Palm 3Q Earnings Call Highlights

Palm held their fiscal third quarter conference call today with analysts and shareholders. Palm's CEO Ed Colligan gave investors an update on Palm Inc's performance during the past quarter and presented his thoughts on the growth of Treo smartphones. He laid out a lot of interesting statistics on smartphone sales, and briefly talked about the upcoming new treo phones due out this year. Read on for the full report.

The conference call started off with Palm CEO Ed Colligan talking about the results of this latest quarter. He feels that Palm Inc executed very well. He mentions that this was Palm 9th consecutive quarter of year over year revenue growth.

He then went over Palm's five corporate objectives:

  • Deliver consistent, profitable growth
  • Invest in and strengthen our product engine
  • Sharpen our attention on the customer experience
  • Expand geographically
  • Collaborate with partners to reach new markets

Colligan went on to state that overall Treo sell-through was a record 569,000 units, up 102% year-over-year, and 756,000 handhelds shipped. Interest in both the Treo 650 and Treo 700w remains high. The Treo 650 is now offered on over 70 carriers worldwide and some carriers even ran low on treo inventory during the quarter.

Smartphone shipments in the US grew by 111%. The Treo now has 30% of the US smartphone market share according to Canalys vs 22% a year ago. Revenue for this quarter was 74% from treo's and 26% from handhelds, reflecting the growth and higher average selling price of smartphones.

Treo 700w demand was strong and the company says that 90% of 700w customers knew they wanted a Treo smartphone and 50% of them were new, first-time smartphone buyers. He also stated that 29% of registered 650 customers switched carriers just to get a Treo 650.

Colligan reiterated that we are still in the early stages of the migration to smartphones. He feels mobile users will eventually look to do more with their current 12-key feature phone based mobile devices and will "naturally evolve" to smartphones. He also mentions that only 2% of business email today is accessed wirelessly. He also talked about Palm's new stagey to drive growth in the European market. Palm is now investing much more in Treo R&D, and the new smartphone development center in Dublin is working on new treo software development.

New Treo smartphones
Ed again confirmed that Palm plans to announce three additional smartphones this calendar year. He said that there are both new Palm OS and Windows Mobile products on the roadmap. These new units will feature a combination of operating system choices, different hardware designs, wireless radios and price points. Rumors about upcoming treo code names Hollywood and Lowrider first leaked last November.

Palm Treo and PDA Sales ChartTreo 700w Patch Coming
Colligan also promised that a system update is being developed for the Treo 700w. This update will introduce the Exchange direct Push email capability, new global address lookup and a "kill pill" ability for wireless data security for IT management.

He also stated that the Palm OS BlackBerry connect email software is completed and is currently pending carrier certification.

Financials
The call was then turned over to the financial side, and Palm CFO Andrew Brown went over the numbers for the quarter. Overall revenue was $388.5 million, a 36% increase from the same period last year. Cash flow from Operations was $91.2 million.

He said that smartphone demand remains strong and there was a increase in both the sell-trough and gross margin. GAAP EPS was .28c per share, EPDS was .19c. Guidance for the 4th Quarter is $400 - $500m revenue and 33.5% to 33.7% gross margin.

When asked if he expected a rise in the handheld sales for the next quarter, Mr. Brown responded that he expects sales to remain flat. This may indicate that we may not see any new PDAs this spring, as this upcoming quarter is traditionally the time that Palm introduces new handheld models. He also stated that the handheld side of the business was still profitable even with flat sales growth.

In the Q&A session Ed Colligan confirmed a rumor that longtime Palm Executive Ken Wirt has left the company. Ed stated that Palm is currently recruiting for a new product marketing manager. He also declined to speculate on any unannounced products, rumors or on the possibility of competing with an Apple smartphone.

Article Comments

 (78 comments)

The following comments are owned by whoever posted them. PalmInfocenter is not responsible for them in any way.
Please Login or register here to add your comments.

Start a new Comment Down View Full Comment Thread

The LD is dead. The TX is dead.

Surur @ 3/23/2006 8:27:00 PM # Q

$100 million/ 765 000 devices = $130 We are talking mostly Zire22's and a smattering of TX's.

When the R&D costs of developing handhelds exceed the profit they generate they will be dropped. This wont be long. As HP said, disconnected handhelds are dead.

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: The LD is dead. The TX is dead.
hkklife @ 3/23/2006 8:57:27 PM # Q
LD aside, Palm hasn't spent any consdierable (note "CONSIDERABLE") R&D $ on a conventional PDA in years...not since the T3, really.

I'd actually say that those numbers average out to a lot of Z22s plus an equal spread of T|E2 and TXs. I have actually been told by people somewhat in the konw on a retail level that the TX & E2 are brisk sellers.

One could argue that Palm has the "traditional" PDA market completely to themselves now. HP is fading fast, Dell might throw in the towel any day now and have zero retail presence. The other players (Asus etc) are minor players at best.

As long as Palm can keep restyling & respinning the tried & true Zire/ T|E/ TX meat'n potatoes, that market has another two years of life left. Zero growth, yes, but still with enough interest from current Palm owners or soccer moms to warrant keeping 2-3 models alive. Just because it makes sense doesn't mean Palm will do it. Surur, look at the Casio/Royal/Sharp "digital diary" market. That stuff was primitive in 1990 but they are stil available today at any good mass merchant, blister packed and priced around $40-$50. If a mono Zire could be sold for $50 and a color Zire for $80-$100, there's a slice of the market where Palm & FrankenGarnet looks absolutely high-tech in comparison to the 128k of memory competition .

At this point, producing, packing & marketing these machines likely costs more than comping up with new industrial designs, especially since the tooling for the old form factors was long ago paid for. In theory you'd think something like the TX or E2 would actually have pretty decent margins-no cellular radio, dirt cheap NVFS RAM, no charge LED or vibarting alarm or voice recorder.

Anyone care to hazard a guess what the margin is on the Z22/T|E2/TX vs a Treo?

Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX

RE: The LD is dead. The TX is dead.
AdamaDBrown @ 3/23/2006 9:00:54 PM # Q
The thing is, those numbers talk about Palm's cut, not neccessarily retail price. A Z22 might make $50 for Palm, an E2 $120, etcetera. So it might not be quite as grim for the handheld side as it's made out to be. The Treos are three quarters of the revenue because they're both popular and very expensive.

RE: The LD is dead. The TX is dead.
Simony @ 3/23/2006 9:02:48 PM # Q
I think you need a lesson in basic cost accounting.

R&D expenditures are normally factored into the 'standard' cost of inventory (ie, on a per unit basis). The 'gross margin' represents the amount by which the net sale price per unit is greater than the 'standard' cost (but it does not including admin, selling and distribution expenses).

The moral to this little parable is that if the 'gross margin' holds up it automatically covers R&D costs.

There is nothing the numbers which indicates that the 'gross margin' on handhelds is being eroded. If anything, the 'gross margin' may hold at current levels since: (a) Palm is already undercutting everyone else in its pricing; and (b) a major player (HP) seems to be exiting this market. In other words, there may be no big market factors which would reduce the 'gross margin'.

If HP really does abandon this market segment, that may provide opportunities for Palm and other players.

I'm speculating, of course - unlike you, I am not able to predict the future.

RE: The LD is dead. The TX is dead.
rsc1000 @ 3/24/2006 12:43:34 PM # Q
Nothing grim or unexpected in these numbers / results at all. Surur is just a funny, funny little man with nothing better to do then sit there - hitting refresh on the PIC home page every 10 seconds waiting for a new story to appear so that he can be the first to post with a negative spin. Honestly Surur - you go to great lengths to try and turn gold into dung. Pretty amusing although it is getting a little thin.

let see now:
- PIC has a story yesterday about a Wall Street Journal article on how Palm capitlized on the RIM uncertainty

Suru missed being the 1st post, so he jumped in with "I find it shocking that their volume of enquirers and number of sales leads were so low. 600 is really unimpressive. It seems they mostly got a bit of PR from the RIM drama, but very little new business."

- Then these 2 stories about the Q3 results (proving his negative assessment of Palms business acumen completey wrong) and he manages 1st post on both, with the Subject headings here:

"The LD is dead. The TX is dead."

...anything for a negative spin!

Wow - palm is crumbling as was prophesized! Falling apart! Oh great TVOR and Surur - guide the way with your supernatural clairvoyance lead us from this digital desert!

LOL. Funny little guy!



RE: The LD is dead. The TX is dead.
Slam @ 3/24/2006 12:58:46 PM # Q
Surur,
The math above may be wrong. Palm had 510,000 units of "sell-in" for PDAs (as stated in their conf call). The number you quote above is "sell thru" reported by the channel. The "Sell-in" number is what you should compare to the $100M in revenue. This brings Palm's PDA ASPs to $196. This suggests some more healthy selling of the higher end PDAs.



RE: The LD is dead. The TX is dead.
Ryan @ 3/24/2006 1:13:46 PM # Q
ok, that explains why the chart says 765 units, while Andrew Brown said 510k. I got confused by that stat as well.
RE: The LD is dead. The TX is dead.
Surur @ 3/24/2006 1:20:55 PM # Q

That does make a big difference, and suggest a balance closer to the TX. The LD is still doomed however.

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

Reply to this comment

Breakdown of the new Treos

AdamaDBrown @ 3/23/2006 9:38:58 PM # Q
Did anyone else hear Colligan (I think it was him) talking specifically about the definition of "new" smartphones? My initial impression from that was that when they refer to three more new Treos coming out this year, that they don't mean the 700w GSM plus 700p CDMA and GSM, they really mean new individual models. Upon reflection, though, I'm not quite as sure. Anyone else have a view?

RE: Breakdown of the new Treos
hkklife @ 3/23/2006 10:22:52 PM # Q
Being the bit of a cynic that I am, I still maintain it's the 700w (Verizon with Sprint coming this summer, CDMA only) followed by:

#2 700p (Sprint CDMA only)
#3 Hollywood (WinMob GSM only)
#4 Lowrider (POS, GSM & CDMA depending on carrier/market, nearly identical specs to Treo 600 but with NVFS).

So if you differentiate between the Lowrider CDMA & Lowrider GSM, that's five total new "Treos" available this year plus the existing stocks of 650s. By any stretch, that's still a LOT of new product for Palm.

I also still think that a T|E3 of some kind will come out this spring or early summer. The E2 is a good seller and needs to brought up to spec with "Palm" branding. The $200 gulf between the Z22 and the TX is just too large.

Dropping the LD altogether and bringing out a mildly refreshed E3 would give Palm a PDA at three consecutive $100 price points and a decent final lineup with one model for each resolution standard (160*160, 320*320, 320*480).

Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX

RE: Breakdown of the new Treos
twizza @ 3/23/2006 11:15:38 PM # Q
You all have an interesting take on those comments, I just posted some thoughts at BPDA a bit ago,but kinda want to repeat them here to get thoughts as well as to just throw something out there.

First about the handhelds comment they made: honestly I think that if Palm wants to keep the PDA dev small and make it a true 25% of revenue, then they will bring out a new model that will call attention again to the PDA side. My guess is either an LD or T with some "needed" specs and a good price. Doing that will keep interest there and make the leaving of HP to the space better for them. It would have to be a good one though, I am in the same belief that the TX and E2 are selling nicely. For Palm to tweak that formula some now would be good.

Second comment (sorry posting via treo and mobile PIC doesn't do discussion posts); I expect Palm to bring out an all new Treo in Oct, maybe Nov. The Holloywood GSM one. The P should be a summer intro, late May keeping the pressure on themselves to execute (my fav word on that call today). A P done right would give the competitive numbers needed to really put palm on the burner come that question in the next earnings call (not like how they skipped around it this time)..

Personally, I was impressed. I knew that Palm was doing well. But they are doing better than I expected. I still think they get one quarter of eww before being totally good again, but I also think this call made them even more ripe for acquisition. The year will be interesting. It is for me, and I only have a Treo to speak for it ;-)

mobileministrymagazine.com
antoinerjwright.com

RE: Breakdown of the new Treos
twizza @ 3/23/2006 11:15:38 PM # Q
You all have an interesting take on those comments, I just posted some thoughts at BPDA a bit ago,but kinda want to repeat them here to get thoughts as well as to just throw something out there.

First about the handhelds comment they made: honestly I think that if Palm wants to keep the PDA dev small and make it a true 25% of revenue, then they will bring out a new model that will call attention again to the PDA side. My guess is either an LD or T with some "needed" specs and a good price. Doing that will keep interest there and make the leaving of HP to the space better for them. It would have to be a good one though, I am in the same belief that the TX and E2 are selling nicely. For Palm to tweak that formula some now would be good.

Second comment (sorry posting via treo and mobile PIC doesn't do discussion posts); I expect Palm to bring out an all new Treo in Oct, maybe Nov. The Holloywood GSM one. The P should be a summer intro, late May keeping the pressure on themselves to execute (my fav word on that call today). A P done right would give the competitive numbers needed to really put palm on the burner come that question in the next earnings call (not like how they skipped around it this time)..

Personally, I was impressed. I knew that Palm was doing well. But they are doing better than I expected. I still think they get one quarter of eww before being totally good again, but I also think this call made them even more ripe for acquisition. The year will be interesting. It is for me, and I only have a Treo to speak for it ;-)

mobileministrymagazine.com
antoinerjwright.com

RE: Breakdown of the new Treos
cervezas @ 3/24/2006 10:46:29 AM # Q
Here's an "outside the box" possibility for the Treo breakdown: one of them may be running Linux. Not ALP, of course, but a distro developed by Palm in partnership with the likes of Wind River.

Consider the hints that make this somewhat credible:

1) The same source that first leaked in Nov '04 (correctly as we now know) that Palm was exploring Windows Mobile also reported that they were exploring partnerships to develop a "a tailored version of the Linux OS to run on its devices." That was almost a year and a half ago. Plenty of time to have their own Linux version of Palm OS ready.

2) Someone close to Wind River dropped hints last year that they were partnering with Palm on a Linux phone. They called it a "feature phone" as I recall, but your guess is as good as mine what that might mean. Maybe it's Lowrider.

3) Moving outside the realm of rumor now, Palm has been hiring Linux engineers for at least the last 7 months that I've been watching. They had postings for 17 Linux engineers back in September and there are 16 engineering positions on their site right now that mention Linux experience as a qualification. Most don't have Linux in the description of duties as they did back in September, but one of them seems pretty revealing:

Linux Engineer- Mobile Handset, Embedded Systems

Overview:

As a senior software engineer, you will play a key role in the architecture, design and implementation of enabling technologies for a new generation of Palm devices. Working as part of the overall system team, you will work on the underlying support for various software applications.

Duties/Responsibilities:

You will be responsible for the design and development of components of a new software platform.

Nothing conclusive that this is about developing a Linux platform for the Treo, although the job title does say "mobile handset." Whatever you make of this it sure doesn't sound like Palm is sitting on their hands between now and whenever ALP is released.

If I were Palm, I'd be working on something like ALP without MAX right now. Just the Garnet emulator and a few native GTK+ apps that could provide multitasking in the area where it's most critical: phone, browser, email, messaging, media playback. Even if you couldn't run multiple Palm OS apps concurrently, you could at least run those native apps in their own processes while a single other Palm OS app was running in the emulator. This would solve the problem of getting the Palm OS to comply with the UMTS standard so it can run on GSM 3G networks. It would position them well from an engineering standpoint if they decide to migrate to ALP when that time comes. And it would give them experience and credibility if they opted instead for a more full-fledged Linux platform of their own in the future.

ALP may not be "Palm OS", but that doesn't mean the next version of Palm OS couldn't still be Linux.



David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
cervezas @ 3/24/2006 11:24:54 AM # Q
Oh yeah. Links.

The first leak that Palm was exploring Linux: http://tinyurl.com/n7zpr

The rumor about Palm partnering on a Linux phone with Wind River: http://www.linuxdevices.com/news/NS5058819583.html (some of this obviously turned out to be false, like the timeline).

And Palm's job postings: http://www.palm.com/us/company/jobs/

It's all a matter pure speculation what this means, of course, but it sure seems like something Linuxy is cooking in the basement at Palm Inc.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
PenguinPowered @ 3/24/2006 5:53:33 PM # Q
It has been my experience that linuxdevices is, um, optimistic in its interpretation of the rumors it hears.

Palm would be remiss if they did not talk to Wind River, but there's a long distance between "talking to" and "developing a phone with".



May You Live in Interesting Times

RE: Breakdown of the new Treos
cervezas @ 3/24/2006 6:16:45 PM # Q
PenguinPowered wrote:
It has been my experience that linuxdevices is, um, optimistic in its interpretation of the rumors it hears.

Agreed. Though you have to ask how much interpretation (ruling out pure fabrication) could result in "Our anonymous sources confirmed that Palm's first Linux-based handsets will use a version of Linux developed by Wind River" meaning something much different from what those sources actually said. If the information is wrong, I'd say it's more likely the fault of the sources than LinuxDevices.

Palm would be remiss if they did not talk to Wind River, but there's a long distance between "talking to" and "developing a phone with".

Sure. So how long is the distance between hiring Linux handset engineers "for the design and development of components of a new software platform" and designing and building a new software platform? I'd say that closes the gap considerably.

It may actually have nothing to do with Treos (could be Palm's "Third Business" ramping up, for example) but from where I stand it sure looks like Palm has moved some kind of Linux project out of the research phase and into development.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
KultiVator @ 3/24/2006 8:07:53 PM # Q
I'm with Cervezas on this story. So many little clues point towards Palm looking to sieze control of their own destiny again and control their prime OS.

WinMob may turn out to be either a nice side-line, or a short-lived survivalist stop-gap - but I'm convinced that Palm want their own in-house platform again and will settle on a policy that focuses more on shifting hardware than it does on licensing software to third parties (although I don't expect Palm to give up licensing their software altogether).

Other Palm job descriptions in recent months have indicated a big rewrite of the basic PIM applications is underway. Can't think why this would be happening unless some fundamental platform changes are 'in the oven'.


KultiVator

RE: Breakdown of the new Treos
PenguinPowered @ 3/25/2006 3:49:43 AM # Q
Sure. So how long is the distance between hiring Linux handset engineers "for the design and development of components of a new software platform" and designing and building a new software platform? I'd say that closes the gap considerably.

I think it's safe to say that Palm is involved in some Linux project, or they wouldn't be hiring Linux people. But that doesn't say who they're working with. Maybe they picked Wind River, although why pick the new kid on the block? Maybe they're working with MontaVista, like PalmSource was? Maybe they're working with Access on an ALP phone?

A company that is buying a Linux solution from a third party is going to hire at least a few Linux engineers to do the integration and so forth in house. Palm is know to have modified large junks of PalmOS even when they were buying it from PalmSource, so buying a solution and then tweaking it would fit that job description.

If I were to speculate, I'd guess Palm is exploring a Linux phone (something they've been rumored to be doing for years, anyway,) but I wouldn't go so far as to guess that they're working with Wind River.

Players in this space these days are: (in no particular order)

MontaVista - has shipped product in the telephony space
Access - has (through CMS) shipped product in the telephony space
Wind River - hasn't yet shipped product in the telephony space, reasonably new to Linux
Handheld.org - well, they exist, anyway.
Trolltech - (see handheld.org)
Nokia - maemo doesn't do telephony.

If I were Palm and interested in adding Linux, I'd be working with my hardware vendors and one or more of MontaVista and Access. I don't think I'd work with any of the others, unless my hardware vendor was working with them and trying hard to convince me to.


May You Live in Interesting Times

RE: Breakdown of the new Treos
AdamaDBrown @ 3/25/2006 5:00:01 AM # Q
If I were Palm, I would be very leery of getting involved with another minor OS provider that could go belly up or get bought out. They need to be looking long-term, and I'd guess that that means either controlling the OS or dealing with someone whose presence is guaranteed. I, personally, would choose the former. Take the simplest available robust mobile OS, and top it with a semi-classic Palm UI and application compatibility layer. No need to get fancy right now.

RE: Breakdown of the new Treos
cervezas @ 3/25/2006 8:23:33 AM # Q
Perhaps I shouln't have mentioned the Wind River rumor in my first post, since I really don't know or care if that rumor turns out to be true. I was mainly trying to point out that where everyone assumes Palm will have to use Windows Mobile for future GSM phones if they are to meet the operator demand for 3G (UMTS) networking, there is a reasonable possibility that we'll see one or more Linux-core Palm OS phones instead.

From what I understand, whether or not they partner with Wind River for kernel-related stuff has nothing to do with the question of whether they're working with PalmSource and ACCESS, since we keep being told ALP can be BYOK (bring your own kernel). If they wanted to do something like what I suggested and build a phone with just a Garnet emulator and a few native Linux apps it would make seem to make a lot of sense to be working together with PalmSource on GHost.

One of the interesting things about the ALP design is that it seems to open the possibility of incremental adoption by current Palm OS licensees. What would keep ACCESS from licensing an early version of ALP to Palm without the MAX framework? Sounds like a good way to convince an important licensee who was otherwise nervous about your platform (particularly its timeline) to give you a try. Fix their Garnet problems first with UMTS support and improved multitasking. Then up-sell them on the whole enchilada when it's ready.

Again, we're completely in the realm of speculation now, but it's interesting to consider that ALP could be quite useful to Palm this year even without the MAX framework.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
Surur @ 3/25/2006 4:19:22 PM # Q

Just as a side question - many people (including me) said Palm went WM to get EVDO, but the Treo 700p obviously disproves that, and its been said the radio is quite separate from the OS, meaning any radio can be plugged in, and used just like any data connection.

Does Palm really need to go Linux for UMTS?

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: Breakdown of the new Treos
PenguinPowered @ 3/25/2006 5:20:45 PM # Q
UMTS makes multitasking demands on the OS that it would surprise me if Garnet are up to. While the radio is a separate device, the overhead of concurrent data and voice is pretty demanding.

Does Palm need to go to Linux? No, they could probably do it other ways. Could they do it in Garnet? Only with a significant rework of Garnet, it seems to me.

May You Live in Interesting Times

RE: Breakdown of the new Treos
PenguinPowered @ 3/25/2006 5:27:09 PM # Q
What would keep ACCESS from licensing an early version of ALP to Palm without the MAX framework?

That would amount to POSE on top of a Linux kernel: all the disadvantages of the PalmOS single tasking model, all the overhead of a multitasking OS.

They could do that, but it would be slow.



May You Live in Interesting Times

RE: Breakdown of the new Treos
Surur @ 3/25/2006 5:27:25 PM # Q

This is obviously a technical question beyond my ken, but how is UMTS different from EVDO in this? Is it the requirement from video and voice at the same time?

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: Breakdown of the new Treos
hkklife @ 3/25/2006 6:03:25 PM # Q
Surur;

Don't hold me to this as I am not an expert (not a GSM user) but I think EVDO, in its current/initial rev, cannot receive a call while in the midst of a data session. Future revisions to the EVDO spec will permit this, much in the way the 700w user can choose to send a SMS to someone if they don't want to take the call.

UMTS can handle simultaneous data & voice transmission and/or have an incoming call come in then resume the data session when the call is finished--if not actually continuing the data seesion in the background while the voice call is in progress.

Anyone else in the know feel free to point out any additional diferences or correct me on this.


It is widely said that EDGE/EVDO is much less taxing on the software/OS side of things as well as easier to implement than their GTSM equivalents. That might explain Palm's immediate attraction to Sprint & Verizon for a final POS Treo or two vs. throwing in the towel on the GSM POS market after the 650 or perhaps the "slow" budget oriented Lowrider.


Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX

RE: Breakdown of the new Treos
cervezas @ 3/25/2006 6:07:41 PM # Q
PenguinPowered wrote:
all the disadvantages of the PalmOS single tasking model, all the overhead of a multitasking OS.

I agree that this would be kind of pointless if the platform ran all its apps in the Garnet emulator. But consider that Garnet's single-tasking could be considerably relieved if only a handful of critical apps could run natively in their own processes, separate from the emulator. The ones that would make the difference would be the ones that relate to the fundamental wireless data and telephony features of the Treo. Again: phone, browser, email client, messaging client, streaming media player. If those apps were written against GTK over DirectFB (no X Windows: they don't care about 3rd party contributions from the Linux side here) it seems to me Palm might have a pretty good stop-gap platform that they could call Palm OS 7. Without running an X server, I suspect the performance could be improved considerably--probably within spitting distance of Windows Mobile at least.

Assuming that Palm didn't clinch the decision to move their Linux experiments out of skunkworks and into production plans until they lost the bid for PalmSource in the Fall of last year (that's when I first noticed all the Linux engineer job postings) don't you think they could have a platform like this ready to ship in some handsets by the end of this year?


David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
PenguinPowered @ 3/25/2006 6:42:14 PM # Q
Assuming that Palm didn't clinch the decision to move their Linux experiments out of skunkworks and into production plans until they lost the bid for PalmSource in the Fall of last year don't you think they could have a platform like this ready to ship in some handsets by the end of this year?

If the 'skunkworks' (which I don't think exists) had a working prototype that needed debugging and productizing, end of this year, for sure.

If the 'skunkworks' had at the end of last year something more or less the equivalent of the PSRC developer conference demo but running on Palm hardware, middle of next year more likely, and I'd bet on the end of next year to be safe.



May You Live in Interesting Times

RE: Breakdown of the new Treos
cervezas @ 3/25/2006 6:42:37 PM # Q
hkklife wrote:
UMTS can handle simultaneous data & voice transmission and/or have an incoming call come in then resume the data session when the call is finished--if not actually continuing the data seesion in the background while the voice call is in progress.

I'm not an expert in this either but it's my understanding that the UMTS standard requires *concurrent* data and voice, and that that is where Garnet gets into trouble. I don't think putting the data transmission on hold during a call meets the standard, even if the transmission is resumed gracefully after the call ends.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
cervezas @ 3/25/2006 6:53:27 PM # Q
PenguinPowered wrote:
the 'skunkworks' (which I don't think exists)...

Ask Ben Combee some time what he does for Palm and you might change your mind.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
Surur @ 3/25/2006 7:04:41 PM # Q

I just tested this with my 3G UMTS phone (Sony Ericsson v800) and to my amazement I would actually download data to my PDA over bluetooth (browsing the web) while in a voice conversation. I did not actually know it was possible, as the phone itself does not provide any multi-tasking interface.

Its a bit telling however that even a dumb phone can do voice and transfer data at the same time on 3 G. I'm sure garnet could do this too.

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: Breakdown of the new Treos
cervezas @ 3/25/2006 7:40:51 PM # Q
Read the manual on your v800, Surer. It's got a multitasking OS, which is why you're able to do this. With existing Palm OS Treos (and I expect this to be true of the 700p) data calls are put on hold during a phone conversation.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog
RE: Breakdown of the new Treos
Surur @ 3/25/2006 7:53:26 PM # Q

I have no idea what the OS is. Its certainly not Symbian, and as Ive said, there is no interface to expose multi-tasking. When I'm in a phone call there is no way to start any app or go to the normal menu. I suspect its as single tasking as garnet.

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: Breakdown of the new Treos
PenguinPowered @ 3/25/2006 11:35:42 PM # Q
From your description, it is an OS with a multitasking kernel, at least with respect to radio streams, that only presents a single user task at a time.

This differs from Garnet because Garnet presented only a single user task and only has a single tasking kernel.



May You Live in Interesting Times

RE: Breakdown of the new Treos
PenguinPowered @ 3/26/2006 12:07:01 AM # Q
FWIW, Sony Ericsson's web site suggests that the v800 is a Symbian phone. At any rate, it is definitely a multitasking OS, since it runs Java.

See http://developer.sonyericsson.com/getDocument.do?docID=75136 for an excellent but detailed explanation of UMTS and when it does or does not require concurrency, by the way.


May You Live in Interesting Times

RE: Breakdown of the new Treos
AdamaDBrown @ 3/26/2006 1:06:33 AM # Q
I have to wonder if UMTS is really going to be an issue for Palm. The UMTS network in the US isn't nearly as extensive as EVDO is, and UMTS in Europe isn't compatible with our frequencies. All things considered, they may simply decide to leave UMTS out of the GSM 700x models.

RE: Breakdown of the new Treos
cervezas @ 3/26/2006 10:20:43 AM # Q
PenguinPowered wrote:
Garnet presented only a single user task and only has a single tasking kernel.

You know your Linux, Marty, but you don't know your Palm OS :-P

The problem is not the kernel. The Garnet kernel and the Kadak kernel that preceded it on the 68k processor are both preemptive multitasking kernels. The problem is the UI framework. All Palm applications run as subroutines of a single UIAS (User Interface Application Shell). It's possible for applications to get the system to spawn new non-UI tasks--always has been. The Sync application has always does this, for example, and so do some audio playback applications that use a new capability for background audio playback introduced with OS 5. What's not possible in Garnet's architecture is for two applications (like phone and browser) to run at the same time and manage application concerns like UI interaction and memory.

It seems to me that Palm might yet be able to hack their way around this if they create a second framework parallel with Garnet. At the most minimal level it might exist only to run the phone application and provide UMTS compliance. As a very short-term solution to the problem that most GSM operators outside the US aren't interested in new handsets that aren't 3G, this might be ok.

But even if they do this, beyond the next year or two Palm is certainly going to be looking for a new platform--either one they'll license from ACCESS/PalmSource, or one they create for themselves that they can call "Palm OS". All other advantages aside, just from a marketing standpoint the latter option may look appealing. The ACCESS acquisition may have pushed them in that direction.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
cervezas @ 3/26/2006 10:45:01 AM # Q
AdamaDBrown wrote:
All things considered, they may simply decide to leave UMTS out of the GSM 700x models.

I doubt that. I think Palm is serious about penetrating the global market, not just selling to the US. They need some way to comply with UMTS and it's going to involve some kind of framework development, either on top of the existing Garnet kernel (temporarily, as I just described) or on top of a new kernel like Linux.

David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog

RE: Breakdown of the new Treos
stonemirror @ 3/26/2006 12:08:27 PM # Q
David's correct here: the Garnet kernel is most certainly capable of multitasking, as was KADAK's AMX kernel, used in Palm OS prior to OS 5.

RE: Breakdown of the new Treos
Surur @ 3/26/2006 12:35:32 PM # Q

David, is POS UMTS capable?

Surur

They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...

RE: Breakdown of the new Treos
PenguinPowered @ 3/26/2006 4:10:30 PM # Q
You know your Linux, Marty, but you don't know your Palm OS :-P

Too true. thanks for the clarification.


May You Live in Interesting Times

Reply to this comment
Start a New Comment Thread Top View Full Comment Thread
Achtung! Only the first 50 comments are displayed within the article.
    Click here for the full story discussion page...

Account

Register Register | Login Log in
user:
pass:

Latest Comments

  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST((SELECT/**/CASE/**/IS_SRVROLEMEM
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST((SELECT/**/CASE/**/IS_SRVROLEMEM
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST((SELECT/**/CASE/**/IS_SRVROLEMEM
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST(db_name()/**/AS/**/NVARCHAR(4000
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST(db_name()/**/AS/**/NVARCHAR(4000
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST(db_name()/**/AS/**/NVARCHAR(4000
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST(db_name()/**/AS/**/NVARCHAR(4000
  • My comments --1' OR UNICODE(SUBSTRING((SELECT/**/ISNULL(CAST(db_name()/**/AS/**/NVARCHAR(4000