2005 the Palm OS Year in Review
In our annual year in review article, PalmInfocenter takes a look at all of the major events that took place in the Palm OS world in 2005. It was certainly an exciting year full of surprise announcements and major transitions. Treo smartphones really began to take off, a lot of innovative devices were released and many surprises and new directions within the Palm economy took shape. Read on for the full recap.
January
In the beginning of 2005 Palm was still known as palmOne. The company announced a number of changes in January. The first sign of change came when the company's CEO resigned. Todd Bradley later moved on to HP to head their Handheld and PC Unit. Ed Colligan stepped up as the new CEO. The Treo 650 became more widely available and enjoyed strong sales and wide praise in the media.
A few new Palm OS devices debuted early in the year. The often delayed Fossil FX2008 wristwatch PDA came to market. Garmin released the iQue 3600a, a GPS handheld designed for aviation, and the Qool QDA-700 became available in some European and Asian markets.
February
In February PalmSource formally completed their acquisition of ChinaMobileSoft. PalmSource acquired the company to leverage the work CMS has done with Linux to provide a new version of Palm OS based on Linux.
In GSPDA news, the company brought the M68 smartphone to Europe. The company also announced they would bring out a Palm OS Cobalt smartphone by the end of 2005. A claim which was again restated during PalmSource's developer conference in May, however to date no manufacturer has announced a Cobalt device and none will likely ever.
Eight months after exiting the US and European PDA markets, Sony put the rumors to rest and officially announced there would be no more Clie handhelds in Japan or anywhere else.
March
In March, industrial handheld maker Symbol announced they had renewed their Palm OS license, however no new devices have been released by the company in a number of years.
Also in the beginning of the March a pair of mystery Treo's surfaced. These images were the first sign of the impending Apocalypse Windows Mobile Treo. palmOne also announced another positive quarter and the opening of a new Smartphone engineering center in Dublin. PalmInfocenter also posted its Treo 650 review.
April
In April palmOne released the Tungsten E2 handheld. This update to one of the company's top selling models included a brighter color screen, flash memory and Bluetooth. Meanwhile rumors and speculation about the upcoming LifeDrive were rampant.
In a strange twist of fate, a company debuted a Palm OS emulator for Microsoft Pocket PC's. Dubbed StyleTap, the new product allows most applications originally written for Palm OS handhelds to run on Windows Mobile-based PDAs.
May
In May the palmOne LifeDrive officially debuted. The LifeDrive was the first Palm OS PDA with a built in hard drive. It featured a 4GB hard-drive for storage, a 320x480 high-resolution color screen, and both Wi-Fi and Bluetooth. Read the review here.
Just a day before the PalmSource developer conference got underway, PalmSource CEO David Nagel resigned his position. Senior VP Patrick McVeigh stepped up as interim CEO, while the company has yet to name a permanent replacement.
The PalmSource DevCon was a very well attended event as the company showed its first glimpses of Palm OS for Linux. Even a Palm OS Cobalt phone was being shown.
palmOne announced at the conference that they had renewed their Palm OS license through 2009. But the big news was that the company worked out a licensing deal with PalmSource to reacquire the rights to use the Palm brand and name.
Summer
The summer months saw the news slow down, with a number of software devices updates here and there. Rumors of a new Tungsten model first showed up.
In PalmSource news, the company announced a slight profit from the Palm trademark sale and also announced layoffs, cutting 16% of its staff. The company also announced one of its first new licensees in some time, as LG Electronics signed an agreement to develop Palm OS based smartphones. PalmSource also revealed that they were shifting all engineering efforts on the future Linux based versions of the Palm OS. The company also put a timetable on its release, stating the version for low end and feature phones is planned to be completed by summer 2006 and a high end version for smartphones is expected to be ready in the second half of 2006.
In July the big news was that palmOne was now officially back to Palm Inc. Palm unveiled their new orange logo and promised new Palm products in the fall.
The summer also saw Tapwave shutting down. The company discontinued their Zodiac handheld business failing to ever grab a significant slice of the portable gaming market amid strong competition from Nintendo and Sony.
Jeff Hawkins made some news when he began to drop hints about his next big invention in development. Hawkins began talking about Palm's secret third business involving the next major generation of mobile computing.
August also saw a big return to the Windows Mobile Treo rumors. A video of it in action was posted, Palm's CFO told the media that "CIOs don't get fired for using Microsoft products" and we learned that HTC would be manufacturing the device. What was years ago totally unthinkable started to appear almost certain.
In Palm OS device news, the GSPDA Xplore M98 smartphone began shipping. Intel also announced their next generation XScale chips code-named "Monahans". A pre-production Monahans chip running at 1.25 GHz was demonstrated palying back a HD quality video on a PDA.
September
In September the latest bomb was dropped in the ongoing PalmSource saga. Japanese mobile software company ACCESS announced they would acquire PalmSource in a deal valued at $324 million dollars. ACCESS pledged to make PalmSource a wholly owned subsidiary and would help the company continue work on Palm OS for Linux. We later learned that Palm and Motorola had both put in bids for PalmSource.
Then in late September it finally happened. Confirming months and months of rumors, speculation and flame wars, Palm and Microsoft jointly announced at a press conference with Ed Colligan and Bill Gates that Palm has licensed the Microsoft Windows Mobile operating system for an expanded line of Treo smartphones. The Treo 700w was to debut on the Verizon network in early 2006. The future would never be the same again.
October
October got off to a bang with Palm releasing two new Palm OS PDAs. The Z22 is Palm's first sub $100 entry level color handheld in a small and light package designed to go after the paper organizer market. The Palm TX handheld was released to positive reviews, featuring a 320x480 pixel screen and both Wi-Fi and Bluetooth.
Later that month, PiTech announced they had developed a 3 megapixel Palm OS smartphone. The GSM smartphone features an auto-focus camera, a quad band wireless radio with EDGE data support and a QVGA display and Bluetooth. It is expected to launch in Europe and Asia in the second quarter of 2006.
November
November opened up with Palm denying a rumor that they were working on a Symbian OS Treo. Palm also announced that they continue to open new retail stores including five new ones in California. Palm CEO Ed Colligan also released a letter reaffirming Palms commitment to the future of the Palm OS amid growing speculation that the company was moving away from its long time operating system.
We also learned for the first time that Palm has a number of new Treos in the works. An investment company released a report detailing that Palm was working on new models including a new high end design code-named Hollywood and a possible entry level Treo model code-named Lowrider.
Novemember was also the month that saw ACCESS completing their acquisition of PalmSource. PalmSource is now a wholly owned subsidiary of ACCESS and is no longer a publicly traded company.
PalmInfocenter also debuted a new Palm OS software store this month. The improved store brings big discounts on Palm applications and also features a mobile store where users can download and buy software on their devices wirelessly.
December
In the last month of 2005 Palm announced very profitable quarter, its eigth consecutive quarter of year over year double digit revenue growth. Strong Treo 650 and handheld sales helped to boost the company's bottom line. Palm also officially revealed that they will announce three new smartphones in 2006.
2005 was definitely a major milestone year in the Palm OS timeline. Treo smartphones really began to take off, a lot of innovative devices were released and many surprises and transitions within the Palm economy took shape. Thanks to everyone for making 2005 a year to remember here on PIC and here's to an even better 2006.
Article Comments
(91 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.
But Wait, There's More!
We'll just have to wait for Skippy to finally find out and blab.
May You Live in Interesting Times
Don't be a hater, Marty.
It's time for you to let go and "move on", Marty. And if you ever need to vent, or need a shounder to cry on, remember: nobody here cares about you.
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: But Wait, There's More!
I never said that what i couldn't say was bad news.
only that you don't know it yet.
May You Live in Interesting Times
RE: But Wait, There's More!
Pat Horne; www.churchoflivingfaith.com
RE: But Wait, There's More!
only that you don't know it yet.
Stay off the drugs, Marty.
It's not a secret anymore. Eleven devices running Cobalt will be released by the end of 2005. Most people know that already.
Actually, they were referring to the end of 2005 in the JULIAN CALENDAR. Technically speaking, this means they still have until January 13, 2006 (Gregorian calendar date) before you can call them a bunch of sleazy, lying dirtbags. I have faith that we'll see at least one Cobalt device released before this weekend.
Pat: Why have you forsaken Palm? Sinner!
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: But Wait, There's More!
It might not be the "mythical color HandEra", but I'm liking my TX anyway.
RE: But Wait, There's More!
I'm standing strong until the end. Just waiting to see if Spring will bring anything worthy to upgrade my Pilot 1000 to.
Pat Horne; www.churchoflivingfaith.com
Et tu, Pat?
I hear you can trade Palm your Pilot 1000 and they'll send you three LifeDrives in return. I'm waiting for them to sweeten the deal by only sending me two LifeDrives in return for my old Pilot 1000...
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
Too late?
Methinks it's time for an "intervention" on Pat's behalf. We CAN and MUST save this lost lamb.
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
Ryan, you left out the BIGGEST PalmSource story of all...
;-O
Do you really not know what's happened, or are you just trying to be nice? Nudge, nudge. Wink, wink.
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Ryan, you left out the BIGGEST PalmSource story of all..
Do you dream of electric sheep, Skippy?
May You Live in Interesting Times
RE: Ryan, you left out the BIGGEST PalmSource story of all...
Good summary outling '05. I had forgotten about the near-vapor releases of some of the Europe & Asia only smartphones.
Now I'd REALLY like to see the Voice's '06 predictions article and the long-awaited VZ90 review/overview. I myself am going to contribute some material to PIC (per the sentiment of the past few days to make our last stand here) with a couple of game reviews in the works at the moment. I also am working on a review of the Palm GPS Navigator kit and, eventualy, a T5/TX case review. Initial impressions are quite positive of the GPS kit even if it basiclaly is just a repackaging of other firms' components in a Palm-branded box.
Oh a "tell all" article from an ex-PalmSourcer would be lovely gossip but that's probably too volatile for "official" PIC content...not to mention a sure fire bet to get someone into legal hot water. I guess we'll just have to wait for the other shoe to drop through the grapevine, huh?
I have to say that the dialogue here on PIC over the past two weeks has been some of the most entertaining and informative/insightful speculation we've had in a long time. I'm actually enjoying the "stir" the 700w has created even if I have zero interest in the actual product itself.
Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX
RE: Ryan, you left out the BIGGEST PalmSource story of all...
Sorry - no can do. That would be illegal! ;-O
and the long-awaited VZ90 review/overview.
VZ90. Mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm. VZ90 + TCPMP + Wi-Fi connection = the ultimate combination media player/Internet tablet/PDA.
Review will come. Soon. I'm also planning to finish up the reviews of a few apps.
TVoR
;-O
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
something decent in news every month
"Do you know the difference between an error and a mistake? Anyone can make an error, but that error doesn't become a mistake until you refuse to correct it."
-Grand Admiral Thrawn
the secret to enjoying your job is to have a hobby that's even worse
My PDAs: Visor --> Visor Neo (blue) --> Zire 71 --> Tungsten T3 (with 4 of 6 screws still remaining) ~?~> zodiac 2?
Cobalt is here at last!!
http://palmaddict.typepad.com/palmaddicts/2006/01/cobalt_10.html
-------------------------------
Editor, http://Pocketfactory.com
Contributing Editor, http://digitalmediathoughts.com
AWESOME!!! I knew the Cobalt upgrade would be released!
We now we have it available for FREE and running. Suckers!
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is here at last!!
The real Cobalt screensaver has never repeating pattern of vapor which begins at the bottom and randomly dissapates as it ascends toward the top. Fools!
Pat Horne; www.churchoflivingfaith.com
"A Cobalt in every pot"
And why are you using the words "Cobalt" and "Vapor" in the same sentence? I'm tempted to tell Dianne Hackborn on you and have her administer a biotchslapping like you'll NEVER believe!
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
Cobalt is REAL.
There's no place like home.
There's no place like home.
There's no place like home.
There's no place like home.
There's no place like home.
There's no place like home.
There's no place like home.
There's no place like home.
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is face down in the grave
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is here at last!!
But Voice, how can an OS that never technically "lived" be face down in its grave? Not to be Dr. O style pedantic (HAH!) here but isn't it more....stillborn than "dead"?
I wonder when Access is going to remove EVERY trace/hint of Cobalt from their website...just kind of magically "wash" it out of existence. I figure it'll be kinda/sorta how Palm tries to pretend that Graffiti 1 never existed.
Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX
RE: Cobalt is here at last!!
Er, wait, no, that was Frodo.
Anyway, PalmOS on Linux is Cobalt, but running on a Linux kernel. Think of it as Robert De Niro with Peter Boyle's heart.
Where's Boris Karloff when you need him?
May You Live in Interesting Times
You guys are just being "pedantical"
Ummmmm... wrong. Cobalt lived in that slow, buggy piece of **** Oswin smartphone for two glorious days at PalmSource's DevCon last year before being (mercifully) put out of its misery and getting flushed down the toilet (along with any remaining shreds of PalmSource's credibility).
Anyway, PalmOS on Linux is Cobalt, but running on a Linux kernel. Think of it as Robert De Niro with Peter Boyle's heart.
Where's Boris Karloff when you need him?
PalmLinux = FrankenPalmOS™:
- Linux kernel
- PACE from PalmOS 5 for PalmOS 1 - 4 compatibility
- Chunky bits of Perfect Code from the Holy Be Engineers (HoBeEn)
- Sweet and sour Linux sauce from caged China MobileSoft codemonkeys
- Touched by the hands of Dianne Hackborn *GASP!*
- Vaporware delay development timeline from PalmOS 6 (Cobalt)
- Rootkit by (disgraced) Marty Fouts
I can't wait to hear the excuses start flying over why PalmLinux won't show up (groundhog-style) in 2006.
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is here at last!!
-- http://sunsolve.sun.com/handbook_pub/Systems/Cobalt/Cobalt.html
I once sent a note off to PalmSource asking about the conflict of THEIR "Cobalt" OS with Sun's...they never replied...giggle.
RE: Cobalt is here at last!!
Thanks for the clarification. Did anyone here actually see/use in PERSON at DevCon that slow, buggy Oswin phone before PalmSource pedanticized it out of its misery?
I thought FrankenGarnet was bad enough...the smorgasbord of code bits outlined in the post above is enough to make my head hurt...and I'm not a developer//programmer/codemonkey!
Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX
RE: Cobalt is here at last!!
Did anyone here actually see/use in PERSON at DevCon that slow, buggy Oswin phone
I did, of course. I only fiddled with the built-in PIM apps for a few minutes so I can't say I did anything that really put it through its paces. But, FWIW, in the 15 minutes I used it I didn't find it slow or buggy. I have heard from other developers who bought them that performance wasn't great, but it was difficult to tell whether that was inherent to the OS or due to beta drivers or what have you. I never heard reports about it being "buggy" but I didn't exactly take a poll.
From what little I know about how Cobalt was written I would expect some performance penalty running existing Palm apps as compared to Garnet. Other things being equal (hopefully they won't be) I would expect the performance penalty when you put Linux under Cobalt's hood to be increased due to the less efficient context switching that's been discussed elsewhere. Of course, with Cobalt developers have a reasonable opportunity to write native ARM code that they haven't yet had--almost all Palm apps to date currently run as emulated 68k code. Taking advantage of that opportunity could resolve all those performance issues. In short, we really don't know right now what Cobalt or Cobalt-Linux would be like in a real production device.
David Beers
Pikesoft Mobile Computing
Software Everywhere blog
www.pikesoft.com/blog
RE: Cobalt is here at last!!
Garnet = OS 4 + some new APIs BUT rewritten in ARM, with the interface exposed as 68K. Thus, API intensive code is fast. Non-API code is slower because all of the 68k code have to have instruction set translated to ARM.
So Garnet/PACE is the efficient way to be backwards compatible.
Cobolt = PACE + Cobolt OS framework & UI + purely native Cobolt API (exposed 'natively' this time) + Cobolt Kernal
PalmLinux = Cobolt - Cobolt Kernal + Linux Kernal WITH modifications to OS & APIs to accommadate the different threading /memory management model of the linux Kernal.
Palm Linux is not a 'messy hack' of Cobolt. The highest level functionality of the OS exists fairly discreet from the Kernal (with the exceptions mentioned for accomadating the different threading/multitasking model). It's not as if Cobolt devices are already out there to complicate the compatibility situation for PalmLinux by forcing them to pull off some sort of crap work around to deal with the new model(the way MS messed up with the no-close-button thing on early the PPC OS and then came up with the pathetic 'smart-minimize' solution so that people never knew if a given app was actually shutting down or just minimizing when you tap the stupid 'X'. God thats dumb).
So for this reason, PS should make sure that nobody uses Cobolt - THAT will actually mess up things more than it will hep at this point.
PalmLinux is no more of a mess than MacOS X is. The Mac has proprietarey Mac UI / Framework / APIs on top of a Unix kernal AND with a compatibilty layer for classic Mac stuff.
Same deal.
Oswin Cobalt FecesFone
Sounds like Beersy's Apologist Hat HAS CUT OFF CIRCULATION TO HIS BRAIN.
Don't get fooled by his B.S., hkklife. Cobalt ran like a drunken, crippled PIG on the Oswin smartphone. Of course Beersy will have you believe it was just a minor revision away from rivalling a Pentagon supercomputer...
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is here at last!!
-------------------------------
Editor, http://Pocketfactory.com
Contributing Editor, http://digitalmediathoughts.com
RE: Cobalt is here at last!!
Thank you. I haven't laughed so hard in days.
PalmLinux is no more of a mess than MacOS X is.
Having seen the insides of both, I would have to describe the above sentence as "damning with faint praise."
Mach was a conceptual hack job on Accent before DARPA suggested to CMU that "Unix compatibility would be nice", after which it morphed into an actual hack job on BSD. Then it was whacked on my NeXT, and finally, Apple, until it became the gordian knot of spaghetti code that it is today.
Knowing that, I can safely say that you are correct, PalmLinux is no more of a mess than that.
(Never inquire too closely into the making of sausages or operatig systems.)
May You Live in Interesting Times
Some people need to buy a vowel
Garnet = OS 4 + some new APIs BUT rewritten in ARM, with the interface exposed as 68K. Thus, API intensive code is fast. Non-API code is slower because all of the 68k code have to have instruction set translated to ARM.
So Garnet/PACE is the efficient way to be backwards compatible.
Cobolt = PACE + Cobolt OS framework & UI + purely native Cobolt API (exposed 'natively' this time) + Cobolt Kernal
PalmLinux = Cobolt - Cobolt Kernal + Linux Kernal WITH modifications to OS & APIs to accommadate the different threading /memory management model of the linux Kernal.
Palm Linux is not a 'messy hack' of Cobolt. The highest level functionality of the OS exists fairly discreet from the Kernal (with the exceptions mentioned for accomadating the different threading/multitasking model). It's not as if Cobolt devices are already out there to complicate the compatibility situation for PalmLinux by forcing them to pull off some sort of crap work around to deal with the new model(the way MS messed up with the no-close-button thing on early the PPC OS and then came up with the pathetic 'smart-minimize' solution so that people never knew if a given app was actually shutting down or just minimizing when you tap the stupid 'X'. God thats dumb).
So for this reason, PS should make sure that nobody uses Cobolt - THAT will actually mess up things more than it will hep at this point.
PalmLinux is no more of a mess than MacOS X is. The Mac has proprietarey Mac UI / Framework / APIs on top of a Unix kernal AND with a compatibilty layer for classic Mac stuff.
Same deal.
Bubba, you need to get a clue.
- Cobalt was already a "messy hack" that would have needed another two YEARS of development (and the accompanying increases in mobile processor speeds expected from the trickle down from Moore's Law) before it could be considered a viable OS for use by consumers.
- PalmLinux is by definition a "messy hack" of Cobalt, as PalmSource was attempting to salvage whatever code they could from Cobalt and cobble together a trendier OS ASAP before the roof came down on them. They simply didn't have time to do a clean sheet design of PalmLinux using some fresh (non-PalmSource) engineers experienced in mobile Linux. So what happens when you do a "messy hack" of a "messy hack"? In you world, the "messy hack[s]" cancel each other out, resulting in Perfect Code, right? Well in the Real World, you actually get Massive Problems that have people up late at night trying to come up with fixes before time runs out and nobody gives a damn anymore about the OS (BeOS-style). If PalmSource releases a STABLE, FUNCTIONAL PalmLinux by Summer 2006 like they claimed they would, I will be shocked but would be the first to say they are geniuses. Having seen the Keystone Kops running the show at PalmSource, this just ain't gonna happen. Sorry, Bubba. Maybe PalmLinux will ultimately be used to power eVilla 2 in 2008 instead?
- Mac OS X was a huge, brutal, "messy hack". The difference between PalmSource and Apple is that Apple managed to pull it off. But don't look too closely under the hood of Mac OS X - you might just throw up if you saw what a nasty, skanky OS your precious, virginal Mac OS really is.
- The word is Cobalt.
- You might want to S T F U before you make get sued for making some poor PalmSource codemonkey choke to death on bananas after reading your comments. Take off those rose colored glasses. Welcome to the Real World, Neo.
TVoR
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
RE: Cobalt is here at last!!
I would never ever argue otherwise. I am cynical, I run windows on my desktop - so i expect a mess from any major OS these days (I gave up on perfection when AMiga dies years ago i was forced to get over it and 'settle' for Windows). BUT it is relative.
My point is this: PLinux can only be better than the w-a-y overdue for replacement Garnet OS - a trully hacked beast at this point. My other point is - with it being 'no more of a mess' than Mac and others and with it being a big improvement over Garnet - then why do we need to hear reflexive b!tching from the annoying, flog-a-dead-horse-beyond-belief, usual suspects like TVOR - who hasn't looked 'under the hood'?
>>But don't look too closely under the hood of Mac OS X - you might just throw up if you saw what a nasty, skanky OS your precious, virginal Mac OS really is.
Check your 'precious, virginal' ego. I don't use Mac, don't own one, and don't like the OS one bit (though i have coded for it). Oh know! - spaghetti code in an OS?????? Are you going to defend your 'precious, virginal' Windows OS? Point: Mac is clearly 'good enough' and - relative to where Palm OS 5 is right now - represents a *huge* leap forward. PalmSource would be doing pretty well right now in the eyes of developers if they had already transitioned to such a 'mess' of an OS - seeing as it is much better to write code for (i.e. actually has an API to to write code against the new OS features vs Palm OS 68k API only + ARM hacking).
Anyways - thanks for yr reflexive non-insightful criticism, TVOR. As usual, NOTHING is added to the discussion. You are not educating me on anything - I live in the real world where i recognize that such a new OS replacing Garnet with pLinux cannot help but be a big improvement over what we have. Don't know what your expectations are - but it sounds pretty naive.
RE: Cobalt is D.O.A. Will PalmLinux show up dead as well?
Again you show how naďve you truly are, Bubba. PalmOS 5 is HERE, has been HERE for 4 years, has proven, widely available development tools, has known bugs + known workarounds, has a decent knowledge base, and ARM apps can be used if developers want to make the effort to do so.
PalmLinux, on the other hand is just VAPORWARE at this point. Whether or not it can function in the REAL WORLD (as PalmOS 5 has bravely been doing for the past 4 years) remains to be seen. In 2004, when PalmSource started to realize that there was a big possibility that Cobalt was not worth trying to save, they should have had the foresight to devote some resources to cleaning up the PalmOS 5 code as much as was realistically possible, since it would be their only shippig OS for 2 or 3 years if Cobalt was shelved. At this point, the smart money would take a known entity like PalmOS 5 over PalmLinux. If you don't believe me, just ask my old buddy, Martha... I mean Marty Fouts.
I don't use Mac, don't own one, and don't like the OS one bit (though i have coded for it). Oh know! - spaghetti code in an OS?????? Are you going to defend your 'precious, virginal' Windows OS?
I never said you used MacOS. And it was Marty - not me - who told you that MacOS is a "gordian knot of spaghetti code... today". And no, I have no need to defend Windows, since I made no claims regarding its elegance or lack thereof. Personally, I feel The Commodore Vic 20 and Coleco Adam are the only systems that matter. Windows XP and MacOS X bow before the might of the Vic 20 and the Adam...
You are not educating me on anything - I live in the real world where i recognize that such a new OS replacing Garnet with pLinux cannot help but be a big improvement over what we have. Don't know what your expectations are - but it sounds pretty naive.
Since I've already edumacated you, I'll leave you to try and figure out the extreme irony of your previous paragraph. Hint: a couple of years ago people were saying "replacing Garnet with Cobalt cannot help but be a big improvement over what we have".
I'll ignore your puerile insults, as I truly hope you'll finally LEARN something for a change.
TVoR [a.k.a. Gandhi]
(Taking the high road, as usual)
------------------------
Sony CLIE UX100: 128 MB real RAM, OLED screen. All the PDA anyone really ever wanted.
------------------------
The Palm eCONomy = Communism™
The Great Palm Swindle: http://www.palminfocenter.com/comment_view.asp?ID=7864#108038
NetFrontLinux - the next major cellphone OS?: http://www.palminfocenter.com/comment_view.asp?ID=8060#111823
Developer rants at PalmOS
Palm: Things That Annoy the Developers (Like Me)
Sun, 11 Dec 2005
Well, its that time of year again where things must be purged, the soul laid bare in preparation for a new year of technology silliness. (Okay,I had pleasant time at a Christmas party last night and a nice little lunch out today but I'm still in a mood to let fly the flames of war ;) While I really enjoy developing for mobile devices (and development in general) and have very much enjoyed being on the front edge of mobile technology
the last half decade or so, I've become a little weary of dealing with PalmOne's (now Palm again) and PalmSource's sillyness to tell you the truth - one can only have so much time wasted through their sheer stupidity before you start to wonder - when will it end? When will they stop being idiots? Their "historical issues" (mistakes, guffaws, unfortunate changes in the timeline) are forever in the developers faces. After writing this huge rant, I realize - it is a wonder people develop for this platform at all. I've been assisting other developers in forums for years, and I know how tough it is for them to get past all this stuff. This is why after a major
Shadow release theres usually a few weeks of quiet .. time to decompress and come to terms with the sheer amount of crap that had to be shovelled. This is why theres so much time between releases sometimes -- its getting harder to get the nerve up to "dive into the mountain of crap" :P
A comprehensive list of things that annoy me would be enormous as I routinely work around dozens of different Palm OS system bugs every day. But off the top of my head right now, heres what comes to mind .. call it the abridged and shortened summary of the greater litany :)
Ben, and my many friends at Palm and PalmSource -
this isn't personal or directed, instead just the frustration as I work arond more and more bugs that make me pull my hair out. Gotta vent and get it off my chest :) I enjoy the
platform, and I admire the work and dedication you guys have put in, but we still suffer :)
Edit, Jan 2006: Remember, I love this platform and work on it every day. I wrote this after working around some really aggravating OS bugs and answering hundreds of support emails about it. Don't be going and thinking I'm anti-Palm OS or anything ;)
Note: I could equally flame Microsft Pocket PC, WinCE, etc. I will,
but after Christmas :)
Backwards and Forwards Compatibility
Microsoft has been doing operating systems for awhile now so I don't
give them much slack when they've blown the game of increasing device
resolution (requiring new developer APIs to support higher resolution
units, instead of one neatly designed API - we all knew that devices
would get better over time, right?) but at least they followed
Palm's example of making low res apps just work on higher res devices.
Well and good. Now, with all Microsoft OS PDAs working the same,
compatibility is pretty high - but not so with Palm OS.
It literally drives me and most other developers absolutely batty dealing
with the volume of OS bugs that infect every Palm OS device - and
do so differently with each device incarnation.
- Bugs - the
developer doing something that works for years and then suddenly on new
devices it makes the unit crash or not work as expected, due to bad
("hurried"?) coding in Palm. Example: On most halfway recent OS 5
devices where you can copy a file out to SD card, and not back into
RAM afterwards. Thast not a feature, folks .. and its been around a year
or two at least now.. unfixed. In some cases it can be worked around but
it often involves risking making problems for syncing later, and such.
Not good. Please.. fix it? - Bad decisions due to 'history' - Great, the T|T1 introduced the
5-way d-pad like controller for better one handed control and an API
to support it more or less, and it could even be used in games. Alas,
the developer API for the 5-way was due to change itself several times
as new devices came out and with the usual lack of PalmSource controlling
licensees developing on its OS, it was only a matter of time when Handspring
came out with its entirely different API for a 5-way controller. Of course,
sure enough when Palm bought Handspring and returned them to the fold, they
adopted the Handspring model so new Tungsten units have the Handspring
control methods.. a little weird. Still, its so typical of Palm's bad
management as to allow dozens of incompatible control types - the jog
dials of Sony versus Handspring etc etc. We developers just love having
spaghetti code - dozens of exceptions in every piece of the code to handle
each device type, or requiring us to make massive compatibility layers
for our applications to operate halfway right. Naturally, Palm encourages
the users to blame the developers for problems or when an application doesn't
feel native to their device... when its Palm OS being changed on every once :/
This is getting better, due to the attrition among OS licensees, but I
know most of us still dont' well support some of the new APIs as we're
too mired down the the dozens of older APIs. ie: I need to better support
the Handspring-style 5-way with its floating blue box, but I still need
to support the 5-way on the older devices. - Lack of standards - Even when you do manage to fully support an API,
that API will itself change from device to device in how it is implemented
within the OS, and more often than not it was never designed in such a way
as to be flexible this way; consider the virtual grafitti, where there is such
a mess of changes that not even Palm can keep its own applications working
correctly with it. (The prefs panel system is notorious for having the
virtual grafitti come up and cover half its options as they forget to resize
properly.) - a "DIA" compliant application will work fine on one device,
and then flicker on another device with too many redraws, or fail to rotate
properly or receive system events in different orders or .. or .. thankfully
it "more or less" works, but it is in fact impossible at this time to make
a bug free virtual grafitti application. So many developers are overjoyed
that the Treo square-screens are taking over so they needn't worry about
virtual graffitti anymore. - Lack of forsight - The PIMs, the lack of more than a few categories.
The PIM (todo, datebook, etc) databases were never really designed to be
extensible. Sure, back in the OS 1 days they had no idea if they would
be successful at all, had limited time and budget and constrained devices
but I would argue this is no reason for shoddy development. For all the
years we've all be saddled with this lack of forsight, so Palm could not
extend their applications to suit users needs (grumbling users the
result, as they watched Pocket PC applications get more funky), and
developers having no way to extend those applications (leading to annoyed
developers making hacks.) Consider - Agendus and DateBk5 (and dozens of
other applications) adding garbled text to the Notes of todo and datebook
items so that icons can be assigned, or other special features added.
It works, but its not really pretty with that garbled text visible in other
applications and the desktop applications. Worse still was when PalmOne
decided they'd had enough and actually upgraded the PIMs - great, the users
will be happy .... - but they again forgot to supply a useful API and
instead just extended with databases without actually making them
extensible per se. So still the garbled text sits in the notes, and
now developers have to support two formats of data instead of one.. more
code duplication.
I always championed PalmSource by
suggesting this is due to the difference in mindset and perhaps a lack
of developers - that Palm pushed for
diversity of devices and would happily try and warp their OS to fit a
new device, while Microsoft clamped down specifications very rigidly. Hence
to this day there are quite a variety of Palm OS units from watches to
game machines to phones all on "one OS" (riddled with bugs), while in
the Microsoft PDA world you'll find a fair variety of devices all
working more or less alike (not as many bugs, but surely some duzzies!)
Good and bad - Pocket PC is fairly easy to develop for, since all devices
have pretty much the same feature set, but Palm OS is fragmented to all
heck. Still, I like the variety of devices.... well, I did, until I had
to support them.
In the end, to summarize - it is a chore to develop for Palm OS
these days; most disregard all the older devices, but I try to maintain
compatibility with and not forgot those trusty old friends... but the time
is coming when I just start ripping out all that code -- it is difficult
enough keeping up with all the new bugs in the OS, let alone old ones.
Developers wish to be productive as time is expensive and tight, yet 90%
of time I find is being consumed finding OS bugs and investigating work
arounds. How much more intuitive would an interface be if we could
focus on improving it, rather than dealing with bugs? How many more
features would there be? How much lower would prices be? We have no
time for 'business logic' when we're wasting it on fluff. This my friends
is one of the biggest challenged in Palm OS development.. they're taking
the fun out of it, and making it hell.
Palm Competing with Developers
So great, we were all happy PalmOne decided to start growing their PIMs.
Typically, they just altered the database format without actually
improving it, but worse still - they didn't document their changes for
years - I can only assume they resisted this because they feared
competition from the very developers they depended upon. What other
explanation is there? Certainly there is no patenting or licensing
issue at large.. though it could just have been lack of manpower to
make it happen. Regardless, it made our lives _miserable_ and to this
day we're all still dealing with the fallout.
Remember, when a tire won't fit on your car, you don't blame GM.. you
blame the maker of the tire. When users found crashes, slowdowns or weird
data loss, they blame the developers. Not PalmOne for breaking their own
specs and then not informing anyone how it worked..
What they did was introduce new databases, undocumented; when developers
applications opened the old databases, the OS would kick in and ferret data
between the new and old. Sounds good on paper, but evidently PalmOne didn't
debug or test this process much, since it was flawed in the design and
buggy in its implementation - frequent device crashes would occur due
to bugs in their 'PIM sync' (my name) code. Things would slow down
enormously (especially with the later NVRAM devices) as data was
forever being copied back and forth, something that never occurred
before. Worse yet is data loss as the PIM sync layer tried to guess
the intent of things going on and often got it wrong. So the developers got
flack again, for PalmOne's mistakes.
They finally did release some documentation, years late. I don't
get it. Months of suffering for no good reasons.
The real tragedy is the lack of learning - this wasn't the first nor
last time the documentation didn't come until devices were on the market
(sometimes for years) and our apps late to be stable. I'm not even going
to talk about the dozens of serious NVRAM flaws - NVRAM being a great
little system that users need, but when a large %age of apps become
unstable it makes you wonder how much testing went on..
Development Tools - or Lack There-Of
Microsoft, aside from its immoral behaviour, knows
how to make some decent tools for developers to work with; their OSes tend
to be backwards compatible and they do a pretty good (if pricey) job of
ensuring developers can get things done. Their Visual Studio suite is
quite popular and its available for C/C++ and Visual Basic development for
Pocket PC - a big comfort to many. With their new .net platform you also
get C# tossed in, and halfway useful drag and drop GUI building tools.
With Palm OS 5 and earlier, the official development suite was always
Metrowerks Codewarrior - a pretty good product. For OS5 "ARM" code
development it was useless for awhile (seriously buggy), but they worked
hard to move CW 9.0 to become CW 9.3 - flaky, but overall pretty darned
decent for doing OS 5 68k+ARM development. A lot of us have been using
CW for years, since it was the best overall system - and it was
discontinued a year or so back or more. New developers have a hard
time locating a copy of the package to use, so they resort to the PRC-Tools
(freeware gcc, a thing close to my heart) which is much harder to get
into.. discouraging away new developers. CW went away likely because it
is hard to make money with development tools, but also because PalmSource
was building their own development environment - PODS - for their OS 6
product. So PODS is out and about, but as you all know OS 6 is not.
So PalmSource spent enormous resources and time working on OS 6 upgrades
and development tools for us, that we couldnt' really use.. as PODS isn't
so good for OS 5 and earlier. Great - Palm depending on out of print
tools, and PalmSource working away for years on things no one is using.
(gcc is great stuff that I use every day, but due to the 68k+ARM
system in OS 5 devices its never been all that convenient.)
So in the Palm OS world, developers have been cobbling things together ..
a sorry state of affairs .. something Palm should be ashamed of.
OKay, theres some other tools so lets flame them too - Java is a popular
thing these days, correct? Well, a shame that Palm only licensed Websphere
Micro for a few of their devices, and isn't really committed to it. Certainly
it is unusable since how can you afford to base an application (months or
years of work) on something which is only for a subset of devices and may
be dropped at any time? Certainly, Java isn't available on Pocket PC at
all since Microsoft isn't all that nice...
The reverse is also true; MS wants everyone to move to C#, but of course
they will never provide a C# environment for Palm OS, and Palm just can't
afford to kick off yet another development tool project.
If you want some cross platform code to work, your best bet is C/C++,
or maybe somethign like NSBasic or SuperWaba. Still, a painful situation.
One other sub-annoyance; people all want java and C# and python these days
as they're sick of building from the metal; sure, productivity is up with
these languages as they've all borrowed some of the best inventions of the
past and mixed in recent patterns and experience; further, they've got
enormous manpower poured ito making new libraries to save time. (C++ and
such never quite got so organized as to have these giant libraries bult
for them in standard ways.) Anyway, people are forever asking why all our
Palm OS apps aren't in java -- they seem to forget the fact that java
isn't really available or fast for mobiles (hell, most phones don't
even do JITting or garbage collecting these days) - that C and C++ are
the really truly only portable codebases out there, to this day. But
try explaining this to the newbies fresh from school, without sounding
like an old codger. They do likes their java ;) Sadly, me too -- but its
not an option so stop bugging me :)
The OS 6 Fiasco
This is much publicized so I needn't go into it much, but in summary -
PalmSource (the OS developer) spent years building and pushing Palm OS 6
as the follow-up to Palm OS 5. Numerous developers (including myself)
started working on OS 6 versions of our applications - either changes to
remain compatibile or re-developed to actually use some of OS 6 features.
Alas, OS 6 was not to come .. despite PalmSource pushing it and announcing
devices and partners, we all quickly realized it was doomed from the
get go, so all that time was pretty much shot. To this day, PalmOne
has been knocking out OS 5 devices with their modified (read, buggy) version
of OS 5 .. a great overall product, but with so many extensions and PalmOne
seemingly out of control with regards to quality and standards, its started
to be referred to as "Frankengarnet" by the developers and users.
Mac OS
Mobile developers aren't big companies with enormous budgets to burn, so
we like to be cross platform (using one codebase for multiple operating
systems) but its nutty over in OSX-land. We have a stable and dependable
conduit for Windows, so that Shadow Plan can sync with Shadow Desktop and
work very well. It would be nice if that same codebase could work under
OSX, but we quickly found out that not only would that not happen, but
that Hotsync Manager for Mac OS X is basicly a butchered up version of the
Mac OS 9 system - that modern OS X style code cannot work with it.
Wonderful how our Mac OS X application is OS X based and due to various
dependancies can't tie into this OS 9 Hotsync Manager system. What silliness!
Its possible to rebuild the needed dependancies, but I am rather sick of
returning to the dark ages, of building up from scratch for every
application somewhere along the line. At least Mark/Space is on top of
the ball and has built their own Hotsync Manager replacement that offers
modern OS X facilities.. but I do not feel it right to saddle our
customers with an extra product (and none to cheap!) just to cover up for
Palm's mistakes. So we .. as usual, work around the system limits and
bugs. Another shameful situation.
Globals, 68k/ARM, and Other 1970s Problems
Again, being spoiled by the wonderful development environments of the
1980s, am sick of painful development situations requiring great devious
minds to actually produce great things. Moving from 68k to ARM was a
difficult yet amazing thing for Palm - its incredible they pulled it off
as smoothly as they have; almost all Palm OS applications to this day are
68k processor applications, running inside of PACE - Palm OS 5's compatibility
system for its ARM processors. Amazing stuff, so that our old apps can run
on the new architecture without changes. Great!
The problem is when you wish to actually write ARM code, to get some
speed. PACE exacts a large performance hit - something you wouldn't notice
for GUI applications but when putting together a high performance
multimedia application such as a game or video player or the like, you need
the real hardware - the ARM system. Ben and the boys worked hard to make
Codewarrior 9.3 stable and its a great product overall.. but due to the
nastyness of the OS and interfaces, we're all forced to jump through hoops.
You see, the 68k suport is quite good of course, but for ARM code you have
almost no access to the rest of the application APIs unless you 'stub them'
over. So to begin with, we all have to make stubs like mad, so we can
make ARM code halfway as easily as 68k code. Ben really helped out here
and the community worked together to produce a lot of useful code, but
a shame PalmSource didn't provide it - they assumed we'd all be on
OS 6 by now, and it just didn't pan out. The whole ARM-68k division is
really quite painful - a lot of us have it pretty well mastered, but we're
sick of the pain; newbies - forget it. Theres just such a learning curve
and hump to get over that few ARM-based applications have come out.
Certainly, to this day its still difficult to port applications from
standard operating systems that use certain features, and you'd better
have CW 9.3 (costware and discontinued) if you want your life to be sane..
using the freeware tools should be enjoyable and powerful, but its taken
years for that community to get to this point. Terrible.
At least I did my part.. I got XCade out pretty fast (one of the first
all ARM applications).. but its not pretty :)
Another thing that annoys me is the whole launch code and globals
situation - its bizarre to be in the year 2005 here and have to deal with
things like an app being launched without globals being available, due
to how Palm OS works. It was innovative and a must-do-to-survive situation
back in OS 1 and 2 and whatnot, but for OS 5 .. we're just sick of having to
warp our code in these terrible fashions to make things run. Lets not even
talk about 64k segmenting and whatnot .. I hate nothing more than copying
functions around, into small little files, so I can re-order them so that
functions can reach each other. Good god, that was the 286 days :)
OS Behind the Times
We all like compatibility (see all that ranting above?), but we're also
cool with well controlled changes. If your OS is broken - fix it. We'll be there
if you manage that process, publishing specs and documentation so that we
can be ready when the newer versions roll out. Bring it on.
It seems alien again to be in 2005 with simple bleep-bleep alarms in
Palm OS; the user can't easily assign mp3s or wav files to alarms or
ringtones, without third party hacks. I've even written and given away
a simple hack for years, to let small wav files be used as alarm tones.
We need an OS change here, so that all applications can have a modern
alarm managing system. Theres some cool stuff in the OS now, so that
alarms can be made to vibrate or be silent from a user configuruation
level withotu changin the apps.. but wheres the modern stuff?
Missing APIs
This ties back to the question of whethor Palm treats developers as
competitors. Where are the APIs that the built in applications can use?
How can we, on a Treo, control the radio to the degree the built in
applications can? Perhaps its documented in some buried Handspring
reference, but when I take a quick look through the usual places (dozens
of out dated documents :) I find it hard to find some of these things.. if
they're to be found at all. Lets cut to the chase - we want a good solid
knowledge base and search, like Microsoft provides. (Well, better than
they provide would be nice ;)
Licensee Madness
This is mostly old news, but it drove us all mental over the years.
I still support many of these older devices so its still relevent to me
but not for much longer I tell you. Especially, Sony as they were King
of Bastards.
Sony pioneered high-resolution because PalmSource couldn't get something
together fast enough I suppose, but they failed to think ahead and made
a pretty dimwitted API. At least it worked more or less... but it was
prone to crash. If the user or developer did certain unknown things,
the device would just hang up - again, the developers took flack and
it took ages to work out the myriad causes. To this day I keep an old
device around for testing, since the problem randomly manifests. Sony
never fixed this bug.. years would pass and the issue only went away
when OS 5 came along and Sony pseudo-adopted the OS 5 high res
system.
Course, Sony was evil for refusing to document how their joystick
add-on operated. Their documentation to the developer consisted of saying
"it just works", but alas it did not - it depended on some undocumented
series of events occuring in order to work, so for all my gaming applications
it worked ramdomly. (That was fascinating.. it should work, or not work.
But working sometimes and not others.. bizarre.) It'd have been nice of
they'd taken 20 mins to document the requirements.
They also refused to document the sound system; as usual, developers
took flack for not supporting special audio on Sony devices, but such
is life. Years passed before Cliepet, a whiz in the community, reverse
engineered it all. Bastards.
I'll try not to go on, but the problem was PalmSource's lack in taking
control of the reigns - why did Sony and Palm and Dana Alphasmart and
Handera all work out different ways of doing things? Shadow Plan
supports some 4 or 5 different incompatible techniques for managing high
resolution. Again, think how much higher application quality, and how much
lower costs would be, if we didn't have to both develop dozens of duplicate
ways of doing things, and then support them over years!
Really, its amazing anyone developed for the platform at all when
you look at this mess :)
At least Tapwave got most of it right.. a shame they tanked :/
Installations
A standard installer process would have been nice; heck, thinking ahead
would've been nice, too, so that Sony and Palm installs could sit nicely
side by side without blowing each other up. We rely on Beik's Pilot
Catapult installer builder, since its always been simple and inexpensive
and fairly stable, but even they can't quite keep it solid due to all
the varieties of weird in the registry that the various licensees put
in. (Microsoft doesn't help with their eternal OS monkeying on the desktop,
making Hotsync Manager sometimes run in Administratior mode while other
apps do not, etc.) A shame when PalmSource laid off the fine fellow working
on a promising new standardized installer.
Quality Assurance
Last but not least is just the lack of QA here; how does it slip through
the testing cracks that sometimes a file can be copied out to SD from RAM
and then not back into RAM? Or that devices crash/reset more frequently
now than ever? Since the new PIM databases weren't documented we
were required ("supposed to") use the old PIM database interface (thus
being denied the new features, plus gaining all the slowdowns and bugs)
and yet the magic PIM sync would frequently crash. Crazy stuff, and
all the blame comes back to developers.. with nothing we can do.
Whew!
When I began this opus I had not intended to write all this out .. I
merely wanted to rant a bit, to get it off my chest how miserable the situation
has been for years.. but out it poured.
Now that I'm all healed, you're suffering. Merry Christmas! :)
They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...
Developer rants at PalmOS - where to go next?
Also
Full article is here:
http://www.codejedi.com/cgi-bin/blog.cgi/technology/mobile/20060110directions.blog
A number of developers I know have been asking the same questions that weigh heavily on my mind - we all watch the tides a' turning and in our markets and domains we have to keep at the front or get left behind, flotsam of the mobile digital age. Of course, being in the front - cresting the wave as it were - is a great risk yet also provides many advantages. When it comes right down to it, being in front is where the fun is, is where the adventure of it all is.. until the dragon catches up.This has been written up on my smartphone (Treo 650) while sitting over a warm drink. Amazing how long you can write on a thumbpad with enough caffeine!
The questions folks ask all boil down to - Where are things going? Alas, if only we knew we could all sit around in hot-tubs after beating the racetrack, right? PalmOne (er, Palm, again. They renamed their software store's internal name my error logs tell me. Amazing the variety of things we deal with every day, no? *g*) is selling a Pocket PC Treo now (the Treo 700w), which looks like a winner of a device (ignoring some first-device woes) especially in the corporate environment. (Remote Desktop from your phone - way cool.) PalmSource is going to be renamed, and developers can't find the official dev-kit for their preferred Palm OS platform. People are buying into the online web-applications craze (Google Maps blazing the way, but is the public ready for trusting a remote online organization with their private or corporate data in Online Office?) The Oqo and other tiny-notebooks are making inroads.. will dedicated handheld OSes still be around in a few years? With handheld sales shrinking (being replaced by smartphone sales, where smartphone customers are less aware they can purchase third party applications) sales for most developers tend to be down a bit. The US Dollar has tanked, meaning those of us outside the mainland US are losing on The Money Multiplier. The online EST stores (the guys selling our apps) are gouging more and more of the profits leaving smaller crumbs for us. Tough times ahead!
The Treo 700w being Pocket PC, all by its lonesome, scares the bejeezus out of a lot of people -- can cats and dogs co-exist under one roof? Certainly, I'm already getting support questions about Shadow Plan running on the new device as people assume they can upgrade from one Treo to another. (StyleTap to the rescue for now, though I really think they should extend Palm OS -- why not offer 240x240 native mode to the Palm OS application, so that we can better support Pocket PC from a Palm OS application in StyleTap? How about different icon-sets for Pocket PC? Supporting Pocket PC virtual grafitti more directly?)
Questions are many; I've long been looking into the Pocket PC and Windows Mobile markets, as they're making up more and more of the whole pie; the VGA devices such as the gorgeous Axim X50v are carving away with wonderful new hardware and solid development platforms (albeit a clunky OS at times.) There are some simple outliners, and some stronger list applications, but nothing like Shadow (and to possible competition.. don't take this as a hint. Let me have my pie, okay? ;) exists per se.
Make an online application, on the assumption more and more devices will have Wireless Everywhere from bluetooth to 802.11 to GPRS and EV-DO? I know here in Canada we pay big for data services, but unlimited plans at $25/mo abound in the US last I checked. Will people trust storing their data on my server and editting via a fancy web application (the so-called Web 2.0 or AJAX movement, which is just a fancy way of saying javascript), or even via a handheld PDA or phone or RIM client? Will they
? Do we put forth the enormous cost of developing and support for Pocket PC, or go for better RIM and Nokia support as those platforms finally and ever so slowly realize they're actually useful for supporting applications beyond email? (Note that I do have some Pocket PC code out there, but so far I have not moved my primary application over to that platform. Its been tough enough keeping up with PalmOne.) Remember, for a lot of us we began in the dark ages of Palm OS and so our code had to run in very low spec environments, and so the 'cohesion' is high -- our code is tied closely to the platform, making porting difficult. Porting to a new platform more often than not (for older applications) means rewriting all or most of a system - very costly. Further, when building new apps for the Windows platforms, do you target portable C/C and the ancient win32 API, or go to C# and .NET - an all new API that essentially locks you into Microsoft platforms, or SuperWaba (a variant of Java, since MS won't support java on their platform) - again locking you into a specific API, that is at least cross platform itself. Tough questions indeed.
As always, I think 2006 will be an interesting year for sure -- things are a' changing, and the technology is getting better and better though surely not fast enough. A lot of developers will be making the hard decisions very soon - adopt a new platform - and which one.. or not. Move forward, take on large challenges and risks.. or be left behind. Swim, or see you later.
This will lead to some attrition; for the heartier folks who always move forward there are risks.. but being first mover, and early mover, there are great rewards. Your beloved Palm OS users who follow their favourite apps to Treo 700w's will loyally adopt your software. Moving to RIMs and Nokia's could well open up whole markets for you.. or bankrupt you.
Now we'll see who the men are ;)
They said I only argued for the sake of arguing, but after an hour I convinced them they were wrong...
RE: Developer rants at PalmOS
-------------------------------
Editor, http://Pocketfactory.com
Contributing Editor, http://digitalmediathoughts.com
RE: Developer rants at PalmOS
As a mere user with NO aspirations or skillz to be a developer/coder, I can still sympathize with the guy as a LOT of the headaches he alludes to (the miserable transition from T|T style 1-handed navigation to Treo/Handspring style) are felt/burdeoned mere users such as myself.
It just sickens me to see the sorry state of the POS software market. You have a handful of legacy apps not updated since '02 or '03 (ie for the T|T) then others that run acceptably well on T5/TX but poorly on all older models. Right now the handheld industry suffers from the worst backwards compatability of any major platform that claims compatability. Ie: You can, with enough effort, get most older DOS or Win 3.1 titles to run reasonably well under XP if you REALLY need it. A GameBoy Advance can still run classic GameBoy titles from 1989. A new PS2 slimline can run PS1 titles from 1995. Only the Palm & Mac platforms seem so impossibly broken--and at least Apple had the good sense to throw out legacy compatability in the name of the future/speed/stability under OS X. Palm's just drifting in circles and losing pieces of the ship everytime a wave hits it.
Pilot 1000-->Pilot 5000-->PalmPilot Pro-->IIIe-->Vx-->m505-->T|T-->T|T2-->T|C-->T|T3-->T|T5-->TX
Click here for the full story discussion page...
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
Ed
Palm m125 December 25, 2002 to March 24 2004 > palmOne Zire 71 March 24, 2004 to March 31, 2005. Tapwave Zodiac 1 April 18, 2005 to November 2, 2005 > palmOne Zire 72 November 2, 2005 to present