Phone, Tablet and Gizmo Thread #0x02

Raja Bose · Post by **Raja Bose** » 08 May 2013 01:42

This brings me to a question I had asked previously - what is DropBox's long-term viability and how do they plan to survive, given vendors such as Chacha and Mickey are themselves providing the same service integrated into their platforms and office suites in addition to providing clients on other platforms to enable x-platform access.

The benefit of such an integration is becoming evident - SkyDrive despite launching after DropBox has 2.5x the number of users today:
http://www.engadget.com/2013/05/07/skyd ... ion-users/

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 02:05

Ramana garu
My understanding at this point is that PI, Trade Secret, and Company Secret are all synonyms out of which Company Secret seems to be not in wide-spread use (I distinctly remember seeing this term in one of the IEEE books). No matter, the question is that what form of protection exists for software. RB was coming from patent side which offers protection for gadget/process where as my/our company concerns are/were more for software protection which is non-existent because firstly, software cannot be patented, and secondly we found out copyright not only does not offer adequate protection, it can even be detrimental. That leads software developers to protection through a mixture of Trade Secrets/PI along with NDAs and Confidentiality agreements, non-public etc. Of course, for big companies who have end-user software products, with substantial amounts of software (say MS with their Compiler/Visual Studio) or Oracle DBMS, this doesn't matter as their customers are end users who are much smaller. These kinds of problems are faced by companies developing software libraries. Usually these companies tend to be small start-ups but their customers are bigger software product companies who have end-user products with hundreds of thousands if not millions of installed seats. Most times, the front-line support is handled by the licensee. This is so because most support calls tend to be UI not behaving according to spec/incorrect documentation/plain user misunderstanding. If the fix for the error involves changing an interface to the licensed library, then the end-user company has to go back to the library developer. Since the library developers are smaller companies, they may not respond in time. The resulting user dissatisfaction reflects on the bigger company as it is their brand name that loses equity in the eyes of the user.

The IP protection issues are not unique to software industry alone. What is different from manufacturing/chemical/process/pharma industries is that software is a bunch of bits, i.e. there is no physicality, and hence there is no protection as such once the sources get out. For most classes of software implementing complicated/complex algorithms, just binaries won't be enough to reverse engineer, but sources coupled with published journal/conference articles would be enough to reproduce the functionality with not much capex. That is not so in manufacturing/process industries, is it not? Even if somebody knows the chemical formula for a drug, capex tends to be substantial because price competitiveness is another name for economies of scale.

On the other hand, combining proprietary algorithms coming out of general theories tailored for a particular application area into a software product/library, testing, and demonstration of the effectiveness of the software, and mass duplication does not require all that much capital (relatively speaking). This is so because all the infrastructural tools are FLOSS - editors, compilers, OS, debuggers, etc. The downside of course is that the opex tends to be high, i.e. major expenses are salaries of the developers/testers/technical marketing. But one can start small and ramp up as time goes by. The above perforce is a simplified general view, of course.

By the way, some of the IP issues are closely linked to algorithmic complexity, information theoretic complexity etc. That is the reason why Software IP lawyers with advanced technical degrees are in great demand and charge big buck$.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 02:56

Raja Bose wrote:^^^Nope it doesn't. If anything the more well researched your prior art is and claims framed accordingly, the harder it is for the examiner to throw your application back at you.

RB, I am not talking about only the initial patent claim - a third party can ask for a re-examination of the patent any time during the patent validity period. The fees for filing a re-examination are substantial ($27K and $17K depending on if the party asking for re-examination wants to be participate in the process or not) but well-funded entities can throw a spanner anytime during the patent validity period. For example, "prior art" is the preferred tool of FSF to challenge frivolous patents. But then software cannot be patented, so patent issues are moot as far as software is concerned. RSA has a patent but I am nor sure how they got a patent on something like that. The algorithm is known and uses well-known mathematical theories. Since there are alternatives (El-Gamal, elliptic key etc.) which are not patented/patentable, RSA is not challenged, I think.

Raja Bose · Post by **Raja Bose** » 08 May 2013 03:00

^^That can happen regardless. Hence, more the reason to ensure that prior art search is rigorously done before/during writing of claims. Personally I have yet to come across any patent which didn't have some prior art. But lack of prior art does not mean a patent cannot be invalidated becoz it may still not meet the "obvious to a person skilled in the arts" criteria. Software patents are at the center of the current patent wars in ITvity. A patent essentially describes an implementation - it cannot describe a concept or an idea onlee. Hence, there is nothing stopping someone from patenting how a piece of software does something.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 03:32

I think we are talking the same thing - software patent is almost useless and worth not the hassles/money to obtain one. One would be better served with maintaining as a trade secret.

Raja Bose · Post by **Raja Bose** » 08 May 2013 04:12

^^^Hehe tell that to FruitCo, Mickey, Sammy, Chacha, GB and others. Who said they are useless? Frivolous, yes. Useless (in practice), no. Remember, patents are both for offense and defense. If you dont have SW patents doesn't mean others will leave you alone. Being SDRE dharmic peacenik like GoI doesn't mean evil neighbors will leave India alone.

Mort Walker · Post by **Mort Walker** » 08 May 2013 04:23

IMHO, the whole s/w patents scheme needs to be re-worked. Algorithms can be patented if there is a clear distinction between them, otherwise s/w should not be allowed to hold.

Raja Bose · Post by **Raja Bose** » 08 May 2013 04:24

Question is, is algorithm == implementation? Patents can be only for implementations.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 04:28

They are enormously useful - for lawyers

and big cos in CE/Mobile space to play some stupid shell games with each other. In the end the users are going to get what they want and the prices are a-falling. I don't think all the patents in the world can save the current bubble from bursting and bottom falling out of the market.

By the way recent Readers Digest has a list of "100 most trusted people In US" and guess what, M. Gates is no. 7 and Mme. Gates is no. 9. I agree with no 1 Tom Hanks and no 10 Julia Roberts though.

Mort Walker · Post by **Mort Walker** » 08 May 2013 04:31

Raja Bose wrote:Question is, is algorithm == implementation? Patents can be only for implementations.

Sometimes. For instance the CDMA 2000 patent by Qualcomm is derived from the CDMA open standard. There is a distinct difference between the two.
The "1-Click" patent that Amazon has is based on dubious reasoning.

Raja Bose · Post by **Raja Bose** » 08 May 2013 04:33

Well Gates is probably there becoz of his philanthropy rather than his philandering.

Patents are increasingly used as strat-e-jee weapons - blame the Mahdi for it.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 04:37

Algorithms cannot be patented because they are constructive proofs of theorems. Can one patent for example any of the various sorting algorithms, FFT, shortest path, Primality testing? I think the answer is no. The algorithms themselves are published and in the public domain. Now let us take the implementation of the sort by Syncsort (which used to be the main sorting library provider to various DBMS companies - may be not currently). What would they gain by patenting the sorting implementation? That is not going to stop anybody (especially Oracle and IBM and M$) from implementing their own if Syncsort becomes a PIA wrt their licensing terms.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 04:52

Mort Walker wrote:The "1-Click" patent that Amazon has is based on dubious reasoning.

It can be argued (and the argument is on-going) that UI is process and hence patentable. Yes, it is implemented in software but the software is not patented, the idea of 1-Click is patented. But mahdi should be blamed for all this nonsense. RMS has some strong opinions about all this tamasha. Also a really smooth UI expeirnce is only part of the whole integrated iEcosystem. Same with Amazon too. Even if somebody copies Amazon's 1-Click and Amazon doesn't go after them, nothing is going to happen to Amazon's market share unless they are also as big as Amazon and have the brand equity and the backend operation and logistics network. None of the patents are going to save Apple's butt either if they falter in service delivery (like Dell, Silicon Graphics who sold out for what some measly $25 million, Motor oil - slightly lucked out because Google had too much money to throw around) or competitors with similar service delivery quality but lower prices arise (Sammy, MS et al).

The success of Apple is not due to some patent shatent stuff but they are good and filled a niche - a pretty big one at that and "skimmed the market".

Raja Bose · Post by **Raja Bose** » 08 May 2013 05:19

Patents are not used for revenue generation primarily despite all the royalty/licensing business. They are used for strat-e-jic purposes onlee.

SaiK · Post by **SaiK** » 08 May 2013 05:24

wasn't different cipher algo patented by many?

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 06:08

saik the one i know is rsa - they cashed out at $250 mil but not as good as Jacobs/Viterbi cdma who started and brought up qualcomm which has $16 B revenue and v. high gross/net margins. afaik, they are still the principles in the co.

Here is the relevant info from Wikipedia
Clicky for RSA details

History

Adi Shamir, one of the authors of RSA: Rivest, Shamir and Adleman
The RSA algorithm was publicly described in 1977 by Ron Rivest, Adi Shamir, and Leonard Adleman at MIT; the letters RSA are the initials of their surnames, listed in the same order as on the paper.[2]

MIT was granted U.S. Patent 4,405,829 for a "Cryptographic communications system and method" that used the algorithm in 1983. The patent would have expired on September 21, 2000 (the term of patent was 17 years at the time), but the algorithm was released to the public domain by RSA Security on September 6, 2000, two weeks earlier.[3] Since a paper describing the algorithm had been published in August 1977,[2] prior to the December 1977 filing date of the patent application, regulations in much of the rest of the world precluded patents elsewhere and only the US patent was granted. Had Cocks' work been publicly known, a patent in the US might not have been possible, either.

From the DWPI's abstract of the patent,

The system includes a communications channel coupled to at least one terminal having an encoding device and to at least one terminal having a decoding device. A message-to-be-transferred is enciphered to ciphertext at the encoding terminal by encoding the message as a number M in a predetermined set. That number is then raised to a first predetermined power (associated with the intended receiver) and finally computed. The remainder or residue, C, is... computed when the exponentiated number is divided by the product of two predetermined prime numbers (associated with the intended receiver).

Clifford Cocks, an English mathematician working for the UK intelligence agency GCHQ, described an equivalent system in an internal document in 1973, but given the relatively expensive computers needed to implement it at the time, it was mostly considered a curiosity and, as far as is publicly known, was never deployed. His discovery, however, was not revealed until 1998 due to its top-secret classification, and Rivest, Shamir, and Adleman devised RSA independently of Cocks' work.

I guess the patent has expired, so RSA implementation is fair game. The real difficulty lies in generating [relative] random primes of sufficient length (in bits) where the algorithm in psuedo-code is something like (in the spirit of Knuith's TAoCP)

1. Generate two random numbers k bits long (k is currently [512, 1024, 2048] bits for really strong encryption)
2. Test for relative primeness of the two numbers generated in step 1. If they are not relative primes then go to 1.
3. Save the two numbers as the private key and the product (and there is an auxiliary number) as public key.

Looks simple, doesn't it? There are several caveats etc. of course.

Vayutuvan · Post by **Vayutuvan** » 08 May 2013 06:55

Raja Bose wrote:Well Gates is probably there becoz of his philanthropy rather than his philandering.

Trusted != philanthropy. Also, I don't think he was ever philandering, in the true sense of the word. Moreover that is between him and Mme. Gates (if it was after tying the knot). Backstabbing is/was the main stay of M$ and is certainly worse, IMHO and all that, than philandering and selfishness/self-interest.

Oh, all those who want to trust M. Gates, here is something to chew on.

Bill Gates: iPad Users Are Frustrated They Can't Type Or Create Documents

Who in their right minds would trust this cr*p? Whut ... Office helps in typing? This guy has given control over to soap saleman and MS is reaping the benefits (but of course all this is no fun till Anmol mian emerges from his exile after another 20 days or so, I guess

)

Mort Walker · Post by **Mort Walker** » 08 May 2013 08:34

matrimc wrote:Algorithms cannot be patented because they are constructive proofs of theorems. Can one patent for example any of the various sorting algorithms, FFT, shortest path, Primality testing? I think the answer is no. The algorithms themselves are published and in the public domain. Now let us take the implementation of the sort by Syncsort (which used to be the main sorting library provider to various DBMS companies - may be not currently). What would they gain by patenting the sorting implementation? That is not going to stop anybody (especially Oracle and IBM and M$) from implementing their own if Syncsort becomes a PIA wrt their licensing terms.

You're thinking strictly of mathematical algorithms which are laws of nature. What I'm talking about are algorithms developed for efficient computing methodologies to use a CPU and or GPU most constructively. This happens quite regularly for signal processing. An early examples is the Red Book Compact Disc Audio standard, establishes 16-bit PCM words sampled at 44.1 KHz, by Philips and Sony in the 1970s. Manufacturers of hardware pay for the license to do this encoding/decoding on equipment every time you see the symbol "CD". Standards such as these are patented by the algorithm not the hardware itself. Hardware such as ADCs, DACs, and processors existed before and are always changing, but to put the label "CD" on your hardware means you must have a license from the developer to do this.

The big dogs in signal processing like Raytheon, Northrup-Grumman, and Lockheed-Martin, who have big business in air & missile defense, surveillance & sensors, and radar have all sorts of patents on algorithms.

Singha · Post by **Singha** » 08 May 2013 08:37

^ interesting about the MIL cos being so deep into signal processing. I used to think they integrated systems but left the heavy lifting to the likes of TI, moto, analog devices, qualcomm etc.

Mort Walker · Post by **Mort Walker** » 08 May 2013 08:45

^^^Aerospace/defense industry develops h/w and s/w for their particular use. Companies will develop specific chip sets for their use, but basically they are clean slates. The rise of Silicon Valley was caused by the push for the US ballistic missile and manned space programs starting from the late 1950s in to the 1960s. Fairchild and the traitorous 8 got their start at Shockley's Lab in 1957. From their it spread like crazy. TI was their competitor doing business for the same MIL industries.

Raja Bose · Post by **Raja Bose** » 08 May 2013 09:16

Singha wrote:^ interesting about the MIL cos being so deep into signal processing. I used to think they integrated systems but left the heavy lifting to the likes of TI, moto, analog devices, qualcomm etc.

Saar if India ever wants to be a super power, it needs to study how to build up these defence cos and use them to lead R&D. Si Valley got its start after WWII from Stan madrassa and LockMart Sunnyvale and during the Cold War the area where LockMart is located in Sunnyvale (which also used to house the ground station for KH spy satellites, namely the Blue Cube) was amongst the top targets for Soviet nuclear missiles. When people jump up and down about India becoming a super power in 2020 they have no clue how wide and deep Khan's R&D runs and just how well funded and non-elitist it is.

I doubt billu dada is trusted becoz of his Mickey antics.

However he more than made up for that with his unique way of doing hands-on philanthropy - I hear that he visits Bihar quite often these days.

Mort Walker · Post by **Mort Walker** » 08 May 2013 09:30

It's not just Stan madrassa, but UC Berkely as well and it's association with DoE. Silicon Valley was just one area in the US that had/has deep R&D. Another famous locale is the Boston area during and after WWII. Hanscom AFB houses MIT Lincoln Lab and MITRE that are completely federally funded, but pretty much hands off by the politicos. The same defense contractors on the west coast are here as well.

Mort Walker · Post by **Mort Walker** » 08 May 2013 09:32

I thought I saw Anmol logged in on this forum?

Raja Bose · Post by **Raja Bose** » 08 May 2013 10:12

He has been sent to sasuraal for 20 days.

Silly-con valley also has one of the highest concentration of chipanda agents for stealing sensitive technology and IP.

Singha · Post by **Singha** » 08 May 2013 10:17

the gates foundating has announced a $100,000 reward for anyone who can redesign the "male rubber" to make it actually give wantable pleasure than be seen as a hindrance.
http://edition.cnn.com/2013/03/25/healt ... -challenge

The rise of Silicon Valley was caused by the push for the US ballistic missile and manned space programs starting from the late 1950s in to the 1960s.

not just SV, but boston as well. apart from the mitre and lincoln, there is also draper labs. they developed the INS guidance system for many missiles as lead agency http://nuclearweaponarchive.org/Usa/Weapons/Airs.html
northrop grumann and raytheon have large facilities and I think folded in smaller cos like sanders. mirte was involved in JSTARs I think.

pretty much all the civilian side embedded sw and HW eng senior talent in that area (ie non desi / no cheeni) came out of the defence industry cutbacks after cold war. many the older labs technicians also came from there.
a few came from the likes of thinking machines, BBN, wang, DEC, groupe bull...

vets from the scene used to tell the work was long term in nature, with projects taking years, with spurts of activity and longish lulls. people sometimes misused the extensive labs for their own ends and one guy was known to be running a taxi agency from his office. they also claimed to have hoodwinked many Jernail types coming for product demos with blinking lights on panels and impressive output on screens with no mission product at back end but just a PC generating the output...because the real stuff was not demo ready...such courageous managers became real legends among the troops

Raja Bose · Post by **Raja Bose** » 08 May 2013 11:07

^^^That cannot beat the 2 enterprising ChipZ employees running a pr0n-on-demand service from the very floor which was segregated for use by the CIA taking advantage of the massive network pipe available.

Mort Walker · Post by **Mort Walker** » 08 May 2013 20:12

Singha wrote:the gates foundating has announced a $100,000 reward for anyone who can redesign the "male rubber" to make it actually give wantable pleasure than be seen as a hindrance.
http://edition.cnn.com/2013/03/25/healt ... -challenge

Not if it's running Windows 8!

We choose to do this and other things this decade, not because they are easy, but because they are hard.

$100K reward is too little and should be more like $100 million. What would be qualifications to do such a design, you may ask? Multiple pee-chaadis in biochemistry, pharmacology, physiology, and material science (chemical engineering dissertation on plastics and rubber). My bet is to fund the hex toy industry with enough funds to research this and they'll do it. Just ask the researchers at Interactive Life Forms in Austin, TX. They've made great strides in various products that would just blow your mind away (no pun intended).

Raja Bose · Post by **Raja Bose** » 08 May 2013 21:08

^^^Mort can be the test consultant to ensure it meets all ISO and CMM multi-letter multi-digit certifications.

Singha · Post by **Singha** » 08 May 2013 21:40

kickstarter could fund lots of small startups in their emerging domain. I am sure somewhere in forgotten files of du pont, bayer, 3m types there are super materials needed for this. but extensive blind trials will be needed for sure.

Mort Walker · Post by **Mort Walker** » 08 May 2013 21:59

^^^The problem is analogous to, could some one make a rain coat that you could wear in the bath that would be more enjoyable than taking a shower without a rain coat?

Nothing is better than going nanga.

Mort Walker · Post by **Mort Walker** » 08 May 2013 21:59

^^^The problem is analogous to, could some one make a rain coat that you could wear in the bath that would be more enjoyable than taking a shower without a rain coat?

Nothing is better than going nanga.

Prem · Post by **Prem** » 09 May 2013 01:16

http://www.wpcentral.com/nokia-teases-m ... elhi-event

Nokia teases more colour with upcoming New Delhi event

Nokia has a number of events lined up for the month of May, including a global announcement tomorrow in New Delhi, India. The company has published a new post on its official blog to detail webcast times for those unable to attend. What are we expecting to see from the company? We'll put some money down on new Asha products to follow the recently announced Asha 210 or even a low-end Lumia Windows Phone.Should you be interested in tuning in tomorrow (May 9th), here are the provided times:
•New Delhi: 12.00pm

•Helsinki: 9:30am
•London: 7:30am

BhairavP · Post by **BhairavP** » 09 May 2013 08:10

^Nokia Asha 501 mostly.. redesigned S40 + 1 HW button.

hanumadu · Post by **hanumadu** » 09 May 2013 16:42

BhairavP wrote:^Nokia Asha 501 mostly.. redesigned S40 + 1 HW button.

Seems to be new platform based on Smarterphone OS. Its being called the Asha software platform. Is this the EOL for S40?

Linux fans can rejoice. Smarterphone is based on Linux.

apoorv · Post by **apoorv** » 09 May 2013 17:07

It is said that new platform in Nokia Asha has been derived from Meego OS.

hanumadu · Post by **hanumadu** » 09 May 2013 17:16

^^The new OS is based on Smarterphone but the UI is based on the swipe concepts first seen on N9 which was running meego OS.

Raja Bose · Post by **Raja Bose** » 09 May 2013 22:04

So the Meego swipe UI lives on....

Nokia Asha 501 Hands On

Anujan · Post by **Anujan** » 12 May 2013 04:40

Just like how Sammy has a phone for every size, Nokia has a phone for every number.

Nokia Lumia 72, 83.5, 446.336 .... sometimes I think FM stations run their branding.

Raja Bose · Post by **Raja Bose** » 12 May 2013 05:46

GB's numbers indicate where in the food chain the phone is. So if X < Y that means phone Lumia X is cheaper/more budget/lower capability than Lumia Y. Sammy uses names instead of numbers to differentiate but I still haven't figured out the logic behind the naming (maybe they use a hat

).

member_20292 · Post by **member_20292** » 12 May 2013 14:54

best naming convention is HTCs.

no one can figure it out, whether One is better than One X or One V or Y or N or Q.

True "they who must not be named"

anybodys got personal experience with the HTC One? Like it?