Fairfax Doubles Stake in RIMM

[StubbleJumper]

All those investments you mention were leveraged companies where it was far from clear whether the assets were greater than the liabilities, and whether the equity would retain any value whatsoever after restructuring. 

 

You are right that all of those companies were levered, but it should be noted that leverage is only one type of risk to which companies are exposed.  There is also operational risk, which for RIMM is enormous (ie, if people hate 2 or 3 generations of their products, the company could be toast and 2 or 3 generations might be only 5 years!).  Just like with RIMM, all of the previous FFH train wrecks that I listed except ABH were not restructuring, but were rather acquired as going concerns.  The market and FFH believed that there was some value there, but ultimately we have seen how it's turned out.  With a technology company, it could be argued that there is even less certainty than with the other five companies that I listed, and it might be farther outside of FFH's circle of competence.

 

 

There is a world of difference with RIM, and I don't understand why anyone would lump RIM into that category of investments.  You are essentially saying that the liquidation value of RIM, which I believe is greater than $0 if you run the numbers conservatively and don't believe there is fraud involved, will decrease over time, even with Watsa on the board. 

 

I completely disagree. 

 

And I hope you are correct.  However, as I suggested, the history of information technology is littered with the carcasses of very successful companies that seemingly had a liquidation value above zero and many of these companies were run by some pretty sharp cats.  While the market for smart phones is growing rapidly, RIMM's market share appears to be dropping almost as rapidly.  Going forward five years, the market for smart phones will be much closer to its saturation point....and then what happens if the trend for RIM's market share does not change trajectory?  What you have in that case is a reprise of  WordPerfect, which was also a product that was a market leader, seemed to have liquidation value and ended up bleeding market share over 5 or 6 years.  In reality, there's not much that Prem can do about that.  Watsa cannot help RIMM make a decent product, he cannot help them maintain a reliable messaging system, and he cannot meaningfully help them with marketing.  At most he can provide advice on corporate finance, and hopefully steer them away from making value-destroying acquisitions. 

 

But perhaps I misspoke by talking about "permanent loss of capital."  I agree that if RIMM is only worth $5 and you paid $20 for your share, you have suffered a permanent loss of capital.  My main point was that to treat RIM as a complete loss is either extremely conservative or disingenuously stated in order to freak people out regarding the potential impact on FFH's NAV.

 

Let's look at this in another way.  If you cannot predict the smart phone market size, RIM's market penetration or margins in 5 years, how can you arrive at a reasonable estimate of value?  Unfortunately most of us end up with a wild-ass guess which could range from a Corel-type result to a Google-type result depending on which ill-informed assumptions that we use.  For people like me, it falls into Mr. Buffett's "too hard" pile.  Since Prem has bought a large slug of this on our behalf, we need to at least understand the worst plausible case.....and unfortunately the worst plausible case is a heavy permanent loss of capital (history of tech investments shows that a zero is unlikely over the next 5 years, but not inconceivable.  A large permanent impairment in value is not unusual).

 

FWIW, we also had these types of discussions about the CDS.  Many of us also "wrote off" the investments in CDS because we didn't know how to value them.  Personally, I also write off FFH's lawsuit against the shorts....  I write off the investments in Irish banks...  The difference between RIMM and some of the other investments that were subject to great uncertainty is that most of those other investments were at least situated close to FFH's circle of competence, so at least investors could console themselves that Prem was an expert in the area he was investing in.

 

 

SJ

[tooskinneejs]

Stubblejumper said: "What you have in that case is a reprise of WordPerfect..."

 

Eight to ten years ago, people probably made that same analogy when discussing Apple's stock. 

[StubbleJumper]

Stubblejumper said: "What you have in that case is a reprise of WordPerfect..."

 

Eight to ten years ago, people probably made that same analogy when discussing Apple's stock.

 

 

 

Yep.  See post #25.

 

 

SJ

[Guest misterstockwell]

I think RIMM could easily go to zero in the chase for relevance. Their moats that once protected them are being filled in and breached. Simple software can take the place of almost everything they once held as sacred. I see no moat at all. BBM is easily duplicated, and folks like Whatsapp have the ability and the customers. They are surpassing a billion messages a day. It is much simpler for people to buy a cool phone and get a free app, and tell their friends to get the free app, than to buy an ugly phone and be forced to use BBM. Any phone, any network.  The secure aspect is also easily replicated. Software firms like MobileIron, Zenprise and Good Technology(to name a few) can provide secure enterprise mobile environments with any device on any platform. App developers have abandoned ship faster than Captain Schettino. I don't see anything that makes RIMM relevant any more.

[BargainValueHunter]

Sorry if this has already been posted...

 

http://www.tgdaily.com/business-and-law-brief/61107-former-rim-ceo-buys-50-million-in-rim-stock

 

Lazaridis's plan to purchase an additional $50 million in stock to supplement his already vast investment in the company shows that he has no hard feelings and remains a loyal steward of RIM, but it also sort of underlines the problem he had as co-CEO.

[Guest Hester]

Stubblejumper said: "What you have in that case is a reprise of WordPerfect..."

 

Eight to ten years ago, people probably made that same analogy when discussing Apple's stock.

 

Maybe Comparing RIMM now to Apple ten years ago doesn't work because the two company's CEO's are different. 10 years ago Apple had some blockhead named Steve Jobs at the helm. On the other hand, currently RIMM has genius visionary Thorsten Heins at the helm.

 

No comparison

 

The two are like apples and oranges palms.

[Uccmal]

I expect we are going to see Rim shift much more to an enterprise model, with less emphasis on handsets going foward.  It already looks as though they are going to release vastly fewer choices with BB10.  The car in Vegas was an example of where they are headed.  QNX/BB10 is already in use in a number of applications requiring robust software, less prone to crashing.

 

Not an investor in Rimm right now, except on a look through basis. 

[Guest valueInv]

 

The two are like apples and oranges palms.

That was a good one  ;D ;D

[tooskinneejs]

Hester said: "Maybe Comparing RIMM now to Apple ten years ago doesn't work because the two company's CEO's are different. 10 years ago Apple had some blockhead named Steve Jobs at the helm."

 

The world's view of Steve Jobs now is much different than it was when everyone thought Apple was doomed to fade away or remain only a niche player and, as a result, the stock was barely trading for more the company's net cash per share.

 

History is most useful when not revised.

[txlaw]

You are right that all of those companies were levered, but it should be noted that leverage is only one type of risk to which companies are exposed.  There is also operational risk, which for RIMM is enormous (ie, if people hate 2 or 3 generations of their products, the company could be toast and 2 or 3 generations might be only 5 years!).  Just like with RIMM, all of the previous FFH train wrecks that I listed except ABH were not restructuring, but were rather acquired as going concerns.  The market and FFH believed that there was some value there, but ultimately we have seen how it's turned out.  With a technology company, it could be argued that there is even less certainty than with the other five companies that I listed, and it might be farther outside of FFH's circle of competence.

 

Sure, but I'm not arguing that there isn't operational risk with RIMM or that RIMM won't crash and burn in terms of market share.  Nor am I arguing with the general statement that with technology companies, there is less certainty regarding the future. 

 

What I'm arguing is that to say that RIMM equity is worth $0 is quite a statement -- a hyperbolic one, IMO.  The reason the distinction must be drawn between RIMM and Canwest (and the other investments you mentioned) is that in those cases you mentioned, there was a real question as to whether all the assets were worth more than all the liabilities. 

 

Now it just happens to be that I think RIMM is probably worth where it's trading at if not more, but that's a discussion we can have another time.

 

And I hope you are correct.  However, as I suggested, the history of information technology is littered with the carcasses of very successful companies that seemingly had a liquidation value above zero and many of these companies were run by some pretty sharp cats.  While the market for smart phones is growing rapidly, RIMM's market share appears to be dropping almost as rapidly.  Going forward five years, the market for smart phones will be much closer to its saturation point....and then what happens if the trend for RIM's market share does not change trajectory?  What you have in that case is a reprise of  WordPerfect, which was also a product that was a market leader, seemed to have liquidation value and ended up bleeding market share over 5 or 6 years.  In reality, there's not much that Prem can do about that.  Watsa cannot help RIMM make a decent product, he cannot help them maintain a reliable messaging system, and he cannot meaningfully help them with marketing.  At most he can provide advice on corporate finance, and hopefully steer them away from making value-destroying acquisitions. 

 

Same comments from above, although I cannot comment specifically on the value of WordPerfect -- I didn't even know that it was a company.  I do know that the product WordPerfect is still used in some places, including in the legal profession.  So the intangible assets of the product itself is still worth something, if not much.

 

To reiterate, you're mixing and matching two things: decline in the business with whether the assets are worth more than the liabilities.

 

Let's look at this in another way.  If you cannot predict the smart phone market size, RIM's market penetration or margins in 5 years, how can you arrive at a reasonable estimate of value?  Unfortunately most of us end up with a wild-ass guess which could range from a Corel-type result to a Google-type result depending on which ill-informed assumptions that we use.  For people like me, it falls into Mr. Buffett's "too hard" pile.  Since Prem has bought a large slug of this on our behalf, we need to at least understand the worst plausible case.....and unfortunately the worst plausible case is a heavy permanent loss of capital (history of tech investments shows that a zero is unlikely over the next 5 years, but not inconceivable.  A large permanent impairment in value is not unusual).

 

FWIW, we also had these types of discussions about the CDS.  Many of us also "wrote off" the investments in CDS because we didn't know how to value them.  Personally, I also write off FFH's lawsuit against the shorts....  I write off the investments in Irish banks...  The difference between RIMM and some of the other investments that were subject to great uncertainty is that most of those other investments were at least situated close to FFH's circle of competence, so at least investors could console themselves that Prem was an expert in the area he was investing in.

 

Again, you seem to be making a general case against RIMM with regards to the uncertainty of its businesses, particularly the selling of Blackberry hardware.  However, I'm suggesting that the safe way to approach RIMM is to look at it from a liquidation/runoff/break up perspective and try to determine how much it is worth dead or broken up versus alive. 

 

That could be $5 per share.  Or it could be $20 per share.  It's very difficult to make the case it's worth $0 per share.  And it certainly isn't appropriate to make the case that it's worth $0 per share by talking about how its going concern value erodes with market share loss.

 

Put it another way: if RIMM were a Graham net-net with no debt, would you be making the same arguments as to why RIMM is worth $0?  I doubt it. 

[Guest Hester]

Hester said: "Maybe Comparing RIMM now to Apple ten years ago doesn't work because the two company's CEO's are different. 10 years ago Apple had some blockhead named Steve Jobs at the helm."

 

The world's view of Steve Jobs now is much different than it was when everyone thought Apple was doomed to fade away or remain only a niche player and, as a result, the stock was barely trading for more the company's net cash per share.

 

History is most useful when not revised.

 

Apple stock more than doubled in the month after Jobs took over as CEO. So perhaps the world's view isn't quite what you think. Of course, at that time he had either founded or co-founded three large successful companies, while Thorsten Heins's biggest accomplishment to date is convincing a multi-billion dollar company to let him be CEO.

 

If you want to compare RIMM to a distressed mobile phone maker, why people pick Apple is beyond me. Actually it's not, it's just survivorship bias. There are hundreds of similar companies that have been distressed, and instead of turning it around went bankrupt or were taken under. Heins was even already a passenger on one shipwreck.

[Ben Graham]

[01/30/2012 08:30 PM]

 

Intel Corp. has signed an agreement with RealNetworks to purchase approximately 190 patents and 170 patent applications worldwide, as well as next-generation video codec software, for $120 million. The acquisition enhances Intel's ability to offer richer experiences and solutions to end users across a spectrum of devices, including through notebooks, smartphones and digital media.

 

http://www.xbitlabs.com/news/multimedia/display/20120130203043_Intel_Buys_Video_Codec_Video_Streaming_Technologies_from_RealNetworks.html

 

RIMM will need to obtain patent rights to compete.

 

It seems like a wise move for Prem Watsa with his new board position at RIMM, would be to partner with IBM & Level 3. The companies entered into a long-term patent cross-license agreement.

 

http://level3.mediaroom.com/index.php?s=23600&item=51470

 

EDIT: Under the terms of the agreement, IBM granted Level 3 licenses to IBM's approximately 42,000 pending and issued patents which cover a broad range of telecommunications services and technologies. In turn, Level 3 granted IBM licenses to those of Level 3's more than 850 pending and issued patents which cover a broad range of information handling systems. The licenses will last as long as the lives of the respective patents. Other terms of the cross-license agreement were not disclosed.

 

"We have long admired IBM's legacy of innovation," said Jack Waters, chief technology officer of Level 3. "This patent cross-license agreement provides Level 3 the opportunity to take advantage of IBM's innovation as we continuously expand the range of services we provide our customers."

 

"IBM enjoys a widely recognized reputation of innovation and invention for technologies related to telecommunications, networking, media and entertainment," said Dan Cerutti, IBM's general manager of intellectual property. "With this cross license, we look forward to deepening our relationship with Level 3, and reaping with them all the benefits of collaboration."

 

[StubbleJumper]

Again, you seem to be making a general case against RIMM with regards to the uncertainty of its businesses, particularly the selling of Blackberry hardware.  However, I'm suggesting that the safe way to approach RIMM is to look at it from a liquidation/runoff/break up perspective and try to determine how much it is worth dead or broken up versus alive. 

 

1) No, I am not making a general case against RIMM.  What I have done is offer a perspective on the potential for a loss of capital.  I do the very same thing for other investments.

 

2) Liquidation/runoff/break-up is only a relevant valuation method if you believe that management would actually do it or that management actually could do it.  With respect, I would suggest that this view of the would would be patently wrong with RIMM.  I'd be very surprised if management liquidated in the face of difficult business conditions.  More likely they would throw more money into it.  And that's the problem with this type of company.  You can fritter away shareholders' capital chasing a market that you've already lost. 

 

 

Put it another way: if RIMM were a Graham net-net with no debt, would you be making the same arguments as to why RIMM is worth $0?  I doubt it.

 

 

3) I never suggested that RIMM is currently worth $0.  I simply suggested that a heavy, permanent impairment of capital is a realistic outcome.  And I suggested that the stock does not need to go to zero for that to occur (heck, even at $5/sh we'd lose what?  Like 75% of invested capital?).  Zero is just the very worst possible case, and it does not strike me as very likely.

 

4) Never say never, but I would probably not buy a tech net-net.  A tech company can be a net-net one year and could be far worse than a net-net five years later due to hubris in management.  Lack of debt and low share prices do not mean that management will drive a positive outcome for owners.  Unfortunately, in tech, it's possible to fritter away large amounts of shareholders' capital quickly.

 

 

I hope this one works out well (which it might!), but I will not be investing in RIMM for my personal account!

 

 

SJ

[SharperDingaan]

There is another way to approach it:

 

Assume tangible equity is worth +/- $2.50. Bad decisions, write-offs, etc. burn off the cash cushion. Positive number, average between $0-5, & existing primarily because of HW presence on the board. The range is a best guess, but if they stopped selling everything today - took the inventory write-off, paid the termination penalties & severance settlements, etc - & were unable to sell any P&E - how much cash would be left?

 

Assume the average patent has a 2 yr 1/2 life of 50%. 4K of patent value is worth 1K at the end of Yr4. They are primarily a single sku company & at an advanced stage in the product life cycle. Most would expect the patents to be currently worth maybe 25% of their peak value - & about 13% ongoing obsolescence for the next 2 years.

 

Development: Lot of talent, patents, & financial strength remain. Even with 75% of the talent gone, the odds are good that there will be totally new lines, product, etc < 5 yrs. They have a very large yolk, & the odds get even better if they drop out of the public eye, go at their own pace, & shed the majority of their analyst following. No more analysts, notebooks, blackberry, etc - the buggys in a automobile world. New product & lines, as in not done yet, & minimal chance of additional financing being required. It does happen, & often (Apple).

 

Impossible to value this future development today, but it clearly has material IV. We also know that industry valuation based on sales must continue to push RIM down, as they lose more share in a pie that has increasingly limited growth. A spur driving technology, mo-mo & technical investors to sell to value investors - & further depress RIM.

 

Take the long view & RIM looks reasonably attractive, but you wouldn't buy it yet as the blood is only just starting to sweep the floor.

 

SD

 

 

 

[racemize]

Assume the average patent has a 2 yr 1/2 life of 50%. 4K of patent value is worth 1K at the end of Yr4. They are primarily a single sku company & at an advanced stage in the product life cycle. Most would expect the patents to be currently worth maybe 25% of their peak value - & about 13% ongoing obsolescence for the next 2 years.

 

You probably were doing this on purpose, but the average patent length being 2.5 years seems very conservative.  Patents last 20 years and a quick search shows them with 2500 issued patents (thought some of them are design patents, which may be of limited use).  For example, the first few I looked at had priority dates >2005.  Even on the 200th patent, the priority date was 2007.  If the patent website weren't being so slow this morning, I could give a better estimate. 

[SharperDingaan]

Racemize: We're speaking to the approximate value of the patent at a point in time following inception, & assuming 50% declining balance amortization every 2 yrs. For a 20yr patent with an inital capitalization of 1,000,000:  Yr 0 1,000,000; Yr 2 500,000; Yr 4 250,000, Yr 6 125,000, Yr 8 62,500, Yr 10 31,250; Yr 12 15,625; Yr 14 7,812; Yr 16 3,906; Yr 18 1,953; Yr 20 0

 

A $1M patent (Blackberry) capitalized in 2005, would have run 6 yrs to the end of 2011 & be worth about 125K, but its future obsolescence cost/yr is small. More notable is that the cost curve closely resembles the inverted decay curve of a long dated LEAP - an opportunity after the intersection point for those so inclined.

 

SD 

[txlaw]

Again, you seem to be making a general case against RIMM with regards to the uncertainty of its businesses, particularly the selling of Blackberry hardware.  However, I'm suggesting that the safe way to approach RIMM is to look at it from a liquidation/runoff/break up perspective and try to determine how much it is worth dead or broken up versus alive. 

 

1) No, I am not making a general case against RIMM.  What I have done is offer a perspective on the potential for a loss of capital.  I do the very same thing for other investments.

 

2) Liquidation/runoff/break-up is only a relevant valuation method if you believe that management would actually do it or that management actually could do it.  With respect, I would suggest that this view of the would would be patently wrong with RIMM.  I'd be very surprised if management liquidated in the face of difficult business conditions.  More likely they would throw more money into it.  And that's the problem with this type of company.  You can fritter away shareholders' capital chasing a market that you've already lost. 

 

 

Put it another way: if RIMM were a Graham net-net with no debt, would you be making the same arguments as to why RIMM is worth $0?  I doubt it.

 

 

3) I never suggested that RIMM is currently worth $0.  I simply suggested that a heavy, permanent impairment of capital is a realistic outcome.  And I suggested that the stock does not need to go to zero for that to occur (heck, even at $5/sh we'd lose what?  Like 75% of invested capital?).  Zero is just the very worst possible case, and it does not strike me as very likely.

 

4) Never say never, but I would probably not buy a tech net-net.  A tech company can be a net-net one year and could be far worse than a net-net five years later due to hubris in management.  Lack of debt and low share prices do not mean that management will drive a positive outcome for owners.  Unfortunately, in tech, it's possible to fritter away large amounts of shareholders' capital quickly.

 

 

I hope this one works out well (which it might!), but I will not be investing in RIMM for my personal account!

 

 

SJ

 

SJ, my original post was directed at a post saying that we should assume RIMM is worth $0. 

 

Not only is it very unlikely that RIMM is worth $0, but anyone who says so is almost insulting HWIC and Prem Watsa.  If HWIC cannot prevent management from burning cash and assets such that the assets are worth less than the liabilities, then they're nowhere near as good as I think they are.  Now, you may not have been arguing that RIMM is a $0 (and I even said later that reasonable people can disagree on what RIMM is worth given an uncertain future).  But you appeared to be overstating your case or responding to something I was never arguing (although at some point, I might try to provide a case for why RIMM is worth X amount).

 

Additionally, what is your evidence to say that management would "throw more money into it" without any concern for ROI?  If you look into RIMM, you will see that the management is not unfocused on shareholder value.  It is that they messed up on the strategic issues and design issues.  That is the pitfall of being in the technology sector.  You can be very focused on getting good returns for shareholders, but you can make bad decisions and you can be outcompeted.  Take a look at Mr. Softie, which still is worth more than $0.

 

Liquidation or break up of RIMM may not happen.  Indeed, the new CEO has suggested that he will keep RIMM together.  That does not mean that RIMM is not worth more than the sum of the parts or that the sum of the parts equals a negative number.

 

 

 

[racemize]

Racemize: We're speaking to the approximate value of the patent at a point in time following inception, & assuming 50% declining balance amortization every 2 yrs. For a 20yr patent with an inital capitalization of 1,000,000:  Yr 0 1,000,000; Yr 2 500,000; Yr 4 250,000, Yr 6 125,000, Yr 8 62,500, Yr 10 31,250; Yr 12 15,625; Yr 14 7,812; Yr 16 3,906; Yr 18 1,953; Yr 20 0

 

A $1M patent (Blackberry) capitalized in 2005, would have run 6 yrs to the end of 2011 & be worth about 125K, but its future obsolescence cost/yr is small. More notable is that the cost curve closely resembles the inverted decay curve of a long dated LEAP - an opportunity after the intersection point for those so inclined.

 

SD

 

Ah, I see, I understand the original sentence now.  However, I'm not sure that using such a model makes sense for patents (when considering real value and not capital schedules), particularly tech ones.  Even with a low future life (assuming still greater than 2 years), if it is a strong patent, it can still be very valuable.  In other words, I'm not sure the patent value decays that fast.  For example, I don't think a patent's value would decrease at all (or at least not that fast) for the first 10 years.  I'd model it as having a steady value for a long period and then a precipitous decline.  Now valuing patents in the first place--that's tough. 

[SharperDingaan]

2 yrs is at the long end of the development cycle. Beginning of Yr3 your competitor introduces their shite competition to your device, but there's still a big gap. Beginning of Yr5 your competitor introduces their 2G, & the gap is much smaller. When they get to 4G, you're the one with the shite device.

 

Each generation usually takes less time than the last. Assuming 4G is achieved in around 6-8 yrs, todays Blackberry would be the shite device - & it appears to be exactly that, versus its competitors. Observing the strength of RIMs competition today, maybe our 2yr time frame should really be 18 months.

 

A patent is worth the PV of its future CF. The problem with tech patents is that once the product life-cycle peak is over, its almost impossible to materially extend the product life by driving down widget costs, & selling at just over cost - you simply get a warehouse full of obsolete & unsellable shite widgets that no-one wants.

 

The good news is that the patent may still have 10yrs+ to run. If you can reuse the patent in a totally new & different product line, you will do very well - & get a deterrent thrown in as well. As at about the time your competitors reach 3G/4G your patent expires, & every copycat, world-wide, will use it to flood the market. If you are the first mover, & the patent has less than 5-6 yrs to run, you'll probably get most of the profit for yourself. 

 

SD

[racemize]

right, the underlying assumption there is that the patent is directly tied to your products.  As a patent agent, a large number of software/tech patents that I write and get issued are fairly generic and are not specific to individual products--thus, they can have a much longer useful life.

[Ben Graham]

BlackBerry London's makeover gets leaked

 

by Dara Kerr | January 31, 2012 8:35 PM PST

 

http://news.cnet.com/8301-1035_3-57369382-94/blackberry-londons-makeover-gets-leaked/?tag=mncol;cnetRiver

 

 

Ten things RIM's new CEO must do right away:  http://news.cnet.com/8301-30686_3-57364084-266/ten-things-rims-new-ceo-must-do-right-away/?tag=mncol;8n

 

 

Things BlackBerry MUST Do!

2012 is the make-or-break year for the BlackBerry; here's how to make it the former.

 

http://cnettv.cnet.com/things-blackberry-must-do/9742-1_53-50119187.html?tag=api&tag=nl.e758

[Ben Graham]

Prem Watsa, make sure your new devices have DASH.

 

DASH promises stutter free streaming video over LTE.

 

We've all been there: fire up a clip from YouTube or a movie on Netflix and things start out great. But, then, after just a few moments, that LTE connection starts to give up the ghost and suddenly you're faced with unbearable stutturing or a video that just dies mid stream. Researchers at the Fraunhofer Institute for Telecommunications are looking to solve that conundrum with DASH, or Dynamic Adaptive Streaming over HTTP. The idea is actually surprisingly simple -- files of different sizes and qualities will be available depending on signal strength and network load, and the stream will be able to seamlessly switch between them as these variables change.

 

http://www.engadget.com/2012/02/05/dash-promises-stutter-free-streaming-video-over-lte/

[Guest valueInv]

Prem Watsa, make sure your new devices have DASH.

 

DASH promises stutter free streaming video over LTE.

 

We've all been there: fire up a clip from YouTube or a movie on Netflix and things start out great. But, then, after just a few moments, that LTE connection starts to give up the ghost and suddenly you're faced with unbearable stutturing or a video that just dies mid stream. Researchers at the Fraunhofer Institute for Telecommunications are looking to solve that conundrum with DASH, or Dynamic Adaptive Streaming over HTTP. The idea is actually surprisingly simple -- files of different sizes and qualities will be available depending on signal strength and network load, and the stream will be able to seamlessly switch between them as these variables change.

 

http://www.engadget.com/2012/02/05/dash-promises-stutter-free-streaming-video-over-lte/

Apple, Adobe, Microsoft, Netflix, Akamai, etc have already implemented different versions adaptive streaming. So most of the videos you are watching on your phone already have it. DASH is simply an attempt to standardize it and it doesn't look like its gaining much traction.

 

[Ben Graham]

Prem Watsa specializes in applying a decentralized approach to running the Fairfax business. Is he now applying a robust, decentralized approach to RF-based localization for upcoming new RIMM devices?

 

Existing approaches to radio frequency (RF) -based localization are centralized (i.e., they require either a central server or the user’s roaming node, such as PDA or laptop, to compute the user’s location) and/or use a powered infrastructure. In a fire, earthquake, or other disaster, electrical power, networking, and other services may be disabled, rendering such a tracking system useless. Even if the infrastructure can operate on emergency generator power, requiring wireless connectivity is impractical when a potentially large

number of wireless access points may themselves have failed (e.g., due to physical

damage from fire).

 

In addition, most previous approaches are brittle in that they do not account for

lost information, such as the failure of one or more transmitters, or perturbations in RF signal propagation. As such, existing approaches are inappropriate for safety-critical applications, such as disaster response, in which the system must continue to operate (perhaps in a degraded state) after the failure of one or more nodes in the tracking infrastructure.

 

In this paper, we present a robust, decentralized approach to RF-based localization,called MoteTrack. MoteTrack uses a network of battery-operated wireless nodes to measure, store, and compute location information. Location tracking is based on empirical measurements of radio signals from multiple transmitters, using an algorithm similar to RADAR. To achieve robustness,

 

MoteTrack extends this approach in three significant ways:

 

First, MoteTrack uses a decentralized approach to computing locations that runs on the programmable beacon nodes, rather than a back-end server.

-

Second, the location signature database is replicated across the beacon nodes themselves in a fashion that minimizes per-node storage overhead and achieves high robustness to failure.

-

Third, MoteTrack employs a dynamic radio signature distance metric that adapts to

loss of information, partial failures of the beacon infrastructure, and perturbations in the RF signal.

 

 

http://www.eecs.harvard.edu/~mdw/papers/motetrack-loca05.pdf