HP Releases More WebOS Components for the TouchPad

About two months ago, HP made the first source releases of webOS components. Conspicuously absent, however, were the sources to the Luna system manager, and it was not possible to build an image for the TouchPad with what was available. On Tuesday, the webOS team released the Luna sources and build tools as "webOS Community Edition." This is a continuation of their previous source releases, and is intended only for the TouchPad; Open webOS is still slated for release in September and will be designed for porting to new hardware platforms. Quoting the developers: "With the release of the webOS Community Edition you can now learn how the TouchPad works, modify your TouchPad experience and then apply that learning to Open webOS 1.0 in the future. We are excited to empower the community to create custom user experiences on the TouchPad. For example, developers can now modify the card view, launcher, notifications, Just Type and more." You can grab the latest over at Github. The developers claim you can build and install it onto actual hardware: anyone want to give it a shot?

Comments

too little too late

[okooolo]

now they release it??

Re:too little too late

[formfeed]

They had to have a meeting about it and that nice conference room with the coffee bar wasn't available till the second half of 2012.

Camera

[headhot]

Will this help the Android guys get the camera to work?

Wonderful mariage (but horrible amount of work) !!

[DrYak]

A fusion between Open WebOS and CyanogenMod would truly *wonderful*, but would require a fuck hell of work (and could get some help from Ubuntu...)

On one hand, this would help bring together the big popular eco-system of Android and the really marvelous "stack-of-cards" based interface of Web OS (i've been using a Pre phone almost since it release in europe and i really love it). Also under the hood, WebOS is full fledged Linux system.
With that Cyanogen would will have a truly wonderful UI, with which it could compete with other Android implementation using custom UI. (like HTC's device using HTC sense instead of the default Android UI). And WebOS could get access to the much successful world of android applications.

On the other hand these OSes are quite different beasts. Android is basically Linux kernel with a custom userland using a few standard embed libraries/utilites, but mostly a huge Android-specific runtime written in Java and running on their own Dalvik (java-like [or java-infringing if you listen to Oracle's BS]) virtual machine. webOS is mostly GNU/Linux : linux kernel, on pretty much standard user land (including DEB-like package management. Really similar to lots of other embed Linux distributions), with the exception of Luna, the graphic UI which was released today and which instead of the usually Xserver + GTK and/or QT, is DirectFB + a special custom Webkit powered "card" interface.
Now that would be in a way interesting for the Android community: propelling Android away from its former "feature-phone" Linux+Java roots into a realm of state-of-the art GNU/Linux OSes (like Maemo, like also small projects like OpenMoko). (I mean nowaday Android *IS* a real smartphone OS. *BUT* it is dissimilar from regular GNU/Linux OSes found elsewhere in the Linux ecosystem).
But that require quite some adaptation and connection between the functionnality and facilities of Android and Luna.

Note that such effort are already underway.
Long time ago, canonical has already demonstrated an Android 1.x running above stock Ubuntu and thus enabling regular Android APP to run on Ubuntu machines.
Recently Canonical has also toyed with the idea of dual-function android smartphone: it's a regular android phone when in your pocket. But there's a full fledged Ubuntu running once the smartphone is connect over HDMI to a real monitor (or docked in brainless-laptop à la AlwaysInnovating's TouchBook).
What helps is that the android-specific modifications (already all the special IPC stuff, but not the yet all the power-saving stuff) are getting incorporated into the 3.x series of linux kernels, so running stock ubuntu on a opensource android is easily possible (albeit with a shorter battery life).
And for the record: WebOS use a rather stock kernel (except for power saving features) so the Open WebOS is supposed to target modern linux 3.x kernels (sans the power saving features). So at least kernel wise open webos + android are as possible as the current android + ubuntu mixes experiments.
What really needs to be done is "only" to get the android parts to "talk" with the open web os parts. But as said, work from canonical could be leveraged here.

(For the record as WebOS is pretty much standard GNU/Linux without much customisation beside power-usage reduction, there are experiments to get Debian and Ubuntu running on webOS devices *along* webos, or at least on devices with enough flash to hold the desktop OS partition)

So an "open webos + android" combo done by collaborating CyanogenMod and WebOS internals teams (with help from canonical) would be really awesome.
But I'm not holding my breath.

(an even more crazy idea would be to get a few of the surviving openmoko guys involved: I mean they managed to do some rather crazy optimisation to get the FreeRunner to boot from several minutes down to mere seconds, and they try to create a standard framework for telephony applications over FreeSmartphone.org but that is probably just useless dreams. webos+android is cool enough).

No

[DrYak]

Saddly, no.

Luna is only the GUI (DirectFB + OpenGL ES + a WebKit based "card" interface).
The webcam lies into the custom kernel used by web OS devices (I don't how much of the camera is an opensource dirvers visible in the kernel patches and how much is a binary-only module), and perhaps a few specific libraries (probably not opensourced yet).

Re:No

[rahvin112]

Large Binary blob, developed for IIRC Kernel 2.6.5. The kernel is so old it predates even Android 2.2 and with WebOS using a more standard kernel and Android having it's modified kernel interfaces its darn near impossible for them to make the port. They only got the mic working because of the one guy that was shipped a touchpad with android 2.2 loaded. It is my understanding from reading the threads that the camera blob will not be easy if it's possible at all. The talk was that it would potentially need a full reverse engineering of the binary blob to get driver written with Android 4 (which is highly unlikely to happen given the amount of effort involved would likely exceed the lifetime of the device).

Like OpenMoko

[DrYak]

It matters because :

- that gives an upgrade path for current webos device owner to see updates to their device even if only as a small comunity of enthousiats (kind of how OpenMoko is still seeing some development around the FreeRunner and its FreeSmartphone.org-based SHR OS. Its small, slow paced, community driven. But it's *there* !)

- webOS GUI is really nice and really does serve well portable devices like tablets and smartphone. much better multitasking than anything else. also nice concepts of apps build around HTML5+Javascript (even Microsoft is toying with the idea in Metro). Opensourcing Luna (i.e.: the GUI itself) is interesting because that means that the GUI could be incorporated back into Android, although that would require a lot of work. But would give to Cyanogen and the Like a really nice and interesting interface which could compete against other such custom interfaces sitting above Android (like HTC's Sense UI).
In short: somewhat merging Open webOS (or at least Luna) and Android would bring a real interesting iOS- / Windows- killer (well, ahem, as if Windows 8 RT needed any efforts in killing...)

- maybe some enthousiats groups will release newer hardware for open web os. I mean openmoko is a really small comunity, and that hasn't stopped the people at golden delicious to design build and ship a hardware upgrade for the freerunner (a newer mobo based around an OMAP similar to what is found in WebOS devices, most android phones and most iphones).

- just look at the GP32, GP2X, Pandora, and the like : a project doesn't need multi-million shipped device to thrive. a small community of enthousiasts is more than enough to keep it alive. and web os has a small comunity of dedicated enthousiast. just look at the official webos app store, or even at the open-source homebrew app store. that's a tremendous production of apps for a platform that is so much dwarfed by the number of Android and iOS devices.

- Who knows, it an Android/open webOS hybrid is well done (please CyanogenMod, please! start collaborating with WebOS internals to make it happen), some smaller device maker might be interested into picking it up. (I mean stock CyanogenMod has already been approached by some manufacturer).

Or even fuse them

[DrYak]

The closer to a gnu userspace the better. Android may have the linux kernel, but it doesn't come close to the hackability of a maemo device. With maemo dead, maybe webos will get us something better than android to start with.

And why not fusing them together to get the best of both worlds? Leverage the big ecosystem of android, while take advantage of the awesome GUI and better GNU/Linux of open webOS ?

Recently the Android peculiar stuff (all of the IPC, although not all the power-saving optimisation) is being backported into the stock kernel. Free-source android can run on a stock Linux 3x. kernel (although with worse battery life).

Palm didn't optimize much beside power-saving, thus open webos should be able to run on stock Linux 3.x kernels too (thats one of their target) - althoug with worse battery life too.

Canonical dual-mode device (an android smartphone when in your pocket, but a full fledged Ubuntu ARM desktop when plugged to a monitor over HDMI) has show that it is possible to get an (almost) stock GNU/Linux OS running above an (almost) stock Android.

Thus with some work, open webOS could be made to run along Android.

Run all the existing Android apps, but get the nice modern multitasking capable UI of Luna's card based interface, and get the ability to develop easily HTML5+Javascript apps thanks to Luna and Enyo frameworks.

That can be a killer feature which could interest a lot of current webos internal developpers, and potential Chinese/Korean/Eastern European manufacturers. It would help them distinguish from regular Android machines.

(Just like HTC is trying to distinguish themself with their own custom UI - HTC Sense).

Re:Adopting it to other OSS project?

[scot4875]

I never own a touch-pad so I do not know how the thing functions or what's so special about it

It's pretty slick. I picked one up during the fire sale to root and use as a cheap Android tablet, but WebOS is really quite nice. In my experience with tablet OSes, I'd have to say it beats iOS and pre-ICS Android hands-down. ICS is a little more of a toss-up, but I still found WebOS to be a more intuitive environment to navigate.

A pity that it'll never really get any use. If it actually gained traction and someone released a distro that would work on one of my phones, I'd definitely give it a trial run.

--Jeremy

and in other news...

[jds91md]

webOS. Pretty soon you'll have to Google that in order to remember what it meant. Too little too late. --Josh (former Palm developer)

Re:Adopting it to other OSS project?

[espiesp]

Before I got ICS installed on my Touchpad, I have to say WebOS > Android except for one major shortcoming: Apps. If WebOS had the same apps as Android I'd still be using it. But now, ICS multitasking is at least as good as WebOS, and it has all the apps. The notification system in ICS brings Android to par with WebOS as well.

Shame WebOS was too little, too late, too small. Because overall it really had a lot of things right.

Re:Which license, bitches?

[LizardKing]

Apache APL2.0 for the GUI stuff, GPL for a lot of the stuff that had already been released.