How PalmOS lost it - readers
Warning: contains strong language
You're not exactly backward in coming forward, are you? In the latest twist of PalmOS soap opera last week, Japanese mobile browser company Access bought PalmSource. Could Palm's fate have been any different, we asked. This selection of views contains pretty strong language - so be warned.
Many Reg readers and Palm fans have pretty strong opinions of how Palm Source lost the plot - with developers and users equally well represented.
Gene Cash, a former Palm fan turned very-reluctant Pocket PC user, is as forthright as any -
"I hope Access buries Palm in a deep hole and shovels nuclear waste on top. And then pees on top of that."
Once Palm had it all, he reminds us.
"People wanted desktop PCs that fit in their pocket for $99, no matter how impractical or impossible that was. They wanted this all-singing, all-dancing device with features they weren't going to use anyway. Palm made its mark offering something VERY easy to use with a VERY well thought out user interface. I could buy gas and write it down in about 6 clicks. It was also quite cheap for what it did.
It didn't do everything, but what it did (be an organizer) it did very well. Now all that simplicity is gone, and the Tungstens & Zires are a pain in the ass to use.
"The original display was black-and-white, but you could read it outdoors just fine, unlike the current color displays that are totally washed out by sunlight and require a battery-eating constantly-on backlight. I go out to my bike, try to look at my grocery list and I can't even tell if my Zire 31 is on! I could go for 2 weeks of constant use without recharging my Vx. The Zire 31 needs to see the charger every other day or else!"
Gene then lists a long list of design errors - no hard case for the Zire, disappearing buttons that compromised ease of use, glued-shut hard-to-repair cases, and models being discontinued - and consequently unsupported - too rapidly.
"Still another problem is it's a bitch to write a Palm app. There's one decent book from O'Reilly, and all the good tools are hundreds of dollars. You can use GCC but it's really painful, and you have to just about glue the bits together by hand," he adds.
He concludes -
"No matter how much I love it, I'll NEVER buy another Palm device, and I tell all my friends not to buy one. I'm sure if they offer Linux, they'll trowel a proprietary interface layer on top of it w/o source, and I'm not playing that game.
Now you CAN get desktop PCs that fit in your pocket, but it's $400 and a major pain in the ass to use, and the battery lasts about 8 hours. I'm even considering a Dell Axim now. Ugh. But only if I can slap Linux and Python on it. Hopefully my m500 will last another couple of years."
Sava Zxivanovich sees PalmOS diminishing importance as part of a bigger trend, but blames the lack of technical investment -
"As a PalmOS developer I really don't think it is a mature communication system and its support is non-existent. Compare it with any other communication/mobile OS and results is obvious and expected."
"I love my Palm. I've found it enormously helpful over the 5 years I've had it," writes Ian Hooker. Ian blames two factors: perceived lack of Microsoft integration, and color screens.
"The end product as sold to consumers wasn't well enough integrated with Microsoft Office. Sure you could always synch with Outlook and there was always Documents To Go, which was later bundled with the PDAs, but it was too little too late. Too many people were either ignorant of or sceptical of whether their expensive new toy would work with Office.
"And in my organization, the IT guy bought Pocket PC devices because he only knows Microsoft software and wasn't willing to take a risk on something unfamiliar. Secondly, people liked colour screens. Colour screens cost a lot in battery life for little extra value over a black and white screen for most applications, but there it is. People like the bling bling. At a crucial point in the history of the PDA, Pocket PCs were a safe option with the bling bling. It might not have been true, but perceptions are as good or better than truth."
It's striking from the mailbag how Palm Inc's poor hardware designs turned people off the "platform". Mike Panero is a good example - and he's even angrier. Comparing two broken machines - a Psion and a Palm - he also comes up with a unique neutron star-dense prose for which no translation is necessary.
Bah, Palm any good? What about Psion?
I had a real cheap Psion, which, of course BROKE
So like a TWAT I got me a PALM
Ha, paid £150 sold it one week later( yes it was FUCKING SHIT*) for £20, my bust Psion was still worth £50 even so you could buy it new for £80
PALM Came with a CD & a fancy kray-dole "Oh USB !== Unused Socket @ the Back"
like F, that grafitee poo poo no S/S crap clock eats batteries (this is a 68K moto FFS, CMOS! HA!!)
PSION 2xAAA== 1 MONTH, PALM 2xAA== 36 Hours !!!
PSION, box no CD no nuddint but WORKS PALM box CD CABLES SHIT LOADS OF SHIT, worth, guess, SHIT!!!
(And someday, perhaps all computer histories will be written this succinctly)
Eric Moon also blames the fact that Palm didn't grapple with early, 'Touchdown'-era design decisions.
"You wrote: 'The decisions it took that year to bring 'Touchdown' - the original Palm Pilot - to market were exactly the limitations that made it a success.'
"I wholeheartedly agree, and as you point out this is decidedly a two-edged sword. Thus, many years later, 3rd-party application developers who learned the original API can still churn out basic Palm OS apps quite quickly - but those wanting to write more involved apps have to jump through a lot of hoops, and the nature of the API somewhat discourages good engineering.
"Every sufficiently large Palm OS app tends to devolve into spaghetti code. Also, the original API is very PDA-centric, assuming a square screen, a stylus interface, and a generally modal UI. None of this is very well aligned with a phone formfactor, particularly for cheaper phones.
But Eric sees a bright side to the move to Linux, as he explains -
"When Palm started licensing the OS to other hardware manufacturers -- and this was even before the PalmSource spinoff -- as they were then in the business of producing different "kits" for different audiences, 3rd-party developers still got the public API, but licensees got a much more configurable toolkit that they could use to customize the OS for their own hardware. The problem there was that this licensee toolkit was quite involved and, generally, licensees needed a lot of handholding and training in the vagaries of proprietary Palm OS internals.
"The big advantage of the Linux move, as I see it, is that there are tons of embedded developers (particularly in Asia) who already know POSIX, the Linux driver model, etc., so the learning curve for customizing a release of Palm OS for Linux would, in theory, be a lot flatter."®