David Beers Weighs in on Foleo (finally)

[SeanH]

Sean, you might notice there is no forum at Brighthand for RAZR under Motorola. It is not a smartphone. The only discussion of RAZR that is considered on-topic is posts inquiring how to pair via Bluetooth with a PDA (a non-cellular device).

The topic is David Beers claimed that the reason the Foloe was not successful was because carries did not want linux devices on there network. The new RAZR is a prefect example of a device that runs Linux for its OS and Linux apps for the multimedia and PIM functions. The four different versions of the RAZR were released for four different carriers. I was using the RAZR as an example of the acceptance of Linux on Mobile Devices.

SeanH

[r0k]

Now that (I hope) this thread has passed a rather tedious phase, I went back and read Beers' article again.

Beers is claiming the carriers said "no native 3rd party apps at all". Perhaps this would explain that some EU carriers are slow to adopt Palmos. But if this were true, how the heck did so many WM phones get out there? I think the carriers will cheerfully tolerate a general purpose OS on a "smartphone" as long as it isn't so unstable their customers cannot run up all those expensive minutes. There is something to be said for a JVM as it allows an app to run in a more or less firewalled environment that has a much lower risk of bringing down the whole phone. But any OS with a protected mode kernel can offer the same degree of stability.

I don't think it was the carrier who told Apple there could be no native apps. Remember that Att wanted the iphone bad. It certainly wasn't winning customers with it's spotty GSM performance in the US. Apple had a closed architecture in mind from the start. Lucky for Palm, this last remaining snafu gives them a window of opportunity to get their stuff together and put out a competitive OS. But I think Apple will wake up long before 2009 so the glacier over at Palm had better thaw a little more quickly lest the ice age become permanent.

[SeanH]

Sean, for my part, I have to ask--why do you still care? If you disliked the Foleo that much, relax, it's dead. Why do you feel the need to continue arguing?

I am responding to your question in this thread because you asked the question in this thread. I am not trying to change the topic of this thread.

I respond to threads about the Foleo because it’s architecture I have studied for many years, way before the Foleo was introduced. Its an ARM based laptop. I watched other devices like it fail over time. I not arguing, I am posting my objective view. I understand that people wanted to the Folio to be successful and they get upset when they read facts on why it failed. When I post my objective view I do not call people names or send links to people to go to hooked-on-phonics.com like some did to me in this very thread. I post facts and I back my facts up with links, pictures and videos. I do not believe its Brighthands practice to filter out objective views on products.

And for that matter, why do you ignore the attempts of the moderation team to get in touch with you?

Since 2003 I have received 14 private messages. 12 of those were mainly people asking questions on what battery config I use in my IBM T42 or how did to get YouTube to stream on your phone. Two were from moderators. The first had nothing to do with phones or the Foleo, it was about a technology I was using. The only post I received from a moderator that had anything to do with the Foleo was a link to a article that asked people to stop comparing the Foleo with a laptop. I did respond and claimed it was a laptop. Most of the articles you read about the failure of the Foleo call it a laptop.

SeanH

[cervezas]

Thanks for the comments on my piece. Just a few clarifications and amplifications.

First, what I mean when I say a "open, native Linux platform" is an operating system that supports user-installable applications written in C/C++ on a Linux kernel. There are plenty of Linux phones out there in the mass market--mostly in Asia--and Motorola has said on more than one occasion that they'll be going completely over to Linux. But to date the only kind of software you can install on the vast majority of these Linux phones is Java MIDlets. The RAZR II is no exception. The Foleo, on the other hand, is a native Linux platform.

I disagree with Adama's characterization of the Foleo platform being "a customized version of straight-up Linux." I got early access to Foleo to develop software for it (not as a blogger) and I can tell you that it's a substantial departure from any other Linux platform that I've seen. It's not just that Palm completely dispensed with X Windows, which is the basis of just about every Linux GUI, including ALP and Maemo. It's that important mods were made to deliver instant-on and instant-application-switching on a PDA-class 400MHz Bulverde processor. In my judgment, the Foleo OS was designed from the very start with the idea that it must be easy to adapt to Palm's smartphones.

When I said that I suspected Foleo was canceled because of the carriers, I didn't mean that the carriers had anything directly to do with Palm's plans to sell it. I meant that I don't believe that Palm could support two fundamentally different operating systems for their smartphones and mobile companions. So when Palm made a decision (because of the carriers or for any other reason) to change the architecture for its smartphones earlier this year, and I think a lot of evidence supports this, they had ultimately to face the hard fact that this meant Foleo could not be released with the original OS.

The suspicion that the cause for change in the smartphone architecture was prompted by carrier requirements is mostly a supposition based on my personal experience as a phone software developer working with the business and developer organizations at Sprint, Verizon and Cingular. I don't claim to have any hard evidence to back this supposition up, but I think if you're looking for an explanation of why Palm would make such a change at such a late date, it's a pretty good place to start. I'm certainly open to others, and if it turns out I'm wrong it doesn't change the thrust of my argument, which was that Foleo was canceled because Palm made changes in their smartphone platform architecture and didn't want to support two different architectures, with separate tools, separate SDKs, even separate programming languages.

I agree that Palm was probably having regrets about trying to rush the Foleo to market without the application support that it needed. And I hope they've recognized that they made some mistakes in the way they tried to market it. But I don't think any of these regrets would have been enough to withdraw it at the last minute, since they could have been addressed over a fairly short timeframe after the initial release. I think the decision was exactly what Colligan said it was: a decision to "offer a single, consistent user experience around this new platform design and a single focus for our platform development efforts." Their Linux OS had become two Linux OSes, and they had to get back to one.

[Adama D. Brown]

The topic is David Beers claimed that the reason the Foloe was not successful was because carries did not want linux devices on there network.

Which is a dubious statement, I agree, though the RAZR isn't the best example. Like most of the Linux phones found in Asia, the RAZR is really just a dumbphone that uses Linux for convenience, rather than as a real platform. However, I don't see the difference in application models between Linux and WM/Garnet/Blackberry to be significant. The carriers have long since learned they can't control applications on a smartphone.

When I post my objective view I do not call people names or send links to people to go to hooked-on-phonics.com like some did to me in this very thread. I post facts and I back my facts up with links, pictures and videos. I do not believe its Brighands practice to filter out objective views on products.

It's not. But when it reaches the level of an obsession, it behooves us to try and move the discussion along, and you don't seem to want to talk about anything else.

The only post I received from a moderator that had anything to do with the Foleo was a link to a article that asked people to stop comparing the Foleo with a laptop.

Odd, since we've been trying to get in touch with you via multiple PMs, and I sent an email via the address listed for your Brighthand account.

[Adama D. Brown]

I disagree with Adama's characterization of the Foleo platform being "a customized version of straight-up Linux."

What I meant was "not a handheld OS." No touchscreen, keyboard and mouse style input, no Garnet compatibility, and from what I saw of the UI, not something that could be easily shrunk. In other words, most of the hard parts would still have to me done fresh for a mobile device OS. It may have had some components ripped out for compactness, but that doesn't neccessarily translate to creating a compact phone-style OS.

It's that important mods were made to deliver instant-on and instant-application-switching on a PDA-class 400MHz Bulverde processor.

That's not exactly new tech, even on mobile Linux. See "Zaurus," as well as "Windows Mobile."

[cervezas]

What I meant was "not a handheld OS." No touchscreen, keyboard and mouse style input, no Garnet compatibility, and from what I saw of the UI, not something that could be easily shrunk. In other words, most of the hard parts would still have to me done fresh for a mobile device OS.

I write apps that need to run on touchscreen and non-touchscreen devices, with or without mouse pointers, and with screens that scale from 160x160 up to full-screen desktop resolution, and I've also worked on platform code that makes this possible. I'm not trying to be humble when I say it's not at all one of the "hard parts." Pen events and mouse-click events map very easily to each other, for example, and whether there is a mouse pointer or not is a simple flag.

It's also not hard to write an application that takes Foleo's side-by-side master-detail panes and displays them as stacked windows the way we're accustomed to seeing in Palm OS. I'd just written a simple Java framework to do this for Treo and Foleo when they canceled the Foleo.

The "handheld OS" features that are in place on Foleo that show signs of being intended as such: the menu bar (almost exactly like Palm OS), the task switcher (similar to tapping and holding the Home button but accessed from a dedicated key), the instantaneous character of application switching (a lot of work there), and the physical navigation controls that map rather conspicuously to those you might find on a phone. Not exactly a D-pad, but could easily map to one. Or (my hope) Palm may be considering a scroll-wheel for their future smartphones, similar to the one that RIM gave up to a loud chorus of complaints.

I've not used a Zaurus personally, but I've never heard anyone say that it was a snappy operating system and I've heard several complaints to the contrary. Foleo is as responsive in most situations as a good Palm OS PDA. And if I remember correctly, the cold reboot (almost never necessary) takes only 25 seconds--much less than most Palm OS devices these days. Foleo does something that Windows Mobile has never done: it puts inactive applications to sleep in a way where the system can recover many of the resources they were using and "reconstitute" them quickly when they become active again. This, too, sounds like something that Palm did to deliver full multitasking within the tight memory constraints of a smartphone or PDA.

None of that proves that was the intent, of course. But I just don't believe that the product guys at Palm would think it was fine for the company to attempt two separate in-house operating systems. I think that state of affairs had to have been something that was thrust upon them by an unexpected circumstance, and one they had to remedy with a painful "exfoleation" when no other reasonable alternative could be found.

[lelisa13p]

The only post I received from a moderator that had anything to do with the Foleo was a link to a article that asked people to stop comparing the Foleo with a laptop. I did respond and claimed it was a laptop.

AAMOF, I have on file saved copies of every PM that I've sent, and the replies, if any. Call it Fact Insurance.

[Adama D. Brown]

I nearly forgot to come back and answer this, but I'm glad that a trip through the PalmInfoCenter archives reminded me of this thread tonight.

The "handheld OS" features that are in place on Foleo that show signs of being intended as such: the menu bar (almost exactly like Palm OS), the task switcher (similar to tapping and holding the Home button but accessed from a dedicated key), the instantaneous character of application switching (a lot of work there), and the physical navigation controls that map rather conspicuously to those you might find on a phone. Not exactly a D-pad, but could easily map to one.

Hasn't a set of directional controls been standard on pretty much every computer ever built? I don't mean to try and sound like I know more about coding than you do, because I don't, but it strikes me that most of what you describe is fairly standard features for any computer. More to the point, if the user interface, APIs, and top level infrastructure were the "easy part," why would it have proven so impossible for PalmSource to do Cobalt well when they had the kernel built, or for that matter why it has taken them so long to do PalmLinux?

I've not used a Zaurus personally, but I've never heard anyone say that it was a snappy operating system and I've heard several complaints to the contrary.

I've heard that the later versions were much improved over the buggy 5000 series in terms of performance, though I've never used them myself either. In any event, how fast it runs was never the question, the point was that the Zauruses did instant-on, as well as instant application switching.

None of that proves that was the intent, of course. But I just don't believe that the product guys at Palm would think it was fine for the company to attempt two separate operating systems. I think that state of affairs had to have been something that was thrust upon them by an unexpected circumstance, and one they had to remedy with a painful "exfoleation" when no other reasonable alternative could be found.

David, I hate to put it this way, but I have a far better record on accuracy then you do. The following was borrowed from a PIC thread in August of 2005, over two years ago.

what evidence is there that Palm has licensed PLinux? All they've done is extend their license for PalmOS. I've seen nothing to indicate that they've committed to delivering Linux or anything else.

So you think Palm plans to just keep patching Garnet through 2009? Honestly, some of you guys are so afraid not to look cynical in this forum that you end up saying things that are downright stupid.

Why would Palm have renewed their Palm OS license for five years if they weren't on board with the roadmap for the OS? People can debate as to whether it's the Cobalt or Linux part of the roadmap that their on board with at this point, but there's no question that Palm has designs on one or the other, if not both. It's only a question of when.

http://www.palminfocenter.com/comments/8042/#111205

I'll also note that this was shortly after I was mostly denounced for speculating that Palm Inc's hiring of Linux developers was either preparation for building their own OS, or replacements for a buyout of PalmSource. They've since attempted both.

Boy, the PIC archives are fun. I'd forgotten how crazy it was there back in the day.

[r0k]

Thanks for the comments on my piece. Just a few clarifications and amplifications.

First, what I mean when I say a "open, native Linux platform" is an operating system that supports user-installable applications written in C/C++ on a Linux kernel. There are plenty of Linux phones out there in the mass market--mostly in Asia--and Motorola has said on more than one occasion that they'll be going completely over to Linux. But to date the only kind of software you can install on the vast majority of these Linux phones is Java MIDlets. The RAZR II is no exception. The Foleo, on the other hand, is a native Linux platform.

I disagree with Adama's characterization of the Foleo platform being "a customized version of straight-up Linux." I got early access to Foleo to develop software for it (not as a blogger) and I can tell you that it's a substantial departure from any other Linux platform that I've seen. It's not just that Palm completely dispensed with X Windows, which is the basis of just about every Linux GUI, including ALP and Maemo. It's that important mods were made to deliver instant-on and instant-application-switching on a PDA-class 400MHz Bulverde processor. In my judgment, the Foleo OS was designed from the very start with the idea that it must be easy to adapt to Palm's smartphones.

When I said that I suspected Foleo was canceled because of the carriers, I didn't mean that the carriers had anything directly to do with Palm's plans to sell it. I meant that I don't believe that Palm could support two fundamentally different operating systems for their smartphones and mobile companions. So when Palm made a decision (because of the carriers or for any other reason) to change the architecture for its smartphones earlier this year, and I think a lot of evidence supports this, they had ultimately to face the hard fact that this meant Foleo could not be released with the original OS.

The suspicion that the cause for change in the smartphone architecture was prompted by carrier requirements is mostly a supposition based on my personal experience as a phone software developer working with the business and developer organizations at Sprint, Verizon and Cingular. I don't claim to have any hard evidence to back this supposition up, but I think if you're looking for an explanation of why Palm would make such a change at such a late date, it's a pretty good place to start. I'm certainly open to others, and if it turns out I'm wrong it doesn't change the thrust of my argument, which was that Foleo was canceled because Palm made changes in their smartphone platform architecture and didn't want to support two different architectures, with separate tools, separate SDKs, even separate programming languages.

I agree that Palm was probably having regrets about trying to rush the Foleo to market without the application support that it needed. And I hope they've recognized that they made some mistakes in the way they tried to market it. But I don't think any of these regrets would have been enough to withdraw it at the last minute, since they could have been addressed over a fairly short timeframe after the initial release. I think the decision was exactly what Colligan said it was: a decision to "offer a single, consistent user experience around this new platform design and a single focus for our platform development efforts." Their Linux OS had become two Linux OSes, and they had to get back to one.

David,

Thanks for these clarifications! There has been a guy from Access posting in another thread saying that there will be ALP devices early next year barring an extinction level event. I hope Palm turns out to be one of the licensees, or at least the devices are available in the US so I can acquire an updated PalmOS2/ALP/anything beyond Garnet device before 2009.