“Caution is the confidential agent of selfishness,” Woodrow Wilson said in a speech in 1909. If not for the fact that film was in its infancy in 1909, you’d be forgiven for thinking Mr. Wilson was talking to a room full of video software engineers.

Truth is, the evolution of the quality of video experience has been, until quite recently, a rather cautious, self-interested affair.

The delivery chain tasked with moving a video from Point A to Point B is fragmented, with each player concerned with the narrow welfare of its own task and its own equipment, with little regard for the impact its decisions have on other links in the chain.

Content creators. Broadcasters. Cable companies. Internet providers. Display manufacturers. Each looks after its own part in the task of making a viewer happy. Few consider how their actions affect others. And nobody takes responsibility for ensuring the overall chain is optimized with a viewer’s eyes in mind.

Deciding What ‘Better’ Means

The problem, ultimately, is that each player defines in its own terms what “better” is going to be. It has been forever thus. The result, unfortunately, is a less-than-optimal experience at the end of the chain — which is to say, for the poor individual sitting on a couch, trying in vain to stream a football game.

Fortunately, with the multitude of delivery solutions, with the new era of cloud-based delivery, and with the improvement in communication infrastructure through wireless and better display devices, technological capabilities are improving and with them, quality of video experience.

But as significant as the improvement has been, there are still enormous gaps that prevent greater advances from being realized.

First and foremost, delivery chain participants have to begin considering quality from the point of view of the end-user’s satisfaction, as defined by a single metric, one that takes into account all elements of the delivery chain. It’s not good enough, for example, for a display manufacturer to tout the latest development in 4K technology, or the number of pixels available to a viewer, when compression, frame rate, brightness have all conspired to leave the viewer less than happy.

The end result, quite aside from an unhappy customer, is wasted time, wasted effort and expectations left unmet — particularly those of the content creator. Crucially, investment is left unprotected.

At the end of the day, quality of experience, by definition, is about the quality of the viewer’s experience, not that of the cable company. Or the camera manufacturer. Or the internet provider.

By agreeing on a single overall metric, delivery chain participants are aligning their goals with a final objective — to make video as faithful to the creative intent as possible.

Doing so — solving for viewer experience — requires answering several fundamental questions in rapid fashion, ideally using an automated process.

1. What is the quality metric in use? Does it tell the entire story? You can go to a doctor, but if the diagnosis isn’t correct, the information isn’t of any use.

2. Where are the problems? In what part of the delivery chain?

3. When did the problem occur? Was it during primetime? Was it when a certain kind of content was being viewed? Data tells a story.

4. Why did the problem occur?

5. How can a solution be achieved?

The answers to those questions need to take into account not just individual elements, but the entire distribution chain. With the ever-growing combinations of sources, formats and devices, the time is right for a uniform metric that considers every element that contributes to viewer experience, but more importantly defines it in an absolute sense, so everybody in the delivery chain knows the part they play in the quality score.

As I said earlier, there is progress being made. Video is being delivered, at every step, much better today than in the recent past.

Metrics Will Help

But if you, a video delivery chain participant, knows what would improve the overall result – Decision A versus Decision B – you can accelerate that improvement by an order of magnitude. In effect, you have the power to bring the future to the present.

Is it a difficult problem to solve? Yes. But early work with unified quality metrics is showing that significant results can be achieved.

Soon, caution and selfishness in video QoE won’t know what hit them.

Dr. Abdul Rehman Ph.D. is CEO of SSIMWAVE, a Waterloo, Ontario-based video technology vendor.

Cable techies are looking around "the fog."

At a Silicon Valley conference early this month, computer and communications experts delved into the IoT and "the fog," a derivative of cloud computing. CableLabs is evaluating the fog's role, especially in emerging wireless services, and Comcast awarded an "Innovation Fund" grant recently to a Princeton University "Cloud to Fog" research project. 

"Fog is 'distributed cloud,'" according to Don Clarke, principal architect at CableLabs, who noted that the concept emerged a couple years ago as a Cisco marketing term. Other experts acknowledged that "fog" and "edge" are often used interchangeably, but they are synergistic – not synonymous.

In a seminal academic paper on "The Fog Computing Paradigm" three years ago, Ivan Stojmenovic and Sheng Wen of Deakin University in Australia described scenarios in which fog services "can be hosted at end devices such as set-top boxes or access points."

By the researchers' reckoning the fog lies between the cloud and the edge. Fog computing is envisioned for use in wireless services, including mobile voice, Internet of Things and connectivity for autonomous vehicles and other new networks.

"Fog can be distinguished from cloud by its proximity to end users ... and its support for mobility," Stojmenovic and Wen wrote. "As fog computing is implemented at the edge of the network, it provides low latency, location awareness and improves quality-of-services for streaming and real-time applications."

Major developments have been taking place quietly since the paper's publication. The 400 attendees at the Fog World Congress in Santa Clara, Calif., Oct. 31-Nov. 1 examined evolving trends in fog computing, including increasing deployments of fog as an early-stage technology and as a solution to Internet of Things latency.

For example, at a session titled "Fog Over Denver," Traci Hiltonberry, director of innovation for the Denver South Economic Development Partnership, explained the creation of a "national model for fog computing and networking in a smart city ... [for] transportation and mobility, public safety, resilience and resource conservation, smart buildings and public health."

Another speaker forecast that fog computing will become an $18.2 billion market by 2022. Christian Renaud, research director-IoT at 451 Research, predicted that fog's primary uses will be in the utilities and energy sectors, followed by transportation, healthcare and industrial. He said he foresees revenue models growing by 37% from 2018 to 2022.

Meanwhile: Fog Forming in the Cable Industry
The potential for this "fog" capability at the network edge is on the minds of several cable operators, although they declined to provide status reports or timetables for implementing the architecture.

"The edge is the new piece of the cloud puzzle," CableLab's Clarke told Multichannel News. He said that identifying "latency and bandwidth efficiency between the cloud and the end user" is a critical issue in edge development.

"It's about relationships," he said, noting that there are "very different dynamics" as various sectors in the network operations business explore "how standards and open-source can become symbiotic."

Clarke said the "edge is where the connectivity provides the latency you need for the service experience," citing issues such as "where does that service terminate and where does the content get delivered?" He eschewed the word "fog," focusing instead on edge computing.

"These days you can put a lot of functionality [into] base station and other less complex, lower power consumption" devices, he said.

"Wireless is an important consideration as we architect these new networks at 5G [fifth generation] and beyond," Clarke said.

Meanwhile, Comcast said that with its grant -- of an unspecified sum -- to the Princeton University project, the company sought to support research into the "cloud-to-fog interface in the areas of storage, communications and management."

In announcing the grant, Jason Livingood, VP-technology policy & standards at Comcast Cable, called fog networking "an architectural approach that seeks to make networks more efficient by pushing network intelligence and processing capabilities closer to end users." Livingood said that fog computing processes would enable the "cloud and edge [to] form a mutually beneficial, interdependent continuum" that would eliminate or minimize the need to determine if a specific task should be handled in the cloud or at a customer's edge device.

"This project highlights the challenges and solution approaches in building a unified interface framework between edge and cloud under the fog-networking paradigm," Livingood said. A Comcast spokesman told Multichannel News that there is no specific timetable for a report on the Princeton research; he characterized the fog exploration as "pretty new."

At Princeton, the project is being supervised by Dr. Mung Chiang, an electrical engineering professor, founder of the Princeton EDGE Lab and a co-founder of the global Open Fog Consortium. Among the supporters of Princeton's EDGE Lab are ARM, Cisco, Dell, Intel and Microsoft.

Chiang has said he helped launch the consortium to address common problems with “edge networks” — the connections at the periphery of a more centralized network, close to the actual devices that use the network. In a Princeton profile of his work, Chiang said, “As we further develop the ‘Internet of Things’ — networked devices in smart cities or connected cars — we have a unique opportunity to bring the ‘cloud’ closer to the edge and users as ‘fog.’”

Photo by d3sign/Getty Images

Going all-Internet protocol is all the rage. Pay TV providers are doing it and over-the-top video services have already done it.

The same holds true for TV programmers looking to put all of their operations and other interworking components under an overarching umbrella of IP technology that will enable them to “virtualize” their infrastructure and develop and launch an array of new, more personalized services.

Turner is well along that path, having started its long, sometimes difficult journey about two years ago.

COMING TOGETHER

In addition to setting the stage for new types of services, migrating a legacy serial digital interface (SDI)-based infrastructure to one underpinned by IP will provide Turner with an operational link between two groups — broadcast and broadband — that have traditionally worked in relatively separate silos. The IP path extends a bridge to those two parts of the business.

“We made a decision as it related to the technology … to essentially look at our network organization function as a feed factory, and that feed factory can spit feeds out in IP , it can spit them out in SDI,” Jeremy Legg, Turner’s chief technology officer, said. “We’re largely indifferent to the format that those feeds actually go out in.”

In addition to providing an operational boost, going with an underlying IP infrastructure will enable Turner and other programmers to start to think differently about their traditional linear TV businesses that have historically been the venue of broadband and OTT.

“That gets [us] into things like personalized streams; it gets into things like personalized content and gets into the targeting of ads and promos and content, and assembling them in certain ways for certain audience segments,” Legg said.

With that as the basis, it’s conceivable that a programmer like Turner could turn one version of Cartoon Network, for example, into multiple versions targeted to different audiences based on shows specific viewers watch and with targeted, more-relevant ads.

“Ultimately, [this transition] leads you to having more versions of a network, not fewer,” Legg said.

But pulling that off using legacy broadcast infrastructure historically hasn’t been possible because it wasn’t IP -based. “You couldn’t write the software to do it,” he said, noting that Turner has begun to hook fiber cables into an IP router in its broadcast plant, and that new IP -based cameras will help to expand and unify the entire video production ecosystem.

IP REVOLUTION

Going with an IP infrastructure, he said, will enable Turner and others to develop software-based systems that can handle elements such as dynamic ad insertion, the multiplexing and personalization of networks that run on an IP -based master control and playout system. That in turn, will support the playout of personalized networks either on-premises or via the cloud.

“In many ways, I think, some of the biggest technology revolutions are actually happening in broadcast,” Legg said.

But this transition isn’t going to happen overnight. Though Turner has been at it for more than a year, there’s still at least a couple of years left to get the job done, Legg estimates.

Some big pieces of the IP puzzle have been completed, though.

Turner, for example, recently signed a deal with Amazon Web Services that moves its physical content into the public cloud, rather than storing everything on-premises. With those master files in the cloud, Turner can then determine, for example, which assets are ad-supported and contain Nielsen watermarks to satisfy a wide range of distribution and service requirements.

Last year, Turner, along with Fox Networks Group and Discovery Communications, made a strategic investment in SDVI, to help them build out media supply chains and ensure that their content is being formatted correctly. Also factoring in is Turner’s acquisition of iStreamPlanet, which plays a key role in how Turner’s content is packaged and encrypted for various streaming platforms.

Another important element was getting the organization to move in the same direction.

“Getting to a place where we had a common vision of where we wanted to go, was probably harder than the technology itself,” Legg said. “But once you get disparate groups all rowing the oars in the same direction, then you can start to move pretty fast.”

Though Turner is developing a hybrid system that relies on cloud and on-premises playout, the trend ultimately will lead to an all-cloud playout approach that could help to seed the move toward more personalized program streams.

LIVE COMES LATER

Some networks, such as Cartoon Network and TC M, don’t rely heavily on live programming or use constantly changing on-screen graphics, but instead lean on playlist-based programming schedules. That makes them possible initial candidates for cloud-based playout.

“We’ll initially look to migrate some networks out that are less complicated to play out,” Legg said. “We’re not in position to start playing CNN out of the cloud yet.”

But networks that are better suited for cloud-based playout could shift to audience-based targeting and tap into usage and viewing data that can help to “influence” the makeup of a personalized network playlist, Legg added.

In 2016, we witnessed the explosion of global video across industries, with no signs of this growth slowing anytime soon. IBM estimates cloud-based video will be a $105 billion market opportunity by 2019. As video-streaming services have expanded and improved, cloud-based video has evolved into a necessary feature of any successful business for reaching internal and external audiences.

As this expansion continues, a few key trends are likely to emerge in 2017. The foundation of each of these trends is data — not just tapping into it, but owning it, and diversifying the data sources used for business decisions. The players that unlock data about user behavior on their platform and apply analytics to gain new, intelligent insights will thrive in the competitive market.

In the media and entertainment spheres, 2016 was the year content owners went direct to consumer. This coming year will be a race to enhance those services and improve the consumer experience to retain and grow audiences. In the enterprise space, 2017 will be the year that more businesses view themselves as content-providers and begin to adopt video as a core tool for communicating with employees, customers and business partners.

As we enter this pivotal time in the video industry, the top cloud video trends to watch for 2017 include the following.

Mergers will change the landscape of the streaming video industry: As the streaming video market becomes increasingly crowded, consolidation within the industry is inevitable. In 2017, we will see more mergers and acquisitions for OTT services as some content-providers combine and others fall out.

Skinny bundles will get smarter: In 2016, skinny bundles were introduced to the television market. Next year, we’re sure to see their continued rise in popularity. Skinny bundles will be created using analytics that provide the right number of channels, the right combination of content, and the right price points for each individual consumer.

Artificial intelligence will play an integral role in the media & entertainment field: As more companies apply data and analytics to video and cognitive capabilities continue to grow, artificial intelligence will begin to play a role in media and entertainment. AI will be used to dramatically streamline editing processes and eventually produce an entire episode of television.

Cognitive capabilities in the cloud will shape product development: Focus groups will expand to include the use of audience-gauging cognitive technology. As companies apply analytic and cognitive technologies to the cloud, they’ll be able to gauge customer reaction of livestreamed product announcements in real time. These companies will tap into this smart data to adjust their products before they even hit the market.

Virtual reality will become real for the enterprise: In 2017, companies will utilize VR to take clients, employees, and business partners to places that would be difficult, if not impossible, without this technology. As these products permeate into the everyday lives of consumers, more mainstream business will explore these new opportunities to engage with and educate customers in more personal ways than ever before.

As we enter the new year, business leaders should prepare for these trends by examining their video strategy and determining whether they’re taking full advantage of data and analytics for all of their video needs. The cloud has opened new opportunities for content providers to reach viewers anywhere and anytime, through almost any device. To win over competitors, businesses will need to apply new technologies to the cloud to tap into the power of advanced data and analytics to gain actionable insights about video-viewing behavior.

David Mowrey is vice president of strategic planning at IBM Cloud Video. 

Charter Communications CEO Tom Rutledge said cable could go a long way toward solving its customer service problem by simply reducing the number of actual face-to-face interactions with customers and helping them solve their problems remotely.

“The inherent problem in all of cable, and always has been, is you have to schedule a job with a person who doesn’t really want you to come to their house and you have to do work of an indeterminate length of time and get to the next job on time,” Rutledge said at the Guggenheim TMT Symposium in New York. “That’s inherently difficult. And all of the business that do that -- plumbers, contractors – everybody can’t stand them, because it’s a difficult transaction to manage. The more you can take that out of the business, the higher the satisfaction goes, just inherently.”

Charter, he said is moving toward that future by putting more functionality in the cloud – its user interface for example, is backward compatible and works with a number of different set-top boxes.

“We want to put advanced user interfaces in all of our customers’ homes and not just in their homes but every outlet they have a TV connected to and every device that they have that they could watch video on," Rutledge said. "We can be state-of-the-art without a physical transaction.”

The Charter chief also had some thoughts on stricter federal regulation in the wake of its $78.7 billion agreement to purchase Time Warner Cable. That deal, expected to close by the end of the year, has yet to begin the regulatory approval process.

Rutledge said he agrees with the basic tenet of net neutrality. But he said that many of the concerns that former FCC chairman Julius Genacowski managed to alleviate through Title I, current FCC chairman Tom Wheeler is seeking to remedy through Title II.

“We agreed with Julius as an industry and as a company with the Title I net neutrality requirements that he wanted us to embrace,” Rutledge said. “Those are the same ones that Tom Wheeler is trying to get using Title II. We embraced the idea of openness to the Internet and everyone having an opportunity, without throttling, without prioritization to use the two-way interactive capability that broadband provides, We were not happy with Title II because of the inability to determine exactly how it will all unfold.”

Rutldge added that Wheeler and Genachowski’s objectives are basically the same, just that Wheeler is going to manage them differently because the process is now largely complaint-driven.

But he added there is a movement afoot that could bring more clarity to the situation.

“Legislation could be passed that would limit net neutrality to a particular set of facts and the risks of other kinds of regulatory creep through Title II; I would prefer that,” Rutledge said.

But Rutledge stopped short of predictions.

“I don’t know if it’s not possible or possible,” Rutledge said. “It’s still being worked on.”   

Sometimes the onslaught of daily deals and lush forecasts obscures their real importance: They are collective indicators of the industry's direction, overshadowing the big picture (a version of the forest/trees conundrum). While last month included a stunning array of "big deals" in telecommunications and media, their significance looms even larger when, as I've just done, you absorb the whole month's news in a single dose.

For reasons I'll explain below, I was off the grid for most of May, and have just been catching up. Wow, a lot happened in a few weeks, but more importantly, the aggregation of deals and analyses confirms the increasing velocity of the cable/telecom industry overhaul.

Obviously, the biggest deals, especially Charter-Time Warner Cable-Bright House Networks plus Altice-Suddenlink, underscore the consolidation of the industry, especially with Altice's appetite for more acquisitions on this side of the Atlantic. The "internationalization" of domestic cable also pokes through, notably with the role that Liberty Media/Liberty Global can play in bringing content deals across the ocean; Altice may do the same.

I encountered plenty of reports about shifting power - away from programmers and toward pay TV system operators, well covered by Mike Farrell in his "Game Changer" report about the Charter deal. I also skimmed through many interpretations about the impact on equipment makers as the MSO customer base narrows. Deals take time, and this sudden tsunami is affecting all segments of the industry.

Meanwhile, the research factories cranked out plenty of forecast fodder -- all showing the accelerating uptake of over-the-top video viewing. I enjoyed the IHS Infonetics prediction that pay TV operators will aggressively deploy OTT services during the next five years, spurred by DISH's Sling TV venture. The study also concluded that pay TV operators (including cable companies) will offer "skinny bundles" to stave off cord-cutting. Who knows exactly how it will play out, but amidst other predictions, the reports reaffirmed the importance of the OTT juggernaut.

And I saw reports from the digital "NewFronts" and TV Upfronts about the growth of "branded entertainment." I've been following BuzzFeed's aggressive multimedia agenda and was impressed to see it was listed along with Fullscreen and Stylehaul as leading proponents that will make branded entertainment a core offering to their advertisers.

If it works, such sponsored programming can substantially revamp the way that advertisers allot their media budgets -- another major distraction from "business as usual." 

Of course, there was plenty of coverage of "the cloud" as a key to the next wave of delivery, including Charter's vision (thanks to its recently acquired stake in ActiveVideo and its CloudTV service) for a world with limited set-top boxes, replaced by virtualized STBs via the cloud. Maybe we shouldn't embrace everything we read about the cloud -- but one month's heavy dose of cloud stories from several industries convinced me that it's happening now. And much more widely and more quickly than I had considered.

In addition to the daily deluge of forecasts about the growing digital penetration, there was the usual slew of deal-making amongst digital opportunists.  Again, the takeaway was the unrelenting flow of activity.  There were several stand-out developments -- possible game changers -- such as Facebook's launch of its "Instant Articles" feature, which lets publishers post videos and stories directly onto  the Facebook NewsFeed.  Among the blue-chip participants are NBC, BBC and the New York Times.  

Google's new plan to create Android TV "channels" opens the door to hybrid TV that blends traditional TV and online content.  It paves the way for viewers of smart TVs to flip easily between any program source available to them. That's another reminder that non-cable entities will be making more deals with content suppliers that once may have been potential cable content  program sources. The process appears to be underway.

Nielsen's purchase of Innerscope Research, a neuromarketing research company, was a reminder of the necessary new tools being brought into play to assess what viewers like about what they are watching. The deal, along with previous Nielsen investments in similar tools plus other neural research ventures I've encountered in the past few years, reminded me of how important it is for programmers and advertisers to know what viewers are thinking and feeling.  It may not be NSA-level invasiveness, but the personal research is becoming a very big priority.

On the dark side, the latest American Customer Satisfaction Index report, which puts cable (especially Comcast and Time Warner Cable) at the bottom of 43 industries wasn't fun reading. It was a humbling reminder of the competitive challenge ahead from wireless and other carriers when and if they have the capacity.

Maybe it was just my overwhelming, sudden immersion into so many aspects of the business that instilled a sense of a "pivotal moment."  The merry month of May 2015 could someday be identified as a changing point for the industry, or maybe it was merely an accelerator for all the daily developments that we see unfolding but can rarely put into context.

As for my personal reason for uncharacteristic obliviousness during May: I was in hospitals for nearly three weeks. No, not a heart attack or bone-crushing traffic accident. It was "just" a triple whammy of unknown origin: staph infection, pneumonia and blood/gastric problems. After great treatments at Northwestern University medical center in Chicago and later at a Johns Hopkins-affiliated hospital in Washington, D.C. near my home, I've recovered well and am nearly back to what passes for "normal" in my life.  Appropriately, I was felled in Chicago on the eve of INTX cable convention.  The title above was inspired by the movie While You Were Sleeping, a 1995 rom-com that appropriately unfolds in a Chicago hospital (although there was nothing funny or romantic about my hospitalization).

Although I had my tablet, smartphone and TV Everywhere account information with me, I cannot say much about the mind-distraction factors while trapped in uncomfortable beds for weeks. (Blame most of it on inadequate hospital WiFi technology.)  I could say a lot about the boredom of TV programming and the proliferation of automotive commercials, but I'll do that elsewhere. The experience did allow me to binge a bit, such as during AMC's Mad Men marathon prior to the series finale (less enjoyable without my DVR). And there's plenty of admiration for electronic medical records and other health technology, on which I will wax elsewhere.

My thanks to the hundreds of kind remarks and encouraging messages that came in when I posted a health status report on Facebook. Although I treasure all of the comments, and I smiled as I read each message, one posting in particular stood out.   

FCC chairman Tom Wheeler (whom I've known and worked with periodically since the 1970s, just two Ohio boys who came to Washington and got into '70s cable) sent me an "OMG" comment. 

Noting that I had to skip the INTX/NCTA convention, Wheeler said, "Those cable shows can be tough." I don't know if he was referring to what he had to handle when he was NCTA president or to the current reception he receives when he speaks to the NCTA audience (probably the latter).   

But I appreciated his point and replied, "You got THAT right about cable shows."

Now I can't wait until the next one -- and to all the industry overhauls between now and then. Based on the past month, fasten your seat belts. It's going to be an intense period of industry change.

You’ve no doubt heard quite a lot about how “the cloud” will revolutionize the way operators do business and you may be skeptical about whether the results will live up to the lofty goals. The reality is that the cloud is already living up to the hype and it is indeed poised to enable operators to do a lot more in the future.

Comcast and Cablevision Systems — cable’s “early adopters” of the cloud — are currently expanding their cloud digital video recorder services, and other operators are in various stages of evaluating and testing their own cloud DVRs. While initially touted primarily as a cost-effective way to offer network DVR services, it has quickly become apparent that having content reside in the cloud enables the delivery of a better DVR service.

For example, whereas set-top box-based DVRs can simultaneously record only as many programs as they have tuners, cloud DVRs open up the gates for much more flexibility. This can be structured to extend revenue to the operator by charging for levels of service for recording, storage and device delivery. Spain-based broadband provider Telefónica believes the subscriber benefits of its popular cloud DVR service are behind the substantial reductions in subscriber churn it is experiencing.

Reduced subscriber churn is a benefit that clearly every operator can understand and appreciate. 

The cloud has been widely deployed for providing complex search and recommendations based on user behavioral data, which also drives up operator revenue through VOD sales. Cloud UI solutions help preserve legacy set-top boxes in the home and keep the price down for new boxes. Features for shopping, applications and advertising all depend on cloud-based technology. These all expand service-based revenue and often drive down capital expenditures.

Additionally, with more and more programming content now residing in the cloud, operators can take advantage of Internet-protocol content delivery to HDMI dongles and other emerging customer premises equipment (CPE). Unlike traditional set-tops, these devices are much smaller, portable and consume much less power, enabling operators to deliver more services in the home and on the go.

Paradigm shift is an overused phrase, but it is applicable when it comes to the cloud. It is affecting virtually every facet of a service provider’s operations, from the headend to the back office, where operators can consider virtualizing their complete user administrative infrastructure with a flexible, scalable solution managed with open-source tools. And again, we have yet to tap its full potential.

The cloud does not just represent a better, lower-cost method to accomplish what was done before. It can dramatically change how the industry operates by improving efficiencies and enabling exciting new services.

Susan Crouse is director of product marketing at Broomfield, Colo.-based video set-top software and middleware maker and technology integration services provider Alticast.