Palm Offers Refund to m130 Owners

EyesWideOpen writes "On Wednesday Palm began notifying registered m130 owners "that they were entitled to a full refund, including taxes paid on the PDA" for misleading them about the actual number of colors the product supports. The m130 was originally advertised as supporting 65,536 colors when in actuality it can only display 58,621. Owners who choose to forfeit the refund and keep the PDA could instead download a free version of the video game SimCity." Looks like a great deal for those who don't care about the bit depth of their PDA, and a way out for those who do.

Comments

64K on colours on a 160x160 screen.

[MavEtJu]

I would be shocked too if I would find out that I can't display all 65536 colours on a screen with 25600 pixels!

Re:64K on colours on a 160x160 screen.

[kaphka]

Funny, perhaps, but the number of colors you can display at one time is irrelevant. What's important is the number of colors you have to choose from, since any particular image can only make use of a small fraction of them. If you're rendering a portrait, for example, 1600 shades of skin tone are a lot better than 100.

Re:64K on colours on a 160x160 screen.

[digitalsushi]

me too.. i thought his sig was going to print out an ascii art bunny... all it did was crash our web server. our mail server didnt crash though.. i tried it on that one too, but they wont give me root on it for some reason. maybe you have to wait longer. i dunno. oops. heat alarm. brb.

Woo Classic Maxis!

[Cyno01]

Why cant i find cool old games like this for my palm? Can anybody direct me to a site that offers cool games for palms pilots (preferably classic games, like that flash version of pitfall somebody posted the other day)?

Re:Woo Classic Maxis!

[BagOBones]

The only site I ever needed to Palm software.
http://www.pilotgear.com/
Just do a search for Simcity in the software section

Re:Woo Classic Maxis!

[questionlp]

There is always Bejeweled/Diamond Mines :)

That game is definitely a productivity virus... even more so than Solitaire (Minesweeper, Hearts, etc.).

Re:Woo Classic Maxis!

[i0lanthe]

Don't forget palmopensource.com and freewarepalm.com which enable a significant time savings (i.e. anyone who looks for games there can automatically skip the "find cracked warez version" step that seems to be dear to the piratical hearts of much of the more vocal /. readership. To whom I say, good grief, get a haircut and a job and cough up the dough already.)

Does it make a big difference to people?

[10 Speed]

Other than bragging rights what difference does the reduced amount of colours make?

I presume people are not purchasing these to watch movies

I think it will be interesting to see how many people ask for the refund...

Re:Does it make a big difference to people?

[RatBastard]

Actually, yes. Paml M130's generate 58,000 colors through some sort of dithering or pixel strobing technique. The display can only generate 4096 actuall colors. The problem is that this makes for really crappy images and, more to the point, is a flat-out lie.

The REAL point of contention is not the number of colors, but the fact that Palm Inc. lied to its customers.

Re:Does it make a big difference to people?

[Cyno01]

dont hold out on us, what handheld are you using? the new clies take memory sticks (magic gate*snicker*) but i doubt those could hold a full length movie

is this really a big deal?

[GoatPigSheep]

Would you return your pda because it only displays 58000 colours instead of 65000? I mean, unless you are doing photo editing on it, it doesn't really matter. Besides, not having to display the extra 7000 colours saves energy.

Re:is this really a big deal?

[Target Drone]

Would you return your pda because it only displays 58000 colours instead of 65000?

Would you return your Pentium because it does almost all divisions correctly?

Like the Pentium bug this isn't a cases of whether users will notice a difference. It's about a company owning up to its mistakes.

Re:is this really a big deal?

[Osty]

Would you return your pda because it only displays 58000 colours instead of 65000? I mean, unless you are doing photo editing on it, it doesn't really matter. Besides, not having to display the extra 7000 colours saves energy.

The problem is not that it can display only "58000" colors, but that it can really only display 4000 colors. That 58,000 number is arrived at by "using a variety of techniques--including turning pixels on and off and combining nearby pixels." (News.com article) So yeah, if Palm advertised that the m130 could display 65536 (16bpp) and it can only do 4096 (12bpp), then I would be complaining. HP had the same problem with earlier Jornadas they released, because they advertised 16bpp and only supported 12bpp (the crazy thing here is that they call the problem a "glitch", when it's a simple fact that the screens they used only supported 12bpp -- sounds like a glitch in the manufacturing process by choosing to use a cheaper screen). Compaq didn't have this problem, because they always advertised at 12bpp, not 16bpp.

In other words, the issue here isn't that the PDA can only do 12bpp, but that Palm advertised it at 16bpp and was caught out in their lie.

Re:is this really a big deal?

[DeadMeat (TM)]

You don't really get 58K colors; that's from Palm counting colors you get when you dither, which doesn't really count. Only 4096 colors are actually available in hardware.

And yes, if I had bought an m130 for viewing photos, I'd be infuriated, because that's blatant false advertisement. At least they're doing the honorable thing, if a bit late.

Re:is this really a big deal?

[martyn s]

That link doesn't work, so try this

I didn't know what dithering was before I looked at the picture but this is what I gather from it. If you have a 1 bit display (just black and white), if you make every other pixel black, and every other pixel white, it will give the appearance of being gray (especially at higher resolutions). That is what dithering is. This is opposed to showing a pixel that is actually gray (half black half white, that is, each sub pixel [red, blue, green] on equal intensity, at half intensity). So the difference between 12 bit dithering, as the m130 does, and true 58,000 colors is considerable. The fact that Palm's spin on it is that it shows 58,000 colors instead of 64,000 leads me to believe that they knew all along about the limitations in the device.

Even if you can get 12-bit (dithered) color to look almost as good as non-dithered 16 bit color (which you can't, but lets just assume), it's still fraud. 16 bit color can be made to look even better if it is dithered. The only way they could've avoided fraud (and even then it would've been sketchy) is if they said "16-bit quality color" or "as good as 16-bit color"

And whoever thinks that the difference between 12 bit color and 16 bit color is just for bragging rights, I suggest they play video games. Even with 32 bit color, if alpha is using some of those bits, you will *still* see color banding, especially in motion. The next generation of videocards is working on 64-bit color (although, they're not actually displaying at 64 bits, just 64 bits are used for calculations, to minimize cumulative color distortion through multiple passes).

Re:is this really a big deal?

[Osty]

And whoever thinks that the difference between 12 bit color and 16 bit color is just for bragging rights, I suggest they play video games. Even with 32 bit color, if alpha is using some of those bits, you will *still* see color banding, especially in motion. The next generation of videocards is working on 64-bit color (although, they're not actually displaying at 64 bits, just 64 bits are used for calculations, to minimize cumulative color distortion through multiple passes).

Just a small nitpick, but 32bpp has a built-in 8 bits for alpha, so you're not sacrificing any color quality by using it. 32bpp means 8 bits red, 8 bits green, 8 bits blue, and 8 bits alpha, or 24bpp + alpha. Yes, you may still see color banding (not as much as you would with 16bpp, of course), but I really doubt you'll notice it so much. As far as the upcoming 64bpp cards go, you're absolutely right -- they'll still only display 24 bit color, but the added bits will allow for greater precision during multiple rendering passes (mmm ... 16 bit red, 16 bit green, 16 bit blue, and 16 bit alpha). If display technology were to catch up to video cards (is that possible? can CRTs do better than 24 bit color? I'd bet LCDs could, but that'll initially be very expensive), then we could actually display 64bpp. Of course, that still wouldn't shut up the people that complain that even though 24bpp allows for 16million colors, there are still colors it can't display (uniform steps between each 256 shades of r, g, and b means that there are some subtle shades that nobody will miss but photo professionals, if even them).

Re:is this really a big deal?

[Dirtside]

Why is this modded "insightful"? The analogy is horribly crippled. A display that has to approximate 10% of its colors is not going to make any material difference. A CPU that miscalculates things is going to cause *actual* problems.

The odd part is, the last sentence ("Like the Pentium bug...") more or less contradicts the entire purpose of the analogy. I'm beginning to think Target Drone didn't say what he meant to say. The Pentium bug IS a case where users will notice a difference -- namely, incorrect results, weird crashes, etc. In the Palm case, most people wouldn't be able to tell a difference.

Not that I'm saying that the company shouldn't own up, but let's not use false logic to make a point.

Re:is this really a big deal?

[xswl0931]

How about if you bought a 650 Mhz processor, but it turns out to be 580 Mhz? Well, the reality is that the m120 is actually 12-bit which displays 4096 colors and cheats to get the 58k magic colors. From seeing it myself, there's no mistake this is a 12-bit display, not a pseudo 15.5 bit display. Just compare it to other 12-bit displays and you'll easily see the difference (or similarities as the case may be). They lied once, and now they lie again. I suppose you could say my Pentium III is a 128 bit processor since it could do 128 bit calculations... right....

Re:is this really a big deal?

[martyn s]

Well, I think CRTs have infinite precision. What I mean is it all depends on the RAMDAC. But as far as I know it makes absolutely no difference. Anyway, thanks for clarifying that 32-bit color includes alpha.

Anyway, I play Halo, for Xbox alot and noticed that there is color banding. If you stop moving, it instantly disappears (in other words you stop noticing it). But it's pretty bad, I mean, I'm not particularly sensitive, as far as I know. So I assumed Halo has 32 bit color, and it almost certainly has alpha. So what's the deal? I guess the only conclusion is that Halo must be only 16 bit, with 4 bits alpha. As much as I love Halo, the color depth and the low resolution (television) ruins it for me. Actually, it's not so much the low resolution, but the jaggies that come with low res. Effective anti-aliasing (mmmm radio 6x sampling) would take care of it pretty well, but unfortunately the Xbox doesn't do anti aliasing without taking a big hit. From what I see it looks like the anti alias only certain polygon edges, but there are still jaggies (big ones!) EVERYWHERE. Whatever.

Re:is this really a big deal?

[Eccles]

Can CRTs do better than 24 bit color?

CRTs are inherently analog in their signaling. The RAMDACs on graphics cards convert the digital output into analog signal levels (thus DAC, digital->analog conversion) which then go along your video cable. If the DAC was capable of converting 16-bit inputs, and the graphics card fed it 16-bits/color, you could have effectively 48-bit color.

DVI signals are digital, and thus (in all specs so far, AFAIK) 8 bits/color or less.

Re:is this really a big deal?

[Osty]

Anyway, I play Halo, for Xbox alot and noticed that there is color banding. If you stop moving, it instantly disappears (in other words you stop noticing it). But it's pretty bad, I mean, I'm not particularly sensitive, as far as I know. So I assumed Halo has 32 bit color, and it almost certainly has alpha. So what's the deal? I guess the only conclusion is that Halo must be only 16 bit, with 4 bits alpha. As much as I love Halo, the color depth and the low resolution (television) ruins it for me. Actually, it's not so much the low resolution, but the jaggies that come with low res. Effective anti-aliasing (mmmm radio 6x sampling) would take care of it pretty well, but unfortunately the Xbox doesn't do anti aliasing without taking a big hit. From what I see it looks like the anti alias only certain polygon edges, but there are still jaggies (big ones!) EVERYWHERE. Whatever.

You need to tell me how you're connecting the XBox to your TV. Are you using composite (single RCA jack for video), S-video (4-pin din plug), or component (3 RCA jacks for video)? Are you playing on a TV that supports progressive scan output? (like 480p on an HDTV) Given that composite is bottom-of-the-barrel quality, and s-vid isn't much better, you're going to see artifacts and general crappiness with both of those, regardless of the bpp being used. Plus, those both run interlaced, so your vertical resolution is basically halved when you move around (turn your "head" side to side). Component is much better, and allows you to use 480p (assuming your TV supports it, and you enable it in the XBox dashboard). Even so, you still may see banding, simply from the aliasing of pixels from a limited resolution (you're essentially playing in 640x480, so there's only so many pixels available to make up the screen). No XBox games that I know of do FSAA yet, nor 1080i (I heard there was one game either out now or out soon that does 1080i, but it didn't sound interesting), but when games start using one or the other, you'll see image quality improve even more (I wouldn't expect both FSAA and 1080i -- at 16x9 aspect ratios, 1080i has more pixels than a 1600x1200 PC display, and FSAA even on the most advanced PC vid cards still chugs along at that res).

Anyway, as far as bpp goes, if you only notice the color banding when you're moving, then it's likely not an artifact of low bpp (you'd notice it when stopped, or even in screenshots, if that were the case). As I said above, I'd be willing to suggest that what you're seeing is aliasing effects from low resolution combined with an interlaced picture. As an example, watch an indy car race on TV some time, and look at the stands as the cars wizz by with the camera moving to follow them. You'll see a kind of moire pattern show up, but if the camera were to stop moving, the stands would look just fine. Halo running on my 46" Mits 16:9 RPTV looks great at 480p, though like you I would like a higher resolution. Never really notice too many jaggies, though.

Re:is this really a big deal?

[Osty]

CRTs are inherently analog in their signaling. The RAMDACs on graphics cards convert the digital output into analog signal levels (thus DAC, digital->analog conversion) which then go along your video cable. If the DAC was capable of converting 16-bit inputs, and the graphics card fed it 16-bits/color, you could have effectively 48-bit color.

Right, what I meant to say was: Is are the phosphors (sp?) used to actually display the colors responsive enough to make 48bpp/64bpp worthwhile? It's all well and good if the video card and RAMDAC can handle the higher bpp, but if the phosphors can't handle the added granularity, what's the point? RGB(0xFFFF, 0xFFFF, 0xFFFF) will look exactly the same as RGB(0xFFFF, 0xFFFF, 0xFFFE) if the phosphors can't handle such a minute difference.

Re:is this really a big deal?

[be-fan]

Actually, with HP, a different display chip was substituted in the manufacturing run by accident.

Re:is this really a big deal?

[martyn s]

Umm, no you're totally wrong. I have a samsung 27 inch HDTV. I have it running in progressive scan too. If you don't see jaggies you must be totally blind. Do you see jaggies in *any* games? Do you know what jaggies are?

The fact that I see banding only when it's moving has nothing to do with video artifacts. I'm no fool; I know the difference between video artifacts and color banding. Trust me, it's absolutely color banding. I can't really come up with a good explanation for why I only see it when it's moving, but I think it has to do with the shading changing as you move or something. You see banding in the shadows. Where there are very subtle differences in color to show shadow.

If Halo ran in 1080i I wouldn't need 6x anti aliasing like the latest radeon has. But it's one or the other: high res, or anti aliasing. I never said I needed both. If the Xbox has a chip equivalent to the Radeon 9700 it would be able to render Halo easily with 6x anti aliasing.

Dude, you don't have to explain what happens when you have an interlaced image. I know. I'll just assume you're explaining it for everyone else's edification.

Anyway, 1080i has more pixels than 1600x1200, but on the other hand, it doesn't. See, even though they say it's 1920x1080, since it's interlaced, only 1920x540 pixels are being shown per frame. But since The Xbox can't pick and choose which pixels to render, if it wants to run in 1080i it has to render all 1920x1080 pixels, and throw out half of them. That's why 720p is much more effective than 1080i for videogames. The resolution is pretty much the same as 1080i, but with 720p there are no wasted pixels, so it's MUCH easier to render.

Anyway, the fact is, there is definitely color banding in Halo, and yes, I can only see it when I move. When I stand still it like fades away in a few seconds. In other words the fact that I can't see it has more to do with my brain than the actual colors. It doesn't actually "fade away", I just fail to see it after a few seconds. So the difference in colors are the same whether I'm moving or not, but I have an easier time perceiving it while in motion.

I just gotta say something: I never understand people who don't see jaggies, in halo, or anywhere else. They're everywhere! I got a bunch of cool images showing the difference between pictures with jaggies and without jaggies. Just email me and I'll give you my AIM name and I'll send them to you if you want. If you can't see jaggies in Halo YOU'RE BLIND.

BTW, the original post of mine should've said "mmmm radeon 6x anti aliasing" not radio

Re:is this really a big deal?

[EvilBudMan]

No, not unless it affected sometthing I did. I would gladly take the game for their fu-bar since I don't have it.

Re:is this really a big deal?

[Dirtside]

Yes, that's a much better analogy that I fully agree with :)

I shouldn't have said that there's no material difference between 58k colors and 64k colors (or 12-bit vs. 16-bit, or however you want to characterize it); that was bad phrasing on my part, and I apologize. My point was that the analogy to the Pentium division bug was flawed, in that the two situations were not comparable.

To put it another way... a division bug of that magnitude renders the CPU essentially unusable for everyone, but a MHz misclassification of that magnitude is simply annoying. Although I suppose you COULD argue that the color difference makes the Palm "unusable," but I still maintain that a majority of people would not consider the Palm unusable even if they DID know about the color difference, and ultimately that the Palm color difference is not fatal the way such a division bug would be.

Re:is this really a big deal?

[devnullify]

Have you ever used a P60 with the bug? I used one for 2 years and never experienced a single problem I could trace back to the bug. Of course Windows crashed, but it did on everyone's machine at the time. I later installed Linux on it once it wasn't really useful as a desktop anymore, and it did that for another 2 years.

It isn't a fatal bug. Merely an annoyance that almost never presents itself.

Re:is this really a big deal?

[Eccles]

Right, what I meant to say was: Is are the phosphors (sp?) used to actually display the colors responsive enough to make 48bpp/64bpp worthwhile?

Yes. There's a question of how much we would be able to see, but I've seen banding on 24-bit images. 30 bits (10 bits per pixel) would probably be sufficient to hide any banding from the human visual system.

Re:is this really a big deal?

[jlv]

Sorry, the division bug of the original Pentium does not "render the CPU essentially unusable". My P60 system is still going strong 6 years later. Way back when, when it was all I had, I even used the darn thing to do my taxes. I checked the math; the rounding problem did not affect the results. OTOH, I did use a test that showed the div bug was indeed there. I never replaced the CPU because the system was use in a critial fashion at the time and I couldn't afford the downtime. For the last few years it's been operating as a firewall, and barely done any FP.

Intel's mistake at the time was saying "this bug won't affect anyone". They didn't intentially create the problem, it was a bug in the chip design. They made a PR blunder by trying to sweep it under the rug, but they finally reversed themselves. It meant a huge earnings hit at the time (although it created a nice aftermarket for cheap Pentium-powered jewerly).

Does the use of 12-bit color make the m130 unusable? No, of course not. It's probably a great 12-bit device -- even better because of the "special dithering" that gets an effective 58000 colors.

The error here is Palm advertising it as a 16-bit device in the first place. The even greater error is Palm continuing to say "it can display 58000 colors, not 65536". They need to fess up with "it only REALLY displays 4096 colors". That they haven't said that is an example of their continued arrogance. I hope the market punishes them.

Darn, I should've sold my Palm stock.

[I was originally going to mod the parent down, but I felt like responding instead]

Re:is this really a big deal?

[AvitarX]

I read that it would take a heavy floating point spread sheet user 100 years on average to encounter the bug. They did the recall to save face, not because it was cousing systems to crash.

Re:is this really a big deal?

[Dirk Pitt]

I don't think the parent actually alluded to the original Pentium bug; if you read more carefully his post is actually correcting the broken analogy from an earlier post. I think he was saying that _if_ a division bug in a processor was the _same magnitude_ as Palm's error in number of colors (12 vs. 16 bit), it would render a processor unusable. You might have inferred incorrectly because of other posts that do mention the infamous bug, so good thing you didn't mod down ;-).

Re:is this really a big deal?

[jlv]

Yes, I'll agree I wasn't very clear when I said "the parent".

The comment I meant was Dirtside's first response to Target Drone , where he says (repeatedly) that the Pentium with the floating point bug were useless.

Re:is this really a big deal?

[Black Perl]

Why is this modded "insightful"? The analogy is horribly crippled. A display that has to approximate 10% of its colors is not going to make any material difference

Actually it has to approximate 93% of its colors (all but 4096 of its "58000").

A CPU that miscalculates things is going to cause *actual* problems.

The miscalculations were actually hard to come by; they only happenened in the FPU and only under rare circumstances. The vast majority of Pentium users never encountered the bug.

let's not use false logic to make a point.

Let's not use misrepresentation to make a counterpoint.

Re:is this really a big deal?

[Dirtside]

Actually, I only mentioned the Pentium bug once in my original post. The other mention (first paragraph, last sentence) was a more general statement delimiting the difference between a CPU that miscalculates things, and a display that has fewer colors.

But I've already got a PortaPam!

[EvilAlien]

What would I want SimCity for?

"We lied to you, so here is a refund... oh, you like the product anyways? Well is is a crappy game for free. Oh, you already subscribe to alt.warez? Well... here... um. *click*"

Re:But I've already got a PortaPam!

[0ptimus]

I've got an idea! Why don't you just chalk up the whopping large registration fee and buy the program!? Wow, that's a novel thought.

Re:But I've already got a PortaPam!

[BigBadBri]

What's wrong with spending 20 bucks to register the damn game?

Tightwad!

Re:But I've already got a PortaPam!

[Qrlx]

Dude: Go buy Bejeweled. Quit being a cheap ass and pay the twenty bucks or whatever.

Your wife will probably think it's sweet if you go out and buy Bejeweled for her, too.

Re:But I've already got a PortaPam!

[outsider007]

Your wife will probably think it's sweet if you go out and buy Bejeweled for her, too.

if she's been playing a pirate copy you can tell everyone you 'made an honest woman out of her' :)

Re:But I've already got a PortaPam!

[unicron]

No, warezing is fine. It's when you think that warezing is a political statement that it gets lame and unacceptable. Here's an example:

"I pirated this because I'd rather spend my 20 bucks on hookers and blow." --ACCEPTABLE

"I pirated this because everytime I download warez the DMCA becomes my bitch!" --UNACCEPTABLE(But this mode of thinking automatically grants superuser access on /.)

Re:But I've already got a PortaPam!

[Bilestoad]

But when you spend more than two hours searching for a $20 program you tell the world:

"Hey, my time is worth less than $10 an hour! Does anyone know of an opening at Jiffy Lube?" --PITIABLE

Re:But I've already got a PortaPam!

[Tsuzuki]

Warezing because you're so poooor isn't much better... how'd you get the computer? The broadband connection? Housing?

$20 wouldn't get you far for hookers and/or blow anyway. ;P

Re:But I've already got a PortaPam!

[Kredal]

Not if you do it at work, and are getting paid for looking for the warez! (:

Re:But I've already got a PortaPam!

[base3]

What I'm wondering is where exactly you're getting hookers and blow for $20/day. And weren't you one of the "copyright infringement is theft" whiners in another story awhile back?

Re:But I've already got a PortaPam!

[twalk]

If you can afford $200+ for a Palm, then you can afford $15 for a game. You just know that you can get away with not paying for the software is all.

NO Palm game developers are getting rich. The very best are just breaking even. Nearly all of them are losing money.

And here you feel that it's acceptable to not pay despite the fact that these developers really need the money in order to pay the bills.

So can I sell SimCity if I don't want it? (nt)

[RallyNick]

nt

Re:So can I sell SimCity if I don't want it? (nt)

[RallyNick]

As far as I know if a company makes false claims about their product to gain customers from competitors, they not only have to give my money back but pay non-zero punitive damages as well so they have an incentive not to do it again.

Comparison

[SiMac]

For the lazy, the comparison between the Prism (real 16-bit) and m130 can be found here.

However, by inspecting this picture, i think that Palm may actually be trying to cover up the fact that there are only 58000-some colors using the dithering technique and that in real life there are actually only 4096 colors.

Re:Comparison

[xactoguy]

Man... You actually expected a Geocities site to stand up to a mad rush of slashdotting??? Whoo boy, I wonder how long it took before that site exceeded the data-transfer amount?

Re:Comparison

[ahaning]

It's done.. I clicked the link and all I got was a note that the site had exceeded it's alloted transfer and this ad for Yahoo!'s porn-ad service. ("Yahoo! is exposing her to millions."? Uhhh.. okay.)

what!?

[Anonymous Coward]

you mean those registration cards actually might have purpose!?!?

Actually,

[shepd]

Its not even 58k colours for real. That's simluated from the hardware limited 12-bit (4k) colour depth. (Or at least that's what TechTV sez).

Palm users were really ripped off, IMHO.

Re:Actually,

[RaboKrabekian]

Last time I checked, a Palm screen was hardly CRT technology. That would make it a bit less portable.

Re:Actually,

[iCEBaLM]

I know a few people who own the M130, the screens are fine. The color is great. This was a complete non-issue but Palm is doing the right thing by standing by their words anyways.

How can you be ripped off if the company offers to buy the damn thing back from you if you don't like it?

-- iCEBaLM

Re:Actually,

[shepd]

Okay, I know quite a bit about how LCDs and CRTs use separate colour pixes and simulate the real colour.

I learned a LOT about the lies of LCD resolutions when I was shopping for a VR/Television headset (that I never bought because _no one_ had them for show in the Kitchener-Waterloo area, except for a barely pre-beta-production pair at the Sony shop that were priced exorbitantly).

Non-consumer LCD specs are rated at their monochrome specifications, that is to say they are rated at 3x their resolution with no colour guarantees (because that's the job of the controller, not the LCD).

Consumer LCD specs are rated at their full colour specifications whenever they mention "colours" in the same line. For example, "Displays 160x160 resolution at 16-bit colour". However, if colours are *not* listed on the same line, its fair game to say its a 480x180 pixel display, _but_ on a fully fledged consumer device one would have to back that up with OS support for a fake monochome display using the separate colour pixels (which Palm does _not_ have).

Now, as far as raw CRTs and raw LCDs actually having bit depths associated with them, this is false. As the raw pixels on LCDs and the minimum size points formed by the shadow mask CRTs are purely analog in nature, you cannot state a bit depth for them. You are only limited by what the controller can do for an LCD, and with a CRT you are unlimited (unless the designer of the controller was on LSD at the time).

Anyways, since you seem so interested in learning how all this works (as you asked me to look it up for you, but I don't need to, since I learned all this in the few EET courses I passed handily) I'll explain why all this is to you. What a nice guy I am, huh?

Okay, lets start with CRTs. These are complicated little beasties when you get into colour, so lets start with monochrome.

The tube you are looking at right now is evacuated of all air. In the rear of it is a heating element, which causes a material in front of it to emit electrons. The amount of electrons emitted is controlled by a control grid in front of this material. This is what allows us to control the intensity, or brightness of the beam. This is controlled through voltage, and therefore is completely analog unless you choose to hook it up to a digital controller. After the beam is attenuated by the control grid, it then passes by "yokes", or electromagnetic coils in a standard CRT, or for an oscilloscope CRT, these are deflection plates. In either case, a voltage is applied to these. A higher voltage moves the electon beam away from that yoke/plate, however a lower voltage does not move it closer (this is why a TV requires at least a 4-way yoke, or 4 deflection plates). Moving the beam causes a spot on the phospor covered, lead impregnated part of the screen you see to light up (it actually excites the phosphor and causes it to emit light waves and x-rays rather than electrons). X-Rays (which are mostly of the soft form anyways) are curtailed by the lead, and the lead is grouned to remove the resulting electrical charge caused by all this electronic conversion away from the screen. Not to mention it keeps the EXTREMELY high voltage used called the "screen" from killing you. Beats me what this was about, nobody ever explained it (could that just be part of why I failed out of EET? :-)

Now we can see if this beam is moved about the screen it will create points of light all over. P22 phospor (which is what is used in starndard computer monitors) does not instantly stop emitting light when charged and, knowing this, we can use it to our advantage and move the beam quickly enough about the screen to keep the entire screen bright.

Now, modulating the yoke and control grid we can produce a picture. NTSC combines all this into one signal (bad). Fortunately, VGA does not, and is still completely analog (and could display google bit colour, if you so desired). VGA uses separate vertical and horizontal deflection signals, and also has separate voltage controls for the different colours red, green and blue (which we're about to get to).

A shadow mask placed behind the phospor on a screen allows the three beams integrated into a colour monitor to selectively hit various coloured phosphors on a computer screen. Basically, I really don't want to go into this anymore, because again, computer monitors are NOT my expertise.

So, as you can see, I've proven CRTs are purely analog, and therefore can display an infinite range of colour (disproving your bit-based theory of CRT colour).

Now to disprove your bit-based theory of LCDs.

LCDs are far more simple than CRTs. A fluid inside an LCD can be polarized at various angles with an applied voltage. The voltage directly controls the angle, and is completely analog. A polarized lense is placed either behind or infront of the LCD. A standard LCD (such as the one in a digital watch) has a mirror behind it which light bounces from when it strikes the LCD.

When a 90 degree twist is applied to the LCD is causes the display to be totally black, because it is a completely perpendicular angle to the polarized glass in front or behind it. If enough voltage to cause a 39.37837 degree twist to be placed on the LCD element, it will show up as a shade of grey, and that shade of grey is different than one at 39.28374 degrees.

When a Red, Green, and Blue colour filter is applied to these elements, you get a colour display, at the cost of requiring three times as many pixels. The display is still analog, and can display an infinite amount of colours, only limited by the controller attached to it.

HTH!

Re:Actually,

[shepd]

>How can you be ripped off if the company offers to buy the damn thing back from you if you don't like it?

Cause they wasted my time should I have bought it and now I have to waste more time returning it?

But hey, if they will pay the entire purchase price, I'd call it even by simply not buying misadvertised Palm produts (ie: Never again).

Fortunately I don't own a Palm. I hate stylus based entry methods. :)

But my mind might change on that when I see all sorts of super-cheap refurbished m130s on the market. ;)

Re:Actually,

[be-fan]

Not to mention it keeps the EXTREMELY high voltage used called the "screen" from killing you.
>>>>>>>>
Oh come on. We played with open monitors lots of times in high-school. They're like 20-30 thousand volts. It'll give you a good shock, but it won't kill you. The real danger is that it arcs several inches, so "insulated" wires look safer than they are.

Re:Actually,

[Duckz]

It's an unfortunate act that Palm is probably really doing this out of the kindness of their legal department, not the way ethics lead them to think.

--
Todd

Re:Actually,

[shepd]

That's it. Slashdot screwed up again, and I just don't feel like writing my comment again.

[PSA mode on]

So I'll keep it short: Kids who just read the above comment, remember, just because the shock doesn't kill you doesn't mean the soldering iron you get thrown on top of won't. Always talk to a TV repair guy (or someone else with experience) before you open up your TV.

[PSA mode off]

Re:Actually,

[ewhac]

Well, you took the EE course and I didn't, but your explanation of LCD technology is not what I learned when I was trying to write display/graphics drivers for BeOS.

As I was led to understand, the individual elements in an LCD display are bi-stable: on or off. Shades of grey are obtained via two primary methods, both performed by the controller: dithering and duty-cycle modulation.

Dithering is what you'd expect, except that the dither pattern changes every field. There are usually four or eight dither patterns through which the controller cycles, thereby avoiding any static pattern artifacts appearing on the display.

Duty-cycle modulation is where you turn the cell on for 50% of the time, then turn it off for the other 50% to obtain a 50% grey. For a 75% grey, you turn it on for 75% of the time, and off for 25%. The rate at which you turn the cell on and off is too fast for the eye to see directly (and LCDs smear out rapid changes, anyway).

And, just to make things harder, every panel model is different. Each panel has different physical characteristics, signalling requirements, etc. Thus, you can't, for example, pull a Citizen panel out of a laptop and replace it with a Matsushita. Even assuming the electrical connections were identical (they're not), the difference in physical response characteristics would cause the image to be anywhere between ugly and invisible. This was one reason why flat panel support was so difficult under BeOS. Each panel model needs the controller software to be hand-tweaked to get the optimum quality image out of it. Exact duty cycle timings need to be experimented with; different dither patterns and pattern sequences need to be explored until the image looks best. All this hand-tooled knowledge is hard-coded in the laptop BIOS. Since BeOS couldn't call the BIOS, we had to punch the registers and hope. Much of the time it worked. But on those occasions where it didn't, we were hosed.

Anyway, that's what I learned. But, again, since I never took an EE course, I'm probably wrong.

Schwab

Re:Actually,

[shepd]

>Anyway, that's what I learned. But, again, since I never took an EE course, I'm probably wrong.

No, you're not too far off, and I did say I flunked out, and there's good reason for that.

It's really not been my week. But I'll make up for my paragrah of less-than-prime information by providing a page of real information!

Anyways, unfortunately, since I haven't had a lot of real-world LCD experience, I can only tell you this -- LCDs are probably AC PWM (or duty cycle modulated) because LCDs last longer under an AC charge, and it is just easier to use PWM rather than a direct change in AC voltage due to the minute changes in AC voltage required to change contrast in an LCD without using PWM.

[Not that you can't change an AC voltage, of course...]

Basically you can either design a really stable circuit that can control the small amount of voltage change necessary to keep the LCD in its middle state, or you can fake it by turning it fully on and fully off fast enough that it appears grey. PWM is used all the time to save a few bucks, and there's no problem with it, except that you trade the effort involved in making a good circuit with the effort involved in finding the "magic numbers" that keep the LCD in the grey zone.

Hey

[mindstrm]

It doesn't even display the 50,000 color number the claim.. it is 12 bit color.

They claim that by 'color mixing' you can get more colors..

Re:Hey

[spongman]

(not trying to be palm advocte here, but...) that's technically true. most displays today (including regular CRTs) can only display 768 colors (256 red, 256 green and 256 blue), it's the 'merging' of those colors that gives the 2^24 combinations. it all comes down to what constitues 'pixel'.

I guess the real problem is that it can't display 50K+ colors at the advertised resolution, since it needs to use several real pixels to make a high-color pixel.

A pathalogical example: a 1024x768x24-bit display can display 1024x768x24 or 1x1x(the total number of different permutations of 24-bit pixels on a 1024x768 display). of course, you'd have to look at the 1x1 display from a long way off for the dithering take effect.

Re:Hey

[unicron]

I can make over 16 million colors using only red, green, and blue, so I'd say they're telling the truth.

Re:Hey

[Ark42]

I bet this gets even more confusing when you look at LCD's where 4 "pixels" can be, say, represented by 2 red areas, 2 green areas, and 1 blue area.

On the other hand, this reminds me of a old ModeX resolution trick for VGA-only boards. You could actually set the screen to 320x600 to simulate 18-bit color at 320x200. It works very well, using every 3rd line as R, G, B with 6 bits per color (64 shades of each, leaving you with 64 unused palette entries)

Re:Hey

[martyn s]

Umm, no. Pixels are made up of subpixels. A red subpixel, a blue one and a green one. All pixels are equal, therefore you can't have a red pixel and a blue and a green one.

Anyway, if you're telling me that my monitor really has 3 times as many pixels as I thought, then yes, you're right. Dithering gets more and more effective the higher the resolution is. But the fact is, the standard is not to call subpixels pixels. They are subpixels. If they really considered them pixels, then they should've advertised that they have 76,800 pixels and not 25,600 (160x160). But according to standard definitions of "pixel" and "resolution" they were being fraudulent.

It's like saying, yeah, this car has 600 horsepower. But our definition of horse power is that 4HP = 1 of what most people call horsepower. We have very strong horses.

Re:Hey

[Osty]

I can make over 16 million colors using only red, green, and blue, so I'd say they're telling the truth.

Uh ... no. You can make over 16 million colors for a single pixel using varying levels of red, green, and blue in that single pixel (16777216 colors, to be exact, with a 24-bit display, giving 8 bits per red, green, and blue, or 256 shades of each merged together using the color properties of light to blend a new color). Palm has 12bpp to work with for a single pixel, or 4 bits per red, green, and blue. That's 16 shades of each, for a combinatorial total of 4096 different, unique colors. Their "blending" involves dithering (if I have a block of four pixels, and set the top left and bottom right to blue, and the top right and bottom left to white, then from a far enough distance, it looks like I have a blue that's 50% lighter than normal blue ...), or using various sub-pixel techniques (if I want a brighter red, I could adjust the red subpixels next to the pixel I'm dealing with and it will look brighter, but it will also be blurrier and could sacrifice the colors in the adjacent pixels), and such (I don't know what else they could do, really). In other words, your "mixing" of red, green, and blue is different than the "mixing" they're doing.

Re:Hey

[RedWizzard]

Well if you want to redefine the definition of pixel, then fine. A CRT driven at 32bit color can manage only 768 colors measured at the subpixel level. And Palm's M130 can only manage 48 and not the 128 they claimed. Which is the point - they lied about the color capabilities and now they're still distorting the truth so they don't look quite so bad. Offering refunds for 58K colors instead of 65K sounds like Palm are great. Refunds for 4K colors instead of 65K sounds justified, nothing more.

Re:Hey

[vadim_t]

Heh, you mean you have really weak horses if you need four times more of them to get the same power.

Re:Hey

[martyn s]

LOL, maybe. I had to think about it for like five minutes; I guess I got it wrong. :) Thanks!

Re:Hey

[ShoeHead]

You must be too old for to own a TI-89.

TI calculators are, afaik, just black and white. however, games developers are able to support at least 7 colors by sacrificing refresh rate for color blending. So, you get grayscales. Sounds like what they're doing. If you can't tell the difference, why make a fuss? I don't see the point in having more than 8 bit color anyways on a stinking PDA.

Re:Hey

[Webmonger]

In the first place, CRTs are analog, so they aren't limited in the number of colours they display. Or rather, they are limited, but only if you start counting each individual electron. That's right: a CRT can display far more than 16.7M colours using only, say, its green electron gun.

The 256-bit-per-channel limitation you describe is in the video adapter hardware, not in the monitor. And video adapters address pixels, not subpixels.

CRTs don't even have subpixels, because subpixels are addressable, and the red/green/blue subcomponents of CRT display (phosphor dots) are not addressable.

So the monitor supports an infinite number of colours. The video card supports 16.7M colours per pixel.

Yes, there is colour mixing going on. No one wants to see 16.7M shades of spectral green. Shove a magnifying glass up against your monitor, and you'll see those red, blue and green phosphor dots.

But the same thing happens with colour photographs, and printing, and pretty well anything that uses a pigment to produce different shades of colour. Everyone agrees that when the mixing below addressable resolution, it's called "a colour", and when mixing at addresable resolution, it's called "a dither pattern".

I have no problem with Palm's original mistake. They happen. But Palm's way of dealing with it has been absolutely atrocious. If they had originally advertised the device as supporting 937,936 colours, they might be justified in claiming its true colour depth was 58,621.

But no one advertises a 16-bit display as supporting 937,936 colours, because it's nonsense. The only reason Palm cares about these "colour mixing" numbers is because Palm's trying to spin this as a 10% reduction in colour depth, instead of a 94% reduction.

That's the "real problem", IMHO.

Class action lawsuit maddness

[idleprocess]

Its most recent decision also follows a class-action lawsuit filed last week in California's Superior Court in Santa Clara County.

I know it's a little off-topic but regardless of how Palm decided to handle this situation, we should all be glad that a class action lawsuit wasn't filed. In Madison County IL. there is a group called ILAW (Illinois Lawyer Abuse Watch (I think))investigating class action lawsuits and some of their findings are scary.

Verizon went through a class action lawsuit and all the participants were awarded some trivial $20 refund, or some voucher for a free month of service while the lawyers raked in millions of dollars.

These 'millions' get written off by the company and get passed to us. Not to say all Class-Action lawsuits are bad, but some are down right scary.

I know off-topic a tad. Oh well.

problem with the story

[JeffSh]

In the last slashdot article about this it was plainly stated that the difference is actually 4-bit (16 bit advertised, 12 bit really) which in actuality means the difference between 60k some odd colors and 4096 colors...

Thats ok...

[dr_dank]

I'll just wait for them to send me the extra 6,915 colors in the mail.

SimCity really worth it?

[Arcaeris]

Well, if you felt like you were getting hosed, this deal is for you. Everyone wins, even those who didn't care.

Acquiring SimCity for PDA without this deal is rather ridiculous. $30 for a game that you can fit on 1 floppy, compared to a regular massive PC game is sad. Sure it's smaller and you can play in class and stuff, but why not just have a Game Boy Advance? Cheaper, and you're not fooling anyone anyway.

Since I am RED-Green Colorblind

[Real World Stuff]

I can only relate to this vicariously through you. Hell, you think Twister is tough for you! "What do you mean left foot red?

Forget all those Colors...I want mine RUGGEDIZED!

[Real World Stuff]

Protected from the elements and able to withstand 4 foot drops drops is what I am looking for! Wher the Hell do you want to go today!

Wow....

[autopr0n]

I'd be supprised if there was any actual, noticable diffrence between 64k and 58,261.

I mean, there are only 25k pixles on the thing.

Never mind that

[Salsaman]

I am still employing Cowboy Neal with crayons and a sketch pad. I thought everybody else was too...

Free SimCity!!! Awesome!!!

[Andre060]

Unless you live in Greece... ;-)

Strobing can work

[autopr0n]

IIRC, the original game boy used black and white strobing to create the four color effect. I've seen calculator programs (for the ti-8x and 9x calcs) that could display color images as well.

Of course, the contrast wasn't as good or anything. It would be interesting to see comparisons between the two. Someone posted a link, but it was to geocities, and obviously it's dead now.

Does anyone have the details on how this supposed color increasing worked? I think it would be intresting to see.

So is this going to become the norm?

[guttentag]

1. Advertise that your product offers much more than it really does.
2. If anyone complains about your false advertising (which is against the law), wait until after the product has secured its place in the market (and in people's homes/offices) before admitting anything.
3. Offer a full refund for the 12 people who would actually rather have their money back than live with their underperforming machines. Placate the rest with a downloadable version of a software product that's over a decade old (after all, the company's only cost-per-download is for the used bandwidth... it's not like they're giving away physical items)
4. Result:
   • the 12 people who knew they were ripped off shut up because get their money back
   • the FTC will never get involved over false advertising charges
   • the company still sells (number of units that would have been sold if its claims had been true - 12) units
   • the vast majority of consumers think they got something for nothing (software) and laud the company
5. Repeat with next product release.

Looks like a great deal for those who don't care about the bit depth of their PDA, and a way out for those who do my a$$... looks like a great marketing/disinformation strategy for Palm.

And no, this is not "the way business is done," this is "false advertising." Unfortunately, false advertising is only against the law if people complain.

Re:So is this going to become the norm?

[Joe Tie.]

It's still one up on microsoft. Their line of working hard on updating their PPC ebook reader to allow reading comercial ebooks changed rather quickly to "Whoops, sorry. Hey you can get it if you buy a new pocketPC though!".

Re:So is this going to become the norm?

[DansnBear]

Quick!!!! I'm currently looking for investors to back the patenting of this business model so we can sue palm For patent violation. Interested parties can contact me at the above address. Thank ou for your time

Not quite sixteen bits...

[chrisgagne]

Ok, so its 15.83913 bit color. That's only .16087 bits short of what they advertised. Now if you convert bits to dollars (2 bits = 25 cents) they've only shorted you 2 cents. They're giving you a game that's worth $29.99 retail. So what are you complaining about?

Shave and a hair cut, two bits.

http://www.chrisgagne.com/hosting

Re:Not quite sixteen bits...

[3583 Bytes Free]

If you read the previous /. article, you'll see that the m130 can only display 4096 colors (12 bit). However, they play some dithering tricks to claim the 58621 colors. Of course, if you did the same thing with real 16 bit color, you could probably claim millions of colors.

The bottom line is that they are quite a ways off. Now, they are saying, "well, we didn't quite give you the full 64k colors, and we're offering a refund even though we're just a few percent off." They look like they are going out of their way for people who wouldn't notice the difference between 65536 and 58621. But in fact they are trying to make sure no one calls them on the fact that they lied - big time, and can only deliver 4096 colors.

I'm feeling some hostility here

[jandrese]

Even though the Palm can only display 4096 colors without resorting to ugly hacks (like pixel flickering), I don't see what the big deal is.

Ok, they lied in their marketing, that's bad. But they seem to be trying to do the honorable thing here. If the color depth is that important do you just get the refund and buy yourself a Handspring.

But lets work the numbers here: A 160x160 pixel screen has 25600 pixels total. The 12 bits per pixel can only display 4096 unique colors. This means that in the worst case scenario, every color will have to be spread across 6.25 pixels. This doesn't seem all that bad to me. In fact it sounds like just the sort of design tradeoff I might have made. Going all the way up to 65536 unique colors is kind of a waste since you'll never be able to get all of those on the screen at once.

Of course Palm should have advertised it as a 12bit screen right from the start, but I'm not ready to hang them out to dry for this. On the contrary, offering Sim City (which is still a fine game, despite what the vitriol filled posts on here might say) seems like a nice gesture to me. Palm certainly could have done worse.

Does anybody remember IOmega and the Click of Death? Years in lawsuits that just make the scum sucking lawyers richer and richer and what do we get? A coupon from IOmega for some paltry sum off of our next purchase of an IOmega product, long after most of us had swarn off IOmega forever. Would you guys have preferred that?

Re:I'm feeling some hostility here

[topham]

The 12bit colour isn't a PALLET of 16.7 million (or 65K) with only 4096 displayed at a time.

It's only 4096 colours total. You don't get to choose which colours are in the pallet.

You get 16 shades of red, 16 shades of green and 16 shades of blue. You get to mix them as well, but thats it.

So, yeah, even though there are only 25,600 pixels on the screen you could still display an image, via scrolling with the full 65K colours. Now your left we fudge tricks to get the same colour range.

I think this move by Palm is a good move though.
Many people are probably more than happy with the display.

Re:I'm feeling some hostility here

[be-fan]

This means that in the worst case scenario, every color will have to be spread across 6.25 pixels.
>>>>>>>>
I have no idea what kind of LSD induced images you're looking at, but the real problem is full-color images use lots of one shade, not a little bit of lots of shades. Thus, worst case, you have something like a gradient, with the 16! values of each shade spread out over 10 pixels. Not a good tradeoff for a machine that's supposed to view photos and whatnot. And p0rn will look terrible with all the skin-tones all stratified like that!

I wonder what palm is going to achieve with this..

[theridersofrohan]

First a points:
Face it: The palm m130 is a cheap 160x160 pixels handheld. 16bit colout on a 160x160 cheap screen is _not_ that much different than a 12bit colour screen. What are you going to do? Run photoshop on your palm or show your vacation pictures to others on a frikin 160x160 screen?

That being said, I own an m130 and have been insanely pleased with it. However, this offer puts me into temptation.. Do i return the m130 and use the money (+£100) to buy the much sexier m515? I might be a righteous person and not do it. But others?

This might be an economical disaster for palm...

Comment removed

[account_deleted]

Comment removed based on user account deletion

58,621 colors?

[Alsee]

If they want to say it has 58,621 colors then they have to say the screen isn't 160x160 anymore, it's 80x160 or 80x80. The only way to get the 58,621 colors is by DITHERING which kills your resolution.

-

Re:58,621 colors?

[rabidcow]

In following the time-honored /. tradition, I have not read the article, but there is a way to simulate more colors than you actually have besides dithering. What you do is to alternate rapidly between two slightly different images. The pixels will appear to be a mixture of them, so you can have halfway in between or whatever. This does not reduce your resulution, but it does introduce some flickering.

Re:58,621 colors?

[Alsee]

Ok, but if they use that method then they have to annoce that the screen refresh rate is only half (or worse) as high as advertized.

-

Re:58,621 colors?

[Junta]

Used a palm? The displays are so low end that redraws should be avoided like the plague. Full motion rapid image changes would just give a blurry mess, the ghosting on those displays is too bad to acheive the described effect.

Re:58,621 colors?

[rabidcow]

I dunno about that, blurring is what you want with this technique. You *want* the two different colors at each pixel to blur together.

But no, I haven't used one, so it might still not work.

Re:58,621 colors?

[twalk]

Palm never advertised the screen refresh rate. And it had already be noted in many reviews that the refresh was terrible.

Why do you assume Palm is lying about the 58k?

[GlenRaphael]

LCDs have a slow refresh rate so it is possible to display more colors by turning the pixels on and off at a high rate. You need to combine this technique with dithering to avoid the appearance of flicker, so the fact that they mention dithering is NOT a cop-out. It doesn't mean they aren't displaying 58,000 possible color values with 1-pixel resolution.

(I used the same technique to get 5-level greyscale on a 1-bit (black & white) Newton in a demo program called Time Domain Grey.)

Re:Why do you assume Palm is lying about the 58k?

[xintegerx]

Yes there is a difference...

The difference is that you can the end result is different between true 64,000 and fake 58,000 colors made out of 4,000.

See http://www.geocities.com/an0nym0vs when it's not slashdotted.

Re:Why do you assume Palm is lying about the 58k?

[xswl0931]

So by your logic, we can call any B&W display (meaning truly 1 bit) greyscale because I can dither or flash them on and off? The hardware can only display 12-bit (4096) color, so they should just advertise it as such.

Re:Why do you assume Palm is lying about the 58k?

[GlenRaphael]

So by your logic, we can call any B&W display (meaning truly 1 bit) greyscale because I can dither or flash them on and off?

If you can get decent greyscale out of it, why not? What matters is the system has sufficient power to display a decent greyscale and the manufacturer provides an API whereby application developers can use it. If developers can call an API to draw a 50% grey pixel at location (x,y) and that's what happens, I call that a device with a greyscale screen. Regardless of what voodoo the OS is doing to get that effect out of the hardware.

(In the case of my Newton application the hires mode I created was hard on the battery and only worked really well in a small window, but with a faster CPU and support from the manufacturer, we could easily have called it a device with a greyscale display.)

The hardware can only display 12-bit (4096) color, so they should just advertise it as such.

What you can address at the hardware level is not the last word on the subject. Or should I have ignored what users saw on their screen and advertised Time Domain Grey as displaying 1-bit color? :-)

Re:Why do you assume Palm is lying about the 58k?

[dfung]

>> So by your logic, we can call any B&W display
>> (meaning truly 1 bit) greyscale because I can
>> dither or flash them on and off?

> Are you some kind of a troll? Those two
> techniques are exactly how greyscale displays > work.

Yow, you're totally wrong on how a monochrome display works here... When you use a grayscale display (or a color one, which works the same way with three guns) the monitor hardware modulates the strength of the electron beam proportionally to the input voltage. The monitor makes grayscale because it was designed to. You could also make a monochrome-only display that doesn't respond proportionally, instead it has a set threshold and anything higher than that is on, anything lower is off.

Inside your computer, the video is all digital. There's a frame buffer that holds a multi-bit value of high bright you'd like the screen to be there. The last step before the signal leaves the computer is that it's run through a digital-to-analog converter that generates the appropriate voltage level out.

In a laptop, there a true digital->screen conversion, but it still looks at that multi-bit value, then transmits that value to the display's drive electronics.

What the original poster in this thread was talking about was something different. He knows that you can light a dot on the screen up, and if you turn it off, it takes a while (more than a refresh cycle) to dim to, uh, white in this case. So, he takes advantage of that hysteresis and the fact that the display's refresh rate much faster than the LCD responds. If he wants a dot to be black (all on), he writes "on" on each refresh cycle. If he wants it to be gray, then he writes "on" one cycle and "off" the rest. Instead of antialiasing in space, he does it it time, which is pretty cool, but probably has the effect of very uneven rendition of the levels (the decay of the LCD won't be linear and you probably can't tweak it much). And, obviously, it will be much less responsive to UI interaction, although you might turn this mode on only when looking at a picture.

I doubt that Palm is doing that, but if the original poster still has his code, I don't think he'll ever get any more money for it than from Palm RIGHT NOW!.

Re:giggle

[Qrlx]

So, which elephant to I have to copy in order to get, uh, punished??

Signed,
CmdrTaco

Re:I wonder what palm is going to achieve with thi

[Qrlx]

Myabe you should get a Sony Clie PEG-SJ30. It really has a nice screen, 320 x 320, with proper backlighting and everything. Here in USA, the Palm m130 is like $250 and the Sony PEG-SJ30 is $300. Definitely worth the extra 50 bucks if you ask me.

Tech plummets in value so fast, I would return the m130 and reinvest in something newer.

You left out the time-domain part

[GlenRaphael]

If you have a 1 bit display (just black and white), if you make every other pixel black, and every other pixel white, it will give the appearance of being gray (especially at higher resolutions). That is what dithering is. This is opposed to showing a pixel that is actually gray (half black half white, that is, each sub pixel [red, blue, green] on equal intensity, at half intensity).

So far so good. But suppose you generate TWO complementary frames of dithered 50% grey. In one frame the first pixel is white, in the other it is black. If "O" is white and "X" is black your two frames look like this:

FRAME #1:
OXOX
XOXO
OXOX

FRAME #2:
XOXO
OXOX
XOXO

Now, alternate displaying frames #1 and #2 in rapid succession on an LCD display with a slow decay rate. The resulting image looks like this:

COMBINED FRAME:
****
****
****

Where "*" looks like a pixel that is 50% grey. Not dithered grey, real grey.

Re:You left out the time-domain part

[xswl0931]

It's not "real" grey, it's dithered flashing grey. I need to set my CRT monitor to 85Hz not to see flickering, for a slow lcd refresh, I'm sure this would bother me.

Re:You left out the time-domain part

[GlenRaphael]

It's not "real" grey, it's dithered flashing grey

If the alternate frames were all black and all white you'd easily see the flashing. That's why the dithering part is required, to break up the patterns a bit, so your brain can't see the solid patterns being painted to the screen.

for a slow lcd refresh, I'm sure this would bother me.

It's the slow decay rate that makes the technique possible. On many LCDs if you paint a color it fades out slowly rather than instantly. The slower this decay rate is relative to the refresh rate, the more color range you should be able to sneak in with this sort of technique. Hmm, I should go do some experiments to see how well this works on current Palm hardware...

palm woes

[sideone]

damn, i wonder if i can get them to replace my old palm pro. still works, but very little memory. Well i guess thats what i get for buying a USR palm! been 5 years with it, and it still works, so i guess memory isnt that big of a deal. Does anyone know any memory hacks for the oldschool palms?

4096 is not 58,621

[frovingslosh]

The m130 was originally advertised as supporting 65,536 colors when in actuality it can only display 58,621.

Lets be accurate here. It can only display 4096 colors. It's a 12 bit color display, not 16. However Palm marketing wants to twist things, it does not serve the user to repeat marketing hype. They sold this thing as a 16 bit display and it was a 12 bit display. Matters a lot if you want to view photos or color images, and that's the reason many paid for a color toy. The problem is more serious than the "only 58,621 colors as contrasted to 64k" marketing hype.

Business ethic

[Banjonardo]

In this business ethic-bashing time we've come to, it's nice to see a company that actually cares about its customers and treats them fairly.

zerg

[Lord Omlette]

Your company was once at the top of its field. Now it's being crushed by competition and you've just been forced to admit to an incredibly stupid blunder, and apologizing is going to be costly as hell.

How do you pull your company out of its rut?

For the record

[El]

I beleive the lie that got Palm in trouble was that it claimed the M130 had "16 bit color" when in fact it has 12 bit color. Dithering has never counted in the past when discussing the number of bitplanes, and it shouldn't really count when discussing the number of colors -- especially on a display with so few pixels. A CGA display is STILL 2-bit color, even if you dither the pixels!

Well, what you say is true..

[mindstrm]

but only if we re-define what is commonly understood on LCD screens to be a 'pixel'.

My 1600x1200 laptop screen has 1600 red, 1600 green, and 1600 blue sub-pixels across.We don't call it a 4800x1200 screen, though.

So when they say it can display 16 bit color on the color LCD screen, the consumer has a right to assume that means they are using a 656 display... six bits for red, five for green, and 6 for blue (or whatever it is..). saying that you can use more pixels to get more color.. that's just bad advertising.

Fuss?

[mindstrm]

The fuss is they told us it had 16 bit color, and it only has 12 bit color.

That's like selling you a car saying it has 300HP, but in fact, only has 120.
what is your argument there, that 300HP is too much anyway, and everyone should be happy?

No. It's fraud, and it's illegal, and they are doing the right thing by offering refunds.

Re:Fuss?

[Forkenhoppen]

That's like selling you a car saying it has 300HP, but in fact, only has 120.

No, it's more akin to saying it has 300HP, when they're really selling a 120HP engine that's been modified and tuned (but mostly tuned) to run at 280HP.

Re:Fuss?

[curiosity]

No, because in that case the car still has 280 horsepower. This is like saying it accelerates like a 280 horsepower car when you drive it down hill.

If they were being intellectually honest, they would have advertised it in the first place as being capable of more than 16 bit color, since you can dither 16 bits just as well as 12. It's the fact that they changed the measuring criteria before they "restated" the capabilities. Duh.

Re:My only question...

[El]

Dithering a color with itself doesn't produce a different color. That eliminates 4096 colors right there. Dithering most colors with black or white doesn't produce a color that couldn't be got from a darker or brighter shade of that color. That gets you a few more colors (about 2744?). Not sure how they got that exact number.

they're doing the right thing

[brad3378]

I can respect a company that can admit it screwed up.
This is going to cost them tons of money, but unlike the actions other companies, Palm may have just earned my trust.

Re:Actually, No, they don't.

[mindstrm]

The Palm device has 12 bit color. Each color sub-pixel (red, green, blue) has 4 bits each, that's 16 shades of color each.

Leaving you with... 4096 colors per pixel.

To claim it does more is to stretch the truth. It is common to accept that when we talk about the # of bits of color, we mean per actual pixel, not sub-pixel.

To group multiple pixels together and then claim it was actually true is just BS

Re:I wonder what palm is going to achieve with thi

[trapvector]

I have never been put into temptation by anything.

That aside, I love my m130. Sure, I was a little skeptical when Palm said 64,000 colors... and that made me less surprised when I turned it on and saw 12-bit color... but I never upgraded for the colors. I switched from my palm IIIc because I needed the expandability and liked the form factor of the m130 so much more. The m515 may be sexier, but it's also a bit more expensive, and I'm already pleased beyond pleased.

Even more so now that I get SimCity for free. Palm rules.

Offtopic

[Osty]

Mozilla, Internet Explorer popup killer (for all flavors of windows).

NoPopIE, proving that anything Mozilla can do, IE can do as well. As for only supporting Win2K/XP, I don't have any Win9x boxes around for testing, nor do I care to.

Intrinsic value of warezing

[0x0d0a]

I'm not sure. The knowledge or experience may have some intrinsic value -- maybe you learn something.

I'll spend much longer when fixing something figuring out *why* it broke then just "getting it working again and forgetting about it" because in the long term, this pays off. Well, at least I hope so.

I remember this being written somewhere in a book or magazine: "They were software engineers, the sort of people that will spend four hours calculating different trip routes to save ten minutes taking the shortest possible trip".

This may be less true for simple piracy, but if you're actually cracking the software yourself, there's some educational value to the whole process.

I'm very pleased as a Palm customer

[defile]

I purchased a refurbished Palm Vx. The device kept losing its calibration which required me to re-run the digitizer. Sometimes it wasn't reachable and required a hard reset. Called Palm, gave them my S/N, a replacement arrived in the mail.

Painless and awesome. Thanks guys.

Offtopic again

[Osty]

Never claimed to be valid CSS, just valid HTML 4.0 (which it is).

Free Time != Money

[Gooba42]

I've never understood this mode of thinking. You do not get paid for 24 hours a day. You get paid for whatever time you spend doing your appointed job. Unless you're taking time away from work to do whatever it is, then there is no time:money ratio.

This mode of thinking suggests that any endeavor for which you don't get paid is a waste of your time which could be better used making money. Is watching a football game or playing a video game or spending time with a loved one worth more hourly than what you get paid at work? Does doing dishes, washing your car or cleaning the litterbox pay you hourly "what you're worth"?

If the answer to these questions is no, and your assertion of time=money is to be believed, then none of these is worth doing. The only thing worth doing under these assumptions is the best paying job you can possibly do for 24 hours a day.

Life isn't, and never will be, cost effective. Don't waste your time making money when you could be making happiness. Go warezing, play games, shower, whatever floats your boat.

The Catch.....

[lateralus_1024]

The only catch is that it's a greyscale version of SimCity.

Misleading?

[DaWorm]

Hmm..the article said Palm "misled" consumers. I believe it's an honest mistake. Come on, they're generous enough to either give registered buyers a refund or a good game.

I don't have an m130 (I use a Zaurus) btw.

How'd all the skript kiddies get that stuff?

[pediddle]

Parents?

In other news, Crayola announces release of new...

[Jafafa Hots]

Crayola today announced the release of thier new 64,000 color crayon assortment.
Upon initial examination, the new assortment looks similar to their old 64 color box, but company officials explained that the 64,000 assortment uses a new crayon technology called "color mixing" which enables users to combine several color crayons to create a unique effect.

Company officials would not comment on the recent lawsuit by Daniel LeFaviour of Dupont, MI wherein he claims to have patented the crayon colormixing technology.

Wheres the link??

[The Real Chrisjc]

I've been poking around on the palm site, and I can't find a link anywhere! I own an m130. Can anyone help me? Please? I want my free copy of Sim City.

Re:Wheres the link??

[Draoi]

Here it is. I'm a Palm m130 owner & I'm seriously considering taking them up on the offer.....

Re:I wonder what palm is going to achieve with thi

[Junta]

There are 25,600 total pixels on a 160x160 display. 12 bit color means you cannot have a scenario with each pixel being a different color. More importantly, it's not so much about simultaneous color, but the total palette. With 12 bit precision, each color has 16 levels. If there is a pure red gradient, it would look like crap, for example. The difference is non-trivial for those who intend to view images on it.

That being said, I have an m130 and don't care about the refund. I don't use my palm to view images, no matter how many colors it could display, the 160x160 restriction is too much to deal with. Even with a 320x320 display, images don't look good enough.

No Color.

[Dwedit]

It's totally bogus. You can not get any color on the TI83+. The only "colors" you get are those from cranking the contrast way up so black pixels become 'blue', and white pixels become 'brown'.

You can put the LCD screen into 'test mode' by sending a specific instruction to the LCD controller. This 'test mode' makes a random line on the screen blue, and shuts off the rest of the display. (You also have to send an END TEST MODE instruction to get it out of test mode) It takes about a second for the blue line to fade back to the normal monochrome screen.

You can't do color on the TI8x's monochrome screens, but you can get grayscale by turning pixels on and off quickly.

Re:So, you failed.

[shepd]

>So, if you were passing your courses "handily", why did you fail out?

I only passed the digital courses "handily".

>Also, I must assume that EET means electrical engineering technician, which isn't even a real electrical engineer.

Well, assume makes an ass out of you. But not me this time -- EET is Electrical Engineering Technologist, which is a combination of the lower end EE math and higher end technician work. I wasn't interested in just sitting in front of a computer designing crap 24/7, and I wasn't just interested in soldering 24/7. Technologist gives a good balance between the two. Lots of troubleshooting via redesigning circuits (rather than just finding the fault and letting the EE know about it).

>An EET is like an EE who never took any math courses and doesn't really understand what's going on.

Nope. Double wrong. But if you want to live in that dreamworld, that's fine by me.

>But, if you failed out of some 2-year EET school (DeVrye?), I feel sorry for you.

No, its a three year program -- soon to be acknowledged as a Polytechnic program (about time too).

You are wrong about a great many things. It's a little sad, actually. :-(

Comment removed

[account_deleted]

Comment removed based on user account deletion

Anyone refunded?

[auyongtc]

Hey - anyone bothered to ask for a refund and got refunded yet?

Re:Actually the photo would do it justice.

[xintegerx]

That's assuming the film mixes the flickering colors the same way that your eyes do at the same time. The photo image would then not be flickering so it could differ.

But because the web site owner bothered with the photos there MUST be some kind of a difference right?

This is the 'net though..