bindings is mentioned with demos several times in this thread. No one is reading the posts because they don't understand the relevance but no one can understand the relevance without reading.

Look at the application list in the Gstreamer picture and plug that into the PS3 XMB features. Chat, media player, serves media to the PSP, video editor .... coming see Collabora projects page.
 
androvsky said:
wild optimism:
If they tested it in private and only just now put the testing pages out for people with debug units to check out, it could happen in a few weeks. Even though the code was recently posted, it was dated from early February. Since they're making their GPL-compliance post now, one could assume that a debug firmware is going out soon for wide testing.

realism:
Summer is much more likely though. It looks like Geoff Levand is the only person working on it, so if any issues crop up, it could get delayed a lot. And those testing pages are pretty basic... I have no clue how far along the port actually is. I'd like to say it's done and ready to be distributed, otherwise they wouldn't have posted the snapshot, but I have no way of knowing. They still haven't uploaded the video clip for the video tag test page. :/

I can't imagine it not being done this year, the fall update at the latest I'd say.

Cairo was included in this disclosure post and has been in use since October 2010. Your right though, Sony has been posting after the release of a software product. This time webkit LGPL was posted before the release of the browser?

There are at least 3 working on webkit:
Geoff Levand <geoff.levand@am.sony.com>
Yuji Mano <yuji.mano@am.sony.com>
Kevin Campbell <Kevin.Campbell@am.sony.com>

Probably more. The names listed are the responsible people in charge of projects not the only people working on them.
 

Massa

Member
jeff_rigby said:
1) Gstreamer, Cairo and Linux 2.6 are mentioned in the Sony Snap developer site
2) The libraries I listed pulled from the .Diff file are Linux (glib for instance) and the list includes Gstreamer and Cairo.
3) Wiki states that Power PC support for Gstreamer was available in 2005 with it's primary use for game consoles. Too soon for the Wii and MS wouldn't use it in the Xbox360 as Gstreamer is a AV standard for Linux platforms.

The PS3 doesn't use either Gstreamer or Linux.
 
Top Bottom