Phone, Tablet and Gizmo Thread #0x02

Mort Walker · Post by **Mort Walker** » 29 Apr 2013 17:56

^^^Accurate time keeping is of utmost importance to me as well. For really important clocks in industrial applications you use oven controlled precision crystal oscillators that cost a few hundred dollars.

negi · Post by **negi** » 29 Apr 2013 20:02

I have plonked close to 2000$ on automatic watches now and one of them is already broken (repair will cost me 300$ + shipping from Germany), there is no sense to this obsession with automatic watches it is just a hobby like collecting fountain pens. A Montblanc will cost you upwards of 300$ and yet it won't be as portable and rugged as a INR 20 reynolds .

Anujan · Post by **Anujan** » 29 Apr 2013 20:04

Well I disagree. Montblanc are smoother than Reynolds. They are more balanced with the right thickness. They to their primary job -- writing -- better than other SDRE options. Question is do costly watches do their primary job -- time keeping -- better than SDRE options?

negi · Post by **negi** » 29 Apr 2013 20:08

^ I should have clarified ; I was talking about fountain pens vs ball point pens . Regardless of how tfta Montblanc's iridium tipped nib is it still needs to be emptied before one boards a flight , it will not write when inclined beyond an angle unlike a ball point pen and biggest problem is every-time when someone next to you asks for a pen and if you cannot say no , your dil will go dhak-dhak when he/she puts the nib to paper (Not many know how to write with a fountain pen). My Boss once took my Cross to a meeting and lost it as if it were some use and throw paper-mate pen.

Also coming to the primary function well as far as accurately showing time is concerned Rolex and Omega models are choronometer certified i.e. within COSC standards , obviously they can't beat the Quartz/Super-Quartz but they do show time within reasonable limits for day to day use. Rolex/Omega aside from having trophy value and history (Likes of Federer and James Bond wear them, first watch on Moon was the Omega's moonwatch and Sir Edmund hillary wore a Rolex to Everest etc etc ) also serve as a piece of jewelry for men.

If one is a believer in function above form then obviously Automatics are not for him; Automatics have same charm as a pre-1983 Royal enfield , it leaks oil has a low fuel economy but it has a better thump than the new ones on streets and has a high re-sale value amongst enthusiasts.

ArmenT · Post by **ArmenT** » 29 Apr 2013 20:23

Anujan wrote:I like wearing watches. I like the analog watch face, dont care about mechanical movement inside. I have a very SDRE casio waveceptor with analog watchface which syncs with an atomic clock twice a day and can outdo any mechanical watch in terms of accuracy. It has automatic daylight savings too and can set various timezones digitally (after which it mechanically moves the watch hands). It is not a $1000 swiss "chronograph".

My usual go-to watch is a Casio calculator watch that I've had since about 1997 or so (this one stores 150 numbers and appointments, which was an upgrade from my older model that only stored 50. I found it very handy back in the days when smart phones weren't common. Casio doesn't make high capacity calculator watches any more, the new ones only store up to 50 numbers

) For swimming and other outdoorsy stuff, I usually fall back on my old Casio DW-280 that was given to me as a birthday present even more moons ago, as my calculator watch isn't waterproof and the DW-280 is waterproof to 200 meters depth. This was a sort of tough watch that Casio made before they introduced their G-shock line. The DW-280 has only had two battery replacements in about 20 years. I also have another waterproof Casio with analog hands (Casio ABX-20) behind a digital LCD face, so you can bring up or hide the digital display as needed and it will look like a regular analog watch.

Speaking of Swiss, I have a collection of Swatches and my fav. one was a red Swatch Paparazzi from a few years ago (they stopped making the model now). This one connects to the atomic clock for ultimate precision and is timezone aware (so I never had to adjust it when I got out of a flight, it simply knew where it was). In addition, it received news, stocks, traffic reports etc. (i.e. it was like a smartphone, but it was on my wrist and this was in 2004!) and also had downloadable watch faces. You could get more content on the watch with a subscription to MSN (e.g. instant messages, more sports news, more stock exchanges etc.). Only downside was that it needed charging pretty frequently (once every few days) and Microsoft MSN stopped supporting it a few years ago. I snapped the strap and haven't been able to get a replacement.

Vayutuvan · Post by **Vayutuvan** » 29 Apr 2013 20:34

That is the problem with swatch unibody

Raja Bose · Post by **Raja Bose** » 29 Apr 2013 20:42

I use the clock on my cellphone - can't get more accurate than that.

vishvak · Post by **vishvak** » 29 Apr 2013 20:56

Any idea why win7 laptops are more expansive than win8 ones?

I am having 2nd thoughts about Lenovo so I was checking out HP/Dell business config laptops and came across a Dell Vostro model that's 40k DOS basic version, 44k win8 version, 50k win7 home premium version and 58k win7 professional for same config! I think for about 2-3kRs. win7 would come bundled along.

Then I was offered Inspiron laptop 47k touchscreen win8 (not business model!) or 64k HP probook/ultrabook) touchscreen one.

Anyone can comment on this? Looks like a way to push touchscreen business model laptops with win7/win8 combo. Matches with messages from some members on this thread.

negi · Post by **negi** » 29 Apr 2013 21:03

Pretty much the same reason as selling Windows XP laptops for 150$ more than the ones with Vista when latter was released.

vishvak · Post by **vishvak** » 29 Apr 2013 21:13

There were some blog pages on Mï¢rö$Ø£T ™ is ditching Win8 enterprise.

Any rough idea if MS 7/8 some version is not to be supported in future 4year period timeframe?

I would rather learn to deal with win8( or downgrade to win7) than spend money on win7 that may not be supported any longer than win8 for 4 years.

=====
Edited to rephrase question

Anujan · Post by **Anujan** » 29 Apr 2013 22:07

Same thing happened with my montblanc. Someone borrowed it and conveniently lost it.

The best hack though is to place a montblanc rollerball refill inside a pilot Body. You have a $300 writing experience in $20 or so. Most websites ask you to use a G2 plastic body, I use a pilot pen with a heavier metal body but forget the model name.

http://www.wired.com/gadgetlab/2010/10/ ... azing-pen/

negi · Post by **negi** » 29 Apr 2013 22:25

^ That sucks; I still have my Hero Fountainpen with which I wrote my XIIth class board exams, it served me well through Engineering until I landed up in the trenches where carrying a fountainpen is considered a sin. I hope to pass it along to my kid.

Vayutuvan · Post by **Vayutuvan** » 30 Apr 2013 00:05

vishvak wrote:Anyone can comment on this? Looks like a way to push touchscreen business model laptops with win7/win8 combo. Matches with messages from some members on this thread.

One possible way out is to buy the cheapest Linux compatible with no OS/DOS and load up with all Windows == Linux pieces.

Prem · Post by **Prem** » 30 Apr 2013 00:06

http://www.electronista.com/articles/13 ... .messages/
Indian company developing world's first Braille smartphone

Researchers at India's Centre for Innovation Incubation and Entrepreneurship are working on a smartphone that will leverage novel technologies in order to allow blind users to receive messages in Braille. The device has yet to be officially named, but it is expected to hit the market late this year.The Braille smartphone, Times of India reports, uses an innovative "touch screen," which can raise and lower a grid of pins in order to form Braille shapes and characters when an SMS or email is received. Shape Memory Alloy technology allows each pin to remember its original flat shape even as it expands, allowing it to contract automatically when necessary. Sumit Dagar, the designer behind the device, says that the initial response to it has been "immense." "It comes out as a companion more than a phone to the user," Dagar told Times of India. "We plan to do more advanced versions of the phone in the future."The device has been in the works for three years, and it may be ready for release later in 2013. The team plans to sell the phone for around $185.

Vayutuvan · Post by **Vayutuvan** » 30 Apr 2013 00:13

What is with the refills for G2? Each refill costs more than the pen and refill. What a waste of resources? Somebody should kick these bozos' a***s.

archan · Post by **archan** » 30 Apr 2013 20:13

vishvak wrote: I would rather learn to deal with win8

The other day also someone asked how coders deal with win8 and I asked if they had actually used win8 or just seen it in pictures. To that, another person said they have, and it is not special. Agreed. It is windoze after all.
However, I wonder why this concern regarding coding and power users. Any moderately advanced user can:
1) configure it so it boots directly into the desktop, no need to see the start screen.
2) Get a menu with shut down, restart ityadi.
3) getting to programs is as fast as it was in win7, it just looks difficult. When you press the start button, the whole desktop disappears and which can make some people feel like they lost some control over their PC. Understandable. But if you want the notepad or the calculator, all you need to do is type the first couple characters and you find the program in front of you. It is fast. I found getting to control panel or mouse settings faster or at the same speed than I do on win7 or XP.
It doesnt stop you from using cmd either or stop you from installing Visual Basic etc.. So how will it hamper a programmer?

I have win8 on my touchscreen home laptop while win7 on my work machine. I dont feel win8 is any worse than win7. I still go back to Ubuntu in a VM even on the work machine, though.

Mort Walker · Post by **Mort Walker** » 30 Apr 2013 22:37

Anujan wrote:Same thing happened with my montblanc. Someone borrowed it and conveniently lost it.

The best hack though is to place a montblanc rollerball refill inside a pilot Body. You have a $300 writing experience in $20 or so. Most websites ask you to use a G2 plastic body, I use a pilot pen with a heavier metal body but forget the model name.

http://www.wired.com/gadgetlab/2010/10/ ... azing-pen/

I used to carry nice Pilot pens, but as I travel frequently it became problematic. I found that in air travel, particularly overseas, the ink cartridge would sometimes leak. Then I had the problem in desh in public places where people always want to borrow your pen if they needed one. Most everyone returned the pen, but on occasion they wouldn't and would walk off with it in a crowd. Now I put put my pens in zip lock bag in my carry on bag.

ramana · Post by **ramana** » 30 Apr 2013 22:38

Blackberry chief says tablets will be obsolete in five years. Is it a prediction or sour grapes?

I use Uniball Signo Model 207 pens. it has almost fountian pen feel.

Negi, First think of getting married.

Mort Walker · Post by **Mort Walker** » 30 Apr 2013 22:44

Raja Bose wrote: Saar then you have no idea what design really means. Hint1: Design is not just about how something looks like. Hint2: GB's low-end S30/S40 phones actually sell more than Sammy's becoz they have great design despite being slightly more expensive - and these are budget-minded "poor" consumers we are talking about in India/S.America/ME/Indo-China.

PS: A good finished product doesn't get made from thin-air. It gets made from a good design process. And Sammy scores quite poorly there.

You're right, I don't have any idea of what design really means in the context of CE devices. I do know a little teeny-tiny bit about integrating electronic components. The S4 is great device with compromises as are all the recent flagship mobile phones. Implying the S4 is a turd is the same as calling the Lumia 920 or iPhone 5 turds.

Vayutuvan · Post by **Vayutuvan** » 30 Apr 2013 22:48

>> Negi, First think of getting married.

but shouldn't that go into L&M also?

Reminds me of the witty Telugu proverb "Alu lEdu choolu lEdu koDuku pEru rAmalingam" (Difficult to translate but roughly no wife no pregnancy yet son's name has already been picked out. It is Ramalingam). Negi ji, just pulling your leg. Please don't take it otherwise.

Mort Walker · Post by **Mort Walker** » 30 Apr 2013 22:50

http://www.youtube.com/watch?feature=pl ... 19vR1GldRI

gakakkad · Post by **gakakkad** » 30 Apr 2013 23:05

Anujan wrote:Well I disagree. Montblanc are smoother than Reynolds. They are more balanced with the right thickness. They to their primary job -- writing -- better than other SDRE options. Question is do costly watches do their primary job -- time keeping -- better than SDRE options?

talking of malfunction , I have a mont blanc fountain pen that someone gifted , it kept blotting even on high quality paper..

Raja Bose · Post by **Raja Bose** » 01 May 2013 01:47

Mort Walker wrote: You're right, I don't have any idea of what design really means in the context of CE devices. I do know a little teeny-tiny bit about integrating electronic components. The S4 is great device with compromises as are all the recent flagship mobile phones. Implying the S4 is a turd is the same as calling the Lumia 920 or iPhone 5 turds.

Integrating electronic components is a small part of what goes into CE devices (and the most commoditized one right now). You can do a perfect job with the HW and still have the final device come out as a turd - that's what makes CE different from selling gray boxes for backend stuff. Its not about the compromises in S4 - all devices have that. It is about Sammy's culture of blindly cramming in stuff with no regard to good UX. It is quite evident in pretty much every single piece of user facing SW they put in their phones. Its almost like the stuff got built by a bunch of robots who were told to check every box in the list and no thought goes in why something should be included or excluded (going by accounts from folks who do work there, the truth is not far from that). Good design is as much about not putting in certain stuff as it is about putting in stuff.

Anujan · Post by **Anujan** » 01 May 2013 02:06

Don't blame sammy for their lack of mobile prowess. They are fundamentally a hardware company and they do things like tack on software because they have to differentiate. Good UX is all nice and dandy, but they are enabled on top of good software engineering. Think of Siri or Google Now. To make something like that, not only do you need to have a good software, but you also need good cloud services. That is a considerable amount of investment in cloud infrastructure, software and engineers. If you are not a software company you can never pull that off.

The beauty about cramming in software is that the users can disable it if they want. It is not like shoving a hardware feature which stays on. OTOH remember that Sammy makes their own components and along with manufacturing and distribution, they are world beaters. Which other company makes everything from RAM to chips and does R&D on displays, makes and distributes their phone at such a massive scale? Fruitco comes close but not quite. Calling their phones a turd is quite a stretch.

As an aside: got my grubby hands on SG4. pretty nice set. HYC one is just a shade better - I like the more solid construction, speakers in front and better low light photographs. Sammy feels thinner though (didn't compare them side by side, going by memory of HTC one). All in all a pretty good phone.

I think phones have reached a point where improvements are likely to be only evolutionary. Like laptops. The days of "wow"!! are gone. This is both a good thing and a bad thing.

Vayutuvan · Post by **Vayutuvan** » 01 May 2013 02:10

>> This is both a good thing and a bad thing.

Good for consumers and bad for manufacturers.
Margins will fall as things become cheaper for reasons of competition.

Mort Walker · Post by **Mort Walker** » 01 May 2013 02:19

Raja Bose wrote:It is about Sammy's culture of blindly cramming in stuff with no regard to good UX. It is quite evident in pretty much every single piece of user facing SW they put in their phones. Its almost like the stuff got built by a bunch of robots who were told to check every box in the list and no thought goes in why something should be included or excluded (going by accounts from folks who do work there, the truth is not far from that). Good design is as much about not putting in certain stuff as it is about putting in stuff.

The software "features" Sammy adds do not all have to be used. Some are good, some are not. If that ruins the UX for you, so be it, but it is still unfair for that reason you imply the S4 is a turd. Calling it a turd for a locked boot loader is legitimate.

Mort Walker · Post by **Mort Walker** » 01 May 2013 02:27

Raja Bose wrote:Integrating electronic components is a small part of what goes into CE devices (and the most commoditized one right now).

Getting CE commoditized takes a fair amount of work years before. To make the next significant breakthrough will take a fair amount of work again in basic sciences of physics and chemistry, and hell of a lot of engineering effort.

Raja Bose · Post by **Raja Bose** » 01 May 2013 02:46

^^Bad UX in a CE product implies product is a turd. Plain and simple. And its not a question of whether I use a feature or not. Simple example, after buying a brand new unlocked GSIII, out of the box, I am bombarded with a million pop up messages telling me I can use XYZ gestures to do everything from taking a picture to flushing the toilet with my phone. And every such notification has a checkbox asking "Do not show this message again" and a "Yes"/"Cancel" button. That is horrible UX and reminds one of old Symbian UIs where the user would get a pop up for every little detail and would have to drop what they are doing to provide explicit response to each of those messages. Basically some Sammy SVP decided "GESTURES ARE COOOLLLL!!! WE NEED GESTURES!!!!" and all the minions decided to jump on the gesture bandwagon and cram every possible gesture driven crap into the device so that at the year-end they can claim in their performance review that they enabled 2 million gestures in Sammy's latest flagship.

Anujan wrote:Don't blame sammy for their lack of mobile prowess. They are fundamentally a hardware company and they do things like tack on software because they have to differentiate.

The excuse of being a HW-onlee company no longer flies in this industry and is no longer a security blanket. It did in the days of Old GB but no longer so and GB learnt it the hard way (back in the days they used to think Apple would never pull off a phone becoz they didn't know how to do telecom HW). HW is an increasingly commoditized business (and that is in line with computer industry trends over the past half century). In fact HW-onlee mobile CE companies will perish, get bought out (eg. Motor Oil) or get commoditized out of business which is why Sammy sees the writing on the wall and is pouring billions into its SW arm. Ofcourse all this money means jack unless they change their product design/development culture. They tried hard end-2011 - mid-2012 and still fell flat due to tons of internal politics and clashes with their corporate culture.

Anujan wrote: The beauty about cramming in software is that the users can disable it if they want. It is not like shoving a hardware feature which stays on.

That is what's good and bad about software - its easy to shove things in and dump a bunch of useless stuff on the user without impacting production cost. In HW you cannot do that - every component needs to be accounted for in the BOM, to the fractions of a cent sometimes. So just cramming in stuff for the heck of it is much harder to pull off there vs software. Dumping a bunch of crap on the users and expecting them to sift thru them and choose what they want or not is not good UX. Its funny how the same arguments repeat in cycles over the years. The above was the exact same argument pushed by the likes of GB/Palm/Sammy back in the earlier days of the mobile industry, when S60v3 etc. used to rule the roost and then they got hit between the eyes with the iPhone.

Anujan wrote: OTOH remember that Sammy makes their own components and along with manufacturing and distribution, they are world beaters. Which other company makes everything from RAM to chips and does R&D on displays, makes and distributes their phone at such a massive scale? Fruitco comes close but not quite. Calling their phones a turd is quite a stretch.

Back in early-mid 2000s, one could say the exact same thing about GB. In fact GB's scale was far larger than Sammy's is right now and we all know how that turned out.

All of the above help in making a good product but doesn't automatically translate into good products. Fruit Co doesn't really do its HW in-house except SoC design - in fact right now the folks who do major HW R&D which goes into their own CE products is pretty much limited to (in order of size of effort): Sammy, GB, Yell-G, Mickey. And out of them only 3 also act as their own ODMs (all partially ofcourse).

Anujan wrote: I think phones have reached a point where improvements are likely to be only evolutionary. Like laptops. The days of "wow"!! are gone. This is both a good thing and a bad thing.

In terms of basic feature set it will be evolutionary since a certain stability/base-line has been achieved around 2010. However in HW certain components will still see revolutionary advances such as camera, speech, battery, display technology and most of all form factor (esp. due to advances in materials). In SW, sky's the limit.

Raja Bose · Post by **Raja Bose** » 01 May 2013 02:52

Mort Walker wrote:
Raja Bose wrote:Integrating electronic components is a small part of what goes into CE devices (and the most commoditized one right now).
Getting CE commoditized takes a fair amount of work years before. To make the next significant breakthrough will take a fair amount of work again in basic sciences of physics and chemistry, and hell of a lot of engineering effort.

CE HW is already commoditized. Otherwise MediaTek would not be in business. That's the problem with HW - its infinitely harder to do that software, requires massive investments of capital, skill and time, has a huge risk and limited RoI. Today a snotty nosed 15 year old kid can make a website for adding goofy effects on photos and sell it for much more than one could sell a solid reputable HW company which did solid advances in low power wireless - the former can be sold for $50 million whereas the latter would be lucky to get $25 million. Its unfair, but thats how it rolls.

Vayutuvan · Post by **Vayutuvan** » 01 May 2013 03:01

There is also early mover advantage whether HW or SW. For example, an exTCS guy sold an MPEG 4 ASIC chip set company for $400 million or thereabouts back when MPeg 4 was new. I don't think anybody can command those kind of prices today for even more advanced ASICs.

Raja Bose · Post by **Raja Bose** » 01 May 2013 03:05

^^^That is why periodically BIL (who is an IC designer) will go

about the unfairness of it all and the ghor kalyug.

Anujan · Post by **Anujan** » 01 May 2013 04:36

The only mistake GB made, IMHO is to not spot the trend and move with the change. At the pricepoint that people were willing to pay for phones, either they had no products or their software, hardware and services *all* sucked**. Companies get comfortable doing what they do best, even when the ground shifts from under them. I am not sure if we can derive a Hardware is getting commoditized, hardware company vs software company lesson.

Take the first counterfactual argument. M$ had a phone OS. On their desktop OS they knew how to create APIs, nurture developers and create a software market and an ecosystem of software and developers.

1. When first Palm succeeded with its PDA and Handspring later folded PDA functionality into a phone. This was Treo 180, way back in 2002.
2. Later FruitCo succeeded with iPod, people wanted music player phone and Moto Came out with Rockr.
3. People loved their Blackberrys. In fact, it used to be called "Crackberry", they liked sending and receiving email and messages on their phones.

All these were years before the iphone.

What prevented M$ from reading these signs and trying to replicate the same model they had with their desktop OS? Essentially it is what FruitCo has done. Created a marketplace of apps and developers. Why didnt M$ produce a phone OS, a marketplace for apps and an ecosystem of developers? They did it with Xbox so doing it in the desktop wasnt a one time wonder. If hardware is so easy to do, or hardware partners so easy to find and all the magic is in software, why was the king of software M$ blown out of the mobile market? 6 years later in 2013 they are still fighting to stay relevant in the mobile space.

Conclusion 1: A software company got its mush kicked in the mobile space, even though "hardware is commoditized and easy to do" because their software wasnt any good.

Take counterfactual argument 2: Blackberry. They designed the hardware, the software and the cloud services. They are passing through a near death experience and their CEO just gave an interview that tablets will die

So you have a "vertically integrated" company which couldnt put out a decent phone to compete with iPhone till Z10 came out. Why? Because they had no software that supported third party development and cloud services that people have come to expect.

Conclusion 2: A vertically integrated company which did hardware, software and cloud services go their mush kicked because they didnt do any of them well enough.

From the ashes of GB, M$, Motorola and other chota motas like LG ityadi, two companies which sell phone hardware which have made billions are (1) FruitCo (software company) and (2) Sammy (hardware company). A spectrum of other companies like M$ (software), BB (integrated), Nokia (mostly hardware but software too) have gotten their mush kicked atleast for now.

So I think the lesson to be learned is not that Hardware is getting commoditized or APIs are getting commoditized (will happen when apps interoperate) or software companies succeed at the expense of hardware companies. It is that execution matters. Moving quickly without inertia matters. Not being comfortable doing something that the companies have always done matters. M$ failed in the first round because their software sucked. Nokia failed in the first round because their software *and* hardware sucked. Moto failed in the first round because manufacturing and distribution sucked. BB's software, hardware *and* cloud services sucked. FruitCo and Samsung simply made better hardware and made (or had access to) better software. it is an all round thrashing.

As an aside: FruitCo are good at software. They are good at hardware. They suck at cloud services -- MobileME was a joke. iCloud is reasonably okay now. Maps was/is a joke. Their email is a joke. They cleverly partnered with Chacha for youtube, Maps, email, contacts sync when iPhunwa was introduced. Now they are building out their own cloud services. Which by the way is a brilliant masterstroke to take on blackberry and exchange sync of M$. This is an example of brilliant execution at either being excellent at everything you do or partnering with someone else in areas you are lacking.

**I am sure that there is some correlation between the fact that (a) cost of cell phone service/data service is going down and the fact that (b) many people have ditched their land lines for cell phones influencing the fact that (c) People are willing to pay more for their phone device. in 2007, if I was ready to plonk down $500 on a phone, iPhone was the only game in town. Nobody produced anything close that justified the price. On top of that, FruitCo's deal with ATT meant that customers were soon plonking only $200 on a new iPhone with contract -- about the same price or slightly more than what they were plonking for a dumb phone/windows mobile phone. That is one of the major reasons for adoption of the iPhone. Pricing trends/distribution deals were probably more instrumental in killing GB's dominance than hardware commoditization. People were simply willing to pay more for their phones in return for more features. None of the traditional phone makers had any phones in that pricepoint. Commoditization hurts companies if average selling price of hardware declines and differentiation goes away. When iPhone succeeded, prices actually rose and iPhone was very differentiated from any other phone anyone else was making. Then how can we claim commoditization killed GB?

The pricepoint argument is perfectly captured by Ballmer gloating 500$ for a phone?! LolzOMFG!!!

Raja Bose · Post by **Raja Bose** » 01 May 2013 05:15

Saying HW is commoditized doesn't mean its easy to do - thinking so is a fallacy. It just means one's profit margins on selling HW onlee are much lower and that is pretty much true for everybody except FruitCo. Just do a calculation on how many units of phones Sammy needs to sell to make a million $ in profits (with capital costs amortized) vs someone selling units of SW like Mickey and you will get the answer. Value addition which results in customers ponying up more cash rarely happens at the HW level now hence, RoI is lower. Is mango abdul customer going to pay me $20.- more if my GPU does extra FLOPS or will he pay me more if my camera software can automatically push pics to Instagram? Add to this the high capital cost of doing HW R&D, development, tooling and not to mention the risk and its evident why being HW-onlee is not sustainable for anybody in the CE business. This is another reason why VCs in the valley shy away from investing in HW startups - too high risk, too low RoI. Having HW-onlee partners suits platforms vendors like Chacha and Mickey but it hurts those companies which are HW-onlee. This is the reason why Sammy and GB are trying to move into services - thats where the margins are high and upfront risk is low plus very important, the cost impact of failure is lower. Being HW-onlee is simply not sustainable from a business standpoint when it comes to the mobile CE industry today. As a business owner I want biggest bang for my capital investment bucks and being HW-onlee does not give you that. This will not make companies like Chacha or Mickey happy which is why both are going/will seriously go vertical at some point. In fact when all major players go vertical, one can assume that some stability condition has been reached.

Anujan wrote: 2. Later FruitCo succeeded with iPod, people wanted music player phone and Moto Came out with Rockr.

Actually Moto's ROKR was what spurred the Mahdi to go for the iPhone becoz it was such a turd. Initially the Mahdi just wanted to stay on the SW side of things and along with the Sprint CEO (iirc) got a collaboration with Moto which resulted in the ROKR. The Mahdi took one look, blew a gasket, invented a few new abuses and put Forstall/Ive/Cue to work on the iPhone. And the rest is history. Sis still has a ROKR from those days (and a pre-BB Motor Oil developed handset which became BB's 1st handset).

Any dominant company will face the incumbent's dial-e-amma and will pay the price for their blind spots (whether its Mickey with mobile or Chacha with social or ChipZ with low power chip solutions). Mickey's blind spot was that it was a desktop centric company (heck its XBox was internally seen as a HTPC more than anything else) so for it mobile was seen as a side-hobby while desktop was seen as the core business (by a certain Sinofsky). History repeats itself and either the company in question perishes or feels the pain and adjusts. In Mickey's case they will try to leverage their dominance in PC and XBox to drive their mobile growth. In case of Chacha, its their online services which drive Android sales. In case of FruitCo, it is the strong brand loyalty and well integrated vertical offerings which drive the iDevice sales not to mention the slick marketing/impulse buy triggers they utilize to make people want their CE products.

Anujan wrote: Conclusion 1: A software company got its mush kicked in the mobile space, even though "hardware is commoditized and easy to do" because their software wasnt any good.

Conclusion 2: A vertically integrated company which did hardware, software and cloud services go their mush kicked because they didnt do any of them well enough.

This is basically another way of saying HW is a commodity. If your HW is good, SW sux you will get your mush kicked. If your HW sux, SW is good, there is a chance you might make money (the old iPhones). If both suck, well your product is dead. So just doing HW good doesn't cut it - it used to back in mid 2000s and before but not today.

Anujan wrote:Then how can we claim commoditization killed GB?

Actually it is commoditization which killed GB (or at least put it in near death mode). When you have commoditization, you need to maintain much higher volume of shipments in order to derive the same amount of profit, simply becoz margins per unit are low. So when there comes a situation where you cannot maintain a high volume of shipments due to some other guys selling shinier stuff or other reasons such as getting snubbed by carriers (which is what happened to GB even when it was #1 - how and why it happened is another interesting story), you will die. Now compare this with someone like Fruit Co. They have lost marketshare to Android in both tablets and phones but guess how much impact it had on their profits from tablets and phones - Zero Nada Zilch. That is becoz FruitCo HW is not commoditized and they enjoy huge margins thru vertical integration in their products plus sweetheart deals the Mahdi crafted with carriers like ATT. The fact that HW is a commodity is not a new phenomenon - it can be traced at least back to the early 1980s if not earlier. And this is becoming even more painfully evident now. This causing even SoC vendors to become more software-centric - just ask Nvidia or QCOM. Nvidia even back in 2009 had more software application engineers as compared to HW engineers. TI ofcourse threw in the towel and winged it for safer pastures like automotive. The next (or next to next) space which will get disrupted by this whole app/services/platform drama will be automotive. An average person spends more time in their car today in massa than they do on vacation.

Anujan · Post by **Anujan** » 01 May 2013 05:19

Raja Bose wrote: Actually Moto's ROKR was what spurred the Mahdi to go for the iPhone becoz it was such a turd. Initially the Mahdi just wanted to stay on the SW side of things and along with the Sprint CEO (iirc) got a collaboration with Moto which resulted in the ROKR. The Mahdi took one look, blew a gasket, invented a few new abuses and put Forstall/Ive/Cue to work on the iPhone. And the rest is history. Sis still has a ROKR from those days (and a pre-BB Motor Oil developed handset which became BB's 1st handset).

This is actually not true. Mahdi was working on the iPhone and purposely crippled the Rockr so that it would die. This is well known. The reverse is true.

If your HW sux, SW is good, there is a chance you might make money

Netbooks were as cheap as iPad. Arguably they could do more things than iPad, yet they died. The hardware was horrible, the software was great and desktop class. On the flip side, the argument towards buying a surface (netbook dimensions but better hardware) seems to be "iPad with desktop software". So why did iPad succeed netbooks didnt and why would surface succeed when netbooks didnt?

And why did the original iPhone suck? It had camera, wifi, cellular data and a touch screen. With enough processor and ram to run conventional browser, maps and youtube. It was miles better than the windows mobile phone hardware with their resistive screen, wimpy processors and constrained RAM.

Raja Bose · Post by **Raja Bose** » 01 May 2013 05:39

Actually yours truly had worked on stuff which went into the ROKR as a contractor (if the display driver on it sucked, you know who to blame

). It was a Moto designed phone thru-and-thru and FruitCo's involvement was limited to the music/iTunes functionality. The Mahdi could not have crippled it becoz he really had no say in its design and development so that's a bit of an urban legend perpetrated by Ed Zander (the same idiot who basically worked the designer of the RAZR to death, literally). In fact I doubt even if given a free rein the Mahdi could have done a better job making it suck as compared to what Moto's own design-by-committee suits did.

Netbooks are an example of where HW and SW both suck. Having SW which rocks on the desktop and putting it on a puny processor means now the SW sux. If HW sux, one needs to design SW to get around that. The original iPhone is a great example of that. It didn't have half the features of the other smartphones, had a low-res screen for the longest time but with superb refresh rate and a capacitive touch screen which the software took full advantage of. As a result people purchased it in droves becoz as a package it didn't suck - it totally rocked. As engineers we make the mistake of thinking, "oh component XYZ sucks so the whole thing must suck" or "component XYZ rocks, clearly the whole thing rocks". That is never the case. Every device is a compromise and making it a great product involves amplifying the strengths and getting around the weaknesses in a way where the customer notices the strengths, not the weaknesses. That is what product design is all about in the end.

Anujan · Post by **Anujan** » 01 May 2013 05:43

Raja Bose wrote:Saying HW is commoditized doesn't mean its easy to do - thinking so is a fallacy. It just means one's profit margins on selling HW onlee are much lower and that is pretty much true for everybody except FruitCo.

Actually this is not true. Commoditization means that the essential attributes are indistinguishable in the eyes of the user. Means you dont care what brand rice it is, just that it is rice. The falling prices is a consequence of commoditization rather than the causal reason of commoditization. If you dont care what brand rice it is, you will pick the cheapest bag, which means that everyone races to the bottom cutting prices and margins. This typically happens when ability to produce the said product becomes so widespread that all features can be replicated by everyone. If making the product is so difficult that only a few can make it, how can it get commoditized? Take desktop processors. Is it commoditized? Only Intel (well maybe AMD) can make them. Then how is it commoditized hain ji?

Add to this the high capital cost of doing HW R&D, development, tooling and not to mention the risk and its evident why being HW-onlee is not sustainable for anybody in the CE business.

This actually refutes your own argument. If the capital cost of doing hardware R&D and tooling is so high, only a few will do so. That means that the market is consolidating and not getting commoditized. Every abdul in China can make a sock. Essentially the same pair of sock irrespective of the brand. It is commoditized. Only a few companies make aircraft engines. Because it requires huge investment in R&D. Aircraft engines are not commoditized.

OTOH what is happening is that innovation is dying down. Things that require high R&D still have high margins, the people who go

about commoditization are the ones who are doing zero innovation. Take Dell for example. Typical example of getting commoditized. Did the do R&D on the processor? No. R&D on Disk/SSD? No. R&D on Motherboard/LCD/Batteries? No. R&D on graphics cards? No. Then what exactly were they doing? They were *assembling* PCs which every abdul in China can do by sourcing components from Intel, Nvidia and Samsung (all of whom have healthy margins). That is why they got commoditized away with shrinking margins. Did they atleast have a fantastic distribution channel? Nope. Amazon and FruitCo had it. Did they atleast have fantastic service? Nope. IBM (during lenovo days) had it. Essentially the marginal value proposition of buying a Dell computer was zero when compared to buying something from HP or Lenovo. Essentially they were doing Yum-Bea-Yae middleman giri with no value addition and then

when every abdul realized he could do Yum-Bea-Yae middleman giri too! For cheaper.

OTOH, FruitCo doesnt research desktop processors, SSDs and LCDs either, but they innovate where it matters. They have the best design for laptops, they have the best retail channel and they have the best service (well it costs, but still..). That is why dell is dying FruitCo is not. Random abduls cannot make unibody aluminium chassis, that requires R&D and capital investment. Thats why Macs command a premium.

I would also claim that "Making a PC" (read assembling a PC) is easy hence assembling PCs is commoditized. Making a PC as good as FruitCo is not (yet) commoditized. So people will pay the lowest price they can for a PC, it is a race to the bottom. Making desktop/server class processors with high investment in R&D and capital cost is not possible by most companies, that business is still not commoditized. Intel charges whatever they think they can short of the extent of killing the golden goose. Last time I checked, Intel's margins were doing just fine and just a few percentage short of M$ and Chacha. In fact they had better margins than M$ in a few quarters. And all this, even when AMDs are for all practical purposes indistinguishable to the user from Intel!

How much of "cell phone commoditization" happened because now we have ready made solutions with baseband/modem/board pre-fabbed and any cell phone company needs to only slap on a plastic case around it? I remember reading about a $12 phone somewhere! Now in such a marketplace, if GB continues to churn out dumb phones and people dont buy it and the margins are shrinking, is it because of some inherent nature of hardware business or is it because essentially a GB phone now is indistinguishable from any other dumb phone I can buy because they havent put anything new in it?

Raja Bose · Post by **Raja Bose** » 01 May 2013 06:15

Anujan wrote:Commoditization means that the essential attributes are indistinguishable in the eyes of the user. Means you dont care what brand rice it is, just that it is rice. The falling prices is a consequence of commoditization rather than the causal reason of commoditization. If you dont care what brand rice it is, you will pick the cheapest bag, which means that everyone races to the bottom cutting prices and margins. This typically happens when ability to produce the said product becomes so widespread that all features can be replicated by everyone. If making the product is so difficult that only a few can make it, how can it get commoditized?

^^^That makes the mistaken assumption that in order to do HW, you need to do original R&D and bear those costs. In fact nothing could be further from the truth. In fact pretty much the whole chipanda industry depends on that very thing especially in mobile!

Only 1 or 2 companies need to do R&D, rest just use that to do mass manufacturing and it has literally become a race to the bottom in terms of unit cost and margins. In fact using your rice argument, that is exactly where most major HW components in mobile devices are at. As a customer do you care whether your mobile device has a Tegra or a MediaTek or Rockchip?? As a customer do you care or even care to know whether your phone has a Samsung display or a CMI display or some local Yangxi display?? You don't. In fact when one designs a mobile device today and then goes to a chipanda ODM, for every component one is asked - do you want name brand or local Chinese part. And I am not just talking about the major assemblies like display, battery, SoC either. Every frickkin part has a doppelganger at least and most have multiple options with very similar specs and price difference down to 1/100th cent. That is commoditization at its most basic avatar. It is not rewarding enough to just do R&D on HW. In this aspect HW is almost like pharma where chipanda are like the SDRE pharma companies - they benefit the consumers but cause the companies doing core R&D nightmares.

Anujan wrote: phone commoditization" happened because now we have ready made solutions with baseband/modem/board pre-fabbed and any cell phone company needs to only slap on a plastic case around it? I remember reading about a $12 phone somewhere! Now in such a marketplace, if GB continues to churn out dumb phones, is it because of some evil of commoditization or is it because essentially a GB phone now is indistinguishable from any other dumb phone I can buy because they havent put anything new in it?

In fact the above is a good example of the impact of commoditization. Back in the days GB which did and still does core research on cell modems, voice, DSPs could sell phones by differentiating on stellar call quality, better battery life blah blah. Then came the chipanda companies which can essentially make the same thing at half the cost. So where does GB gets back its RoI on the years of research that went into making phones last longer with great call quality? It can't. It only gets a slice of the returns becoz rest are eaten by folks who didn't do the R&D but did just did the production (like Dell). So now GB has to push itself even harder to differentiate. Now where is the money going to come from to do this next level of R&D? And this assuming all R&D succeeds, is on time, doesn't go over-budget etc. So even in an ideal situation its a slippery slope, forget about average case.

Commoditization of HW ensures that you cannot just keep differentiating on HW and the cost of differentiation in HW becomes higher and higher with lower and lower RoI. Hence, that is simply not sustainable. That is why LG and HTC get the jitters. Companies which are HW onlee in mobile right now are not so due to choice. After all which CEO is mad enough to go in for higher risk when there are lower risk/more sustainable ways of making money. Hence, the push by traditionally HW-onlee companies like Sammy and GB to get into the SW/apps/services game. Just like SW companies can get into HW (Fruit/Mickey) the reverse can happen too. GB is already showing signs of that with their nav and other stuff. Sammy will follow. All it takes is the right culture in place with strong financial backing and some runway.

PS: Dell's entire business actually was built on JIT manufacturing and delivery - they just happen to be in the business of making PCs. I would say they had a pretty darn good supply chain going at one point, almost rivaling that of GB back then.

Raja Bose · Post by **Raja Bose** » 01 May 2013 06:27

Comparison of the desktop processor space with mobile processors is fundamentally flawed. In desktop ChipZ does both the architecture and manufacturing. How the 64 bit architecture came to be amd64 instead of x64 was another smart anti-trust move by ChipZ. So if you want to buy a desktop processor, you have to buy a ChipZ or AMD product. Hence, ChipZ can afford to fund all sorts of R&D (a lot of which end up in failure as expected) without breaking the bank becoz their cash flow from current and legacy product line is assured. OT but basically in this context, Mickey and ChipZ operate in similar manner - they have strong existing revenue sources for decades which they use to fund other stuff - mostly they fail and sometimes when they succeed they make a killing.

In contrast, ARM just does the architecture but not the manufacturing - its fabless. So it licenses it to as many willing folks as possible. You can even take their base reference design and add your own stuff like what FruitCo/Sammy/Nvidia/QCOM/TI/MediaTek/Rockchip/Foxconn do. That has resulted in mobile processors becoming an commodity. On top of that the rise of Android has ensured that the cost of switching for a product owner is even lower. Today if I put my SW stack on top of Android, I can pretty quickly switch between QCOM, Sammy, Nvidia, Rockchip and in some cases MediaTek/Foxconn without any major porting efforts. So my primary consideration becomes price rather than the manufacturer of the SoC (SoC is no aircraft engine, its more like a bag of rice

). For all I care it can be a gray square with no name on it as along as it works reliably. And my customer doesn't care either. That way I can eke out a few more cent or if Allah is most benevolent and ever merciful, even $2-$3 extra margin per unit.

Raja Bose · Post by **Raja Bose** » 01 May 2013 07:47

Google Glass review (Explorer Edition)

The technology may be raw and early stage but the industrial design on this one is top-notch given the limitations imposed by the display technology and interaction modalities. All credits for this go to a certain Scandinavian motorhama named Isabelle.

Anujan · Post by **Anujan** » 01 May 2013 08:22

Raja Bose wrote:Comparison of the desktop processor space with mobile processors is fundamentally flawed. In desktop ChipZ does both the architecture and manufacturing. How the 64 bit architecture came to be amd64 instead of x64 was another smart anti-trust move by ChipZ.

This wasnt an anti-trust move. Intel moved to IA-64, originally designed by HP and co-implemented by Intel and HP in their Itanium line. HP wanted to move their superdome from PA-RISC to itanium. Under a deal, Intel bought the itanium architecture from HP, and it would be their next server class (and ultimately desktop class) processor.

Itanium was a VLIW (or EPIC as they would call it) architecture and not out of order superscalar like IA32 (x86). I was part of the team that implemented the trimaran opensource compiler for IA-64 architecture, it was opensource version of a reference compiler implemented by HP while researching Itanium. (so I have undeniable upper hand* in such issues

). AMD took the more conservative approach and moved to AMD64, essentially IA32 with 64 bit extensions.

Nobody bought Itanium. It had emulation mode for legacy software but if you wanted full performance benefits, you needed to recompile the software. Last nail in the coffin was chacha refusing to buy Itanium because it was a power hog. So it didnt make any dent into (the then nascent) data center market either. I was interning in HP those times, still remember the BS market predictions by Gartner that Itanium will take over the world, duly circulated among us to keep our spirits alive

This was before the bottom collapsed and people couldnt hide that Itanium was a failure.

The only ones which would use Itanium was HP nonstop line (used to be called Compaq tandem) used heavily in the banking and financial industry. It had two processors which would do calculations and the results would be compared.

Long story short, Intel had to go with their tail tucked in and implement AMD64.

Postscript 1: Trimaran still lives on on a research compiler. Yours truly's name is on the website too.

Postscript 2: Itanium had a fantastic floating point pipeline. Better than AMD's (I think at that time it was called sledgehammer). As the joke went "Intel's Alpha engineers were better than AMD's Alpha engineers"). At that time DEC Alpha was the best architecture around, but the company went bankrupt and half their engineers joined Intel and the other half joined AMD

All because M$ refused to port NT to Alpha because they picked up a fight with Compaq