LiPS service [LWN.net]

One might think that there are already enough industry bodies working on Linux in telephone applications. But, on November 14, a new group, called the "Linux Phone Standards Forum" (or "LiPS") announced its existence. According to the release:

The LiPS Forum will accelerate the adoption of Linux in fixed, mobile and converged devices by standardizing Linux-based services and APIs that most directly influence the development, deployment and interoperability of applications and user-level services.

In essence, LiPS wants to push toward the creation of a standard low-level phone platform which allows vendors to focus their efforts on the higher-level features which set their offerings apart. The appeal of this idea is not that hard to understand. As an operating system for telephones, Linux is hard to beat: it can be customized to taste, it is efficient, and it lacks per-unit royalty costs. In addition, mobile platforms have become powerful enough to run Linux, and many mobile applications are sufficiently demanding to require a complete operating system like Linux. On the other hand, Linux lacks the features specific to telephony which can be found in a proprietary platform like Symbian. By filling in that layer of telephony-specific features, LiPS hopes to create a competitive platform for future products.

LiPS will probably be successful in scheduling meetings, generating white papers, and cranking out press releases. But if LiPS truly wants to turn Linux into a platform it can rely upon in the future, its management may want to consider engaging openly with the development community; "cooperating with OSDL" is not sufficient in this regard. If LiPS sees itself as another proprietary, members-only consortium, it will cut itself off from much that the community can provide.

A good start would be to admit some community projects to the group. For example, since they claim to be trying to build platforms for telephony in general - not limited to mobile devices - the LiPS member companies might well benefit from having somebody from the Asterisk and Bayonne projects at the table.

Even better would be to work with the community directly. A look at the list of companies which have joined LiPS (ARM, Cellon, Esmertec, France Telecom/Orange, FSM Labs, Huawei, Jaluna, MIZI Research, MontaVista Software, Open-Plug and PalmSource) and the other companies which have been active in Linux-based telephones (Motorola, Haier, Nokia, NEC, Panasonic, Samsung, ...) has few intersections with the list of companies participating in Linux kernel development. If the LiPS members truly want to get the most out of Linux, they will be better off working with the development community and contributing back their improvements. The recent announcement by the Consumer Electronics Linux Forum that it had hired a Linux kernel developer is a step in the right direction, but it is only a beginning.

Finally, if LiPS truly wants to achieve world domination with Linux-based phones, it should give some thought to the creation of a user-hackable platform. A phone which can be extended to perform functions never envisioned by its creators will be a far more valuable device, and it should find a wider market. Unfortunately, the mobile phone market tends to be dominated by companies which behave like, well, telephone companies, with the result that even routine features (such as Bluetooth) can be locked down, and user-hackable devices are a rarity. When a device is fully locked down, it matters little to the user whether it is running Linux or something else altogether. If LiPS were sufficiently enlightened that it could go against the closed nature of the industry and specify the creation of Linux-based phones which have not had the natural freedom of Linux stripped out of them, it could be the start of something truly interesting.

to post comments

LiPS service

Posted Nov 17, 2005 5:18 UTC (Thu) by JoeBuck (subscriber, #2330) [Link] (2 responses)

Unfortunately, user-hackable phones are the last things these guys want. The reason is that the customers for cell phones in the US, as well as in a number of other countries, are really the big carriers (Verizon, Cingular, etc), and control over what can be run on the phone is something they see as a key asset, to be converted into money. In many cases, a cell phone from, say, Samsung will lack an advertised feature when you want to use it with, say, Verizon because Verizon doesn't want its customers to have that feature.

If people can directly control the sound the phone makes when a call comes in, there goes the highly profitable ring tone business. Similarly, if the carrier makes extra money for video messages on camera phones, they will prevent anyone from programming a way to send that video through a free Wi-Fi connection; better to force the customer to make a call through their network so they can charge the fee.

Maybe a maverick carrier could break up the system. Or maybe governments could force the carriers to allow customers to choose whatever phones they want. Or maybe if the carriers refuse and Wi-Fi keeps taking off, everyone will just bypass them. Or maybe the big guys will just buy up enough congresspeople to protect their businesses.

LiPS service

Posted Nov 17, 2005 17:23 UTC (Thu) by iabervon (subscriber, #722) [Link] (1 responses)

The LiPS members aren't the carriers, though. I bet they'd love to have hackable devices if they could. And there's no reason they couldn't make a platform which could be extensible or could be locked down, depending on the configuration. If they were using a sufficiently flexible underlying platform, they could make it so the only thing that cared that the user hadn't changed stuff was the program to connect to the phone network. And they could allow the option of VoIP with wifi and let hackers play with it in that configuration.

LiPS service

Posted Nov 18, 2005 0:08 UTC (Fri) by sepreece (guest, #19270) [Link]

Actually, LiPS was founded specifically to create an API driven by the operators' perspective. It is heavily driven by carrier requirements.

That said, the Forum's goals aren't particularly aimed at keeping phones non-hackable, but at making it possible for the carriers to develop their own software, implementing their own look-and-feel/branding, and have it be portable across handsets from multiple vendors. There may be security requirements that end up having the effect of making the phones hard or impossible to hack, but the primary reason for those requirements would be to keep other people from messing with your phone. Keeping you from messing with it just happens as a side effect.

People who make handsets and people who sell handsets are really, REALLY eager to keep the virus problem out of phones and off their networks. Making phones hackable won't happen unless the vendors and carriers are convinced that (a) it helps them make money and (b) it doesn't open the phones and network to malware.

DISCLAIMER: I work for a handset manufacturer.

LiPS service: LWN meta

Posted Nov 17, 2005 8:51 UTC (Thu) by Duncan (guest, #6647) [Link] (4 responses)

As is often the case, I saw noise about this from a couple other sources
first, but count on LWN to have an article bringing some sense to the
noise. One continuing reason LWN is worth my subcription money!

Duncan

LiPS service: LWN meta

Posted Nov 17, 2005 8:56 UTC (Thu) by Duncan (guest, #6647) [Link] (3 responses)

BTW, why does the weekly front-page story continue to report ZERO
comments, when there was one when I clicked to submit my own, and after
submitting it, two when I updated the front page, only to see ZERO still
listed?

Often, the comments are quite enlightening, so I'll click thru to the
story to read the comments if it says there are any. If it lies and says
there are none, I could be missing some interesting comments!

Duncan

LiPS service: LWN meta

Posted Nov 17, 2005 12:29 UTC (Thu) by Duncan (guest, #6647) [Link]

... And now there are five posts, this will make six, and it /still/ says
no posts on the weekly cover article! Time to file a bug I guess!

Comment count

Posted Nov 17, 2005 14:55 UTC (Thu) by corbet (editor, #1) [Link] (1 responses)

Very strange. There will normally be a delay before the comment count is updated (the weekly page gets regenerated in the background), but it shouldn't be this long. I'll have to look into it...

Comment count

Posted Nov 17, 2005 18:34 UTC (Thu) by Duncan (guest, #6647) [Link]

FWIW, I mailed lwn@, with the response here shortly thereafter, and a
reply to my mail a bit after that with a summary of the fix (thanks!).
Turned out it was a caching daemon that needed kicked.

It's working again now!

Thanks for the great site, and thanks for the resolution report saying
what it was. Those are nice to get, for folks like me that take the time
to "file the bug".

Duncan

LiPS service

Posted Nov 17, 2005 10:57 UTC (Thu) by nedrichards (guest, #23295) [Link]

Very *very* little of what goes into making a really good phone has anything to do with the kernel. At least once you've got beyond a certain base. Sure there's always optimisations to be made (and Symbian have done well by introducing a hard realtime kernel recently) but really, almost all stuff that needs doing is in userspace. And if you look at what some of the companies (well, at least the one I work for) are doing you'll see that there is engagement going on in individual userspace projects. Not that we couldn't do a lot better (and really should be) but the kernel alone isn't everything or even anything near enough when it comes to making a really compelling mobile device.

A really good way of doing this sort of thing (and producing a hackable platform along the way) is the way Nokia have worked with the open source/free software community over <a href="http://maemo.org">Maemo</a>. I don't think that LiPS is intended to be that sort of thing but it'd be nice if it would be.

Insert tiresome disclaimer about not speaking for employer here.

LiPS service

Posted Nov 17, 2005 11:40 UTC (Thu) by ikm (guest, #493) [Link]

> When a device is fully locked down, it matters little to the user whether it is running Linux or something else altogether.

I would like to note that it still matters whether the device runs Linux or not, since the kernel is still under the GPL, allowing some explorations and modifications to start with at least.