Saturday, 7 October 2017
Thursday, 5 October 2017
Saturday, 2 September 2017
Tuesday, 29 August 2017
Wednesday, 9 August 2017
The 5 Technologies We Need to Change the World
I just finished reading an interesting hard science fiction book called The Punch Escrow, by Tal M. Klein (a movie is in the works).
What makes the difference between hard and soft science fiction is that hard science fiction is based on science, while soft is, let's just say, far more imaginative. To be honest, I enjoy both types, and the soft stuff is a ton easier to write. Those pesky physical rules don't get in the way, and you don't have to do research.
The story takes place several decades in the future, and it revolves around the idea of quantum foam and teleportation. It points out why teleportation never may be practical, but it brings up the idea of human 3D printing, which could be used more effectively for space exploration.
However, it also would have a massive number of other uses, both good and bad, which got me thinking about what else could change our future in a massive way. I came up with a list of five potentially world-changing technologies.
I'll close with my product of the week: a book on management that could have a massive effect on your company's success, based on the black boxes used in airplanes. It's called Black Box Thinking.
Technology 1: Organic Printing
We can use 3D printers for plastics, ceramics, metals and some blends, but our efforts even to print food have been more in line with automated icing machines for cakesthan printing food.
If we could print food affordably using nonperishable components, it would mean not only that we would be better able to address the massive amount of global hunger that exists, but also that we potentially could cut the cost of food manufacturing and eliminate most food-borne illnesses.
There is an amazing amount of activity in this area, suggesting that by 2030 we actually might have something like the Star Trek replicator in our homes.
Given that this same technology likely could manufacture drugs and better prosthetics, this single step could have a massive impact on how we live -- far beyond the way we eat.
Technology 2: Advanced Bio-engineering
A division of Google is releasing millions of bio-engineered mosquitoes to eliminate those that carry sicknesses. Granted, I do remember that many apocalyptic movies start this way.
The ability to manufacture insects that can address certain problems could have a massive impact, good and bad, on our environment. The bad would come from a mistake, or if someone decided to create militarized mosquitoes.
In the world of The Punch Escrow, there are mosquitoes that have been engineered to eat pollutants in the air and pee H20 -- and characters have to dodge constant pee drenchings from the mosquitoes.
Still, bio-engineered life forms could offset much of the damage we've done to the world -- addressing global warming as well as land, sea and air pollution -- and go places that people currently are unable to go.
Technology 3: AI Salting
Artificial intelligence salting is another concept author Klein introduces as a major plot element in The Punch Escrow. AI salting isn't meal preparation, for when we humans eat AIs (boy, talk about a concept that could start a Terminator event) it means a specialized technician teaches an AI to think more like a human.
Basically, it is individual AI deep learning of human behaviors. The underlying concept, making computers think more like humans, is critical to make them more effective at interacting with humans and interfacing with us more effectively.
If we really can't tell the difference between an AI and a human, or if an AI handling a human-related task could be made to be empathetic, the improvement in the interaction and the effectiveness of the AI would be improved vastly.
However, few are focused on the human part, and the challenge to train AIs to be more human-like could change forever the way we interact with and use them. At the very least, it would be a huge step in creating robots indistinguishable from humans and making the Westworld experience real.
Technology 4: Ultracapacitor Batteries
As Elon Musk repeatedly has said, batteries suck. Ultracapacitors can be charged and discharged almost instantly. They don't have the level of temperature problems that batteries currently exhibit. They are much lighter, which increases efficiency in things like cars, and their life cycle is vastly longer than current batteries.
The problem is, they don't do a good job of storing energy for any length of time. Some recent promising news from the scientific community suggests we may be close to sorting this out.
Batteries that could charge instantly and produce far more energy without problems would be a huge step toward making off-grid home power and electric-powered cars far more convenient.
Technology 5: Wireless Power
Ever since Nikola Tesla started talking about being able to broadcast power, it has been a known game-changer. Granted, Tesla may have gotten his ideas from aliens, but if you don't need batteries, then electric cars, planes, trains and personal electronics become smaller and far more reliable.
Qualcomm is working on a technology called "Halo", initially to charge electric cars without having to plug them in. However, its vision includes putting this technology in roads so that you'd never have to charge your car again -- it would charge while you were driving.
Rather than replacing a gas pump with a far slower charging station, you would just get rid of it. While not as good as true broadcast power, technology like this could work in cars, planes and offices, and we would never have to worry about charging our personal stuff or cars ever again.
A similar technology from WiTricity is being used to develop wireless charging for all our devices and currently being built into Dell's laptop charging docks.
Wrapping Up
Put these technologies together, and we'd have our food coming to us anyplace in any form and at any time we wanted. We'd have bugs making the world a better place to live.
AIs would be our friends -- not the problem Elon Musk is envisioning (though I kind of question his idea that government should fix this, given how bad it is at fixing things), or they'd just be much better at "taking care" of us -- but not in a good way.
Finally, if we can get better energy storage and distribution, we end up in a far more reliable and less-polluted world, coming damn close to a future Utopia. Though, as The Punch Escrow points out, if we can't fix ourselves, the result still could be pretty nasty.
Just think of the implications of printing people... As the only sure thing about the future is that it will be very different than the world of today, here's hoping that is a good thing.
Summer is the time I get caught up on my reading, and after reading The Punch Escrow, I moved to another recommended book that is far more practical. Black Box Thinking is based largely around comparing the healthcare industry to the airline industry, and pointing out that airlines have become massively safer over the years. However, hospitals may be the third biggest killer of people, largely because airlines have black boxes.
The reason this hits home for me is that it points to hospitals as places where errors are covered up aggressively to avoid liability. Black boxes, which capture errors but can't be used in litigation, are used to determine fault -- not to assign blame, but to ensure that the mistake never happens again. This one practice has helped transform air travel from one of the least safe ways to travel to one of the safest.
The big takeaway is that if you and your company can focus more on mistakes as learning opportunities and on ensuring that they are one-time events, rather than focusing on shooting the poor sap who made the mistake, which is much more typical, you'll end up not only with a far less hostile working environment, but also a far more successful company.
One of my big personal concerns is that we'll transfer this process of blame and covering up mistakes to our coming wave of ever-more-intelligent machines, which could speed up the related problems to machine speed. I doubt we'd survive that.
So, a book that makes workplace environments better, companies more successful, and humans more likely to survive is worth reading, I think, and it's my product of the week.
CoreOS, OCI Unveil Controversial Open Container Industry Standard
CoreOS and the Open Container Initiative on Wednesday introduced image and runtime specifications largely based on Docker's image format technology.
However, OCI's decision to model the standard on Docker's de facto platform has raised questions. Some critics have argued for other options.
Version 1.0 provides a stable standard for application containers, according to Brandon Philips, CTO at CoreOS and chair of the OCI Technical Oversight Board.
Having a standard created by industry leaders should spur OCI partners to develop further standards and innovation, he said.
Reaching the 1.0 mark means that the OCI Runtime Spec and the OCI Image Format Spec now are ready for broad use. Further, this achievement will push the OCI community to help stabilize a growing market of interoperable pluggable tools, Philips added.
The industry-supported standards also will provide a sense of confidence that containers are here to stay, he said, and that Kubernetes users can expect future support.
"The outcome is really good. The certification process is under way now," Philips told LinuxInsider.
Collaboration Challenges
Open standards are key to the success of the container ecosystem, said Philips, and the best way to achieve standards is by working closely with the community. However, reaching agreement on version 1.0 was more time consuming than expected.
"Early on, the biggest challenge was coming to terms with the model of how the project releases would work and how to get the project off the ground," Philips recalled. "Everyone underestimated how much time that would take."
Coalition members dealt with mismatched expectations about what they wanted to do, he said, but in the last year or so, the group got the referencing expectations done and more testing came through.
Quest for Standards
CoreOS officials began discussing the idea for an industry-approved open standard for the container image and runtime specifications several years ago. That early quest led to the realization that agreeing on a standard image format was critical, Phillips said.
CoreOS and container technology creator Dockerannounced OCI's formation in June 2015. The coalition started with 21 industry leaders forming the Open Container Project (OCP) as a non-profit organization seeking minimal common standards for software containers for cloud storage.
The coalition includes leaders in the container industry -- among them, Docker, Microsoft, Red Hat, IBM, Google and The Linux Foundation.
OCI's goal is to give high confidence to application developers that the software deployed in their containers will continue to work when newer specifications come out and people develop new tools. That confidence must be met for both proprietary and open source software.
It does not matter if the tools and applications are proprietary or open source. With the specifications in place, the products can be designed to work with any container configuration, Philips said.
"You need a conscious effort to create standards outside of people writing code. It is a separate effort," he added.
As part of the coalition, Docker donated its de facto image format standard technology to the OCP.
It included the company's container format, runtime code and specifications. Work on creating an Open Container Initiative Image Specification began last year.
The standards milestone gives container users the capability to develop, package and sign application containers. They also can run the containers in a variety of container engines, noted Philips.
A Choice of One?
The coalition faced two ways to pursue open standards, observed Charles King, principal analyst at Pund-IT. The first option was to gather like-minded people to hash out differences and build standards from scratch.
The coalition members seemed to settle for the second option, which involved adopting a powerful, market leading platform as an effective standard, he said.
"Docker's contributions to The Linux Foundation put the OCI firmly on the second path -- but those who are concerned about Docker's approach or its market position may feel there are better options," King told LinuxInsider.
In fact, one OCI member -- CoreOS -- leveled some strong criticism of the group's general direction at the beginning of the effort, he said, "so it will be interesting to see how V1.0 does/doesn't address those concerns."
Faster Path
Docker's widely deployed runtime implementation is a suitable foundation for building an open standard. It already was a defacto standard, according to David Linthicum, senior vice president at Cloud Technology Partners.
"It's also important that we get this working for us quickly. The waves of standards meetings, dealing with politics and things such as that, just waste time," he told LinuxInsider.
Right now, though, there are no better options, Linthicum added.
The runtime Docker uses is runC, which is an implementation of the OCI runtime standard, according to Joe Brockmeier, senior evangelist for Linux Containers at Red Hat.
"So, runC is a suitable foundation for a runtime standard, yes. It is broadly accepted and forms the basis for most container implementations today," he told LinuxInsider.
OCI is far more than Docker. While Docker did commit the underlying code from which the OCI specification is derived, the lineage stops there, said Brockmeier, and no truly viable alternatives exist.
Docking the Question
Adopting an industry-wide standard likely will simplify and speed container adoption and management for many companies, suggested Pund-IT's King. It also is likely that some key vendors will continue to focus on their own proprietary container technologies.
"They'll argue that theirs is a superior path -- but that will effectively prevent the OCI from achieving market-wide leadership," he said. "Starting out with a standard that's more or less complete, as OCI has, may not perfectly please everyone, but it's likely to move forward to completion more quickly and effectively than other options."
Containers have standardized deployment to cloud, with Docker clearly leading the way, said Marko Anastasov, cofounder of Semaphore.
Docker's de facto standard container does represent the best foundation for developing an open standard, he said.
"How Docker's commercial interests will influence the scale of its involvement in OCI remains to be seen," he told LinuxInsider.
Opposing Viewpoint
An open standard is not the end-all for adopting more containers in cloud deployment, contended Nic Cheneweth, principal consultant with ThoughtWorks. A better approach is to look at the impact of the server virtualization segment of the IT industry.
"The principal driver for continued growth and widespread adoption was not in the statement of an industry standard but in the potential and realized efficiencies obtained by use of any of the competing technologies, such as VMware, Xen, etc.," Cheneweth told LinuxInsider.
Aspects of container technology, such as the container itself, lend themselves to definition of a standard. Until then, healthy competition guided by deep open source software involvement will contribute to be a better standard, he said.
A standard around the orchestration of containers is not particularly important to the continued growth of the space, according to Cheneweth.
However, if the industry insists on locking into a de facto container standard, the model OCI chose is a good starting point, he said. "I don't know that better options are available, but certainly worse ones exist."
Microsoft Rolls Out Linux Support in SQL Server 2017 Release Candidate
Microsoft on Monday announced the availability of its first public release candidate for SQL Server 2017, which includes full support for Windows, Linux and Docker containers.
SQL Server on Linux improves on earlier previews with several key enhancements, including active directory authentication; transport layer security to encrypt data; and SQL Server Integration Services that add support for Unicode ODBC drivers.
SQL Server 2017 has demonstrated faster performance than competitive databases or older SQL Server versions with new benchmarks, Microsoft said, including the world record TPC-H 1-TB non-clustered data warehousing benchmark achieved this spring using SQL Server 2017 on Red Hat Enterprise Linux and HPE DL380 Gen 9 hardware.
Among early adopters are Dv01, a financial technology startup, which began using an open source database on a rival's cloud, but got 15 times faster performance on SQL Server 2017, according to Microsoft.
Another customer, Convergent Computing, has moved some Tier 2 applications to inexpensive, white box servers using SQL Server 2017 on Linux.
Convergent started fiddling with SQL on Linux last year, said Rand Morimoto, president of the company.
After evaluating the initial performance of the new platform, Convergent added applications as Microsoft added functionality, he told LinuxInsider.
Convergent now has half a dozen business applications -- including financial systems data analysis and client communications management systems -- running successfully on Linux, Morimoto said.
The company can get the same performance while running these SQL/Linux instances on lower power, he said, allowing it to recapture resources allocated to other applications.
"We're anticipating over the next 24 months that will translate to a decrease in costs of at least 34 percent," Morimoto said. "As we scale our SQL instances using SQL/Linux, we feel we can effectively increase capacity by 40 to 50 percent without having to allocate more resources."
Advantage Azure?
The move is the latest indication of Microsoft's growing warmth toward open source. CEO Satya Nadella has shifted the company's philosophy away from competition and toward cooperation.
"It recognizes that forcing customers onto Windows servers isn't a good long-term play for a couple of reasons," noted Rebecca Wettemann, vice president of research at Nucleus Research.
For starters, Microsoft is laser-focused on bringing customers to Azure, making the desire to fight the Windows server battle less important, she told LinuxInsider.
"Windows is still a dominant player in the enterprise server space," Wettemann acknowledged, but "with the growth of people writing for Linux, and the growth of Linux servers in the enterprise, Microsoft recognizes that customers want to run their database of choice on Linux -- so opening SQL to Linux puts Microsoft into the hands of more potential users."
The release seemed timed to coincide with last week's announcement of Microsoft's Azure Stack, which likely will have a huge impact on the private cloud market, noted Paul Teich, principal analyst at Tirias Research.
The availability of SQL Server on Linux will ensure that Microsoft customers that don't want a full off-premises commitment to the Azure cloud have a familiar -- and licensed -- database available for on-premises deployment of Azure Stack, he told LinuxInsider.
"It's a half step for customers who want to modernize their code a little bit but have concerns that keep them out of a multitenant cloud," Teich said. "They want the flexibility to take advantage of a hybrid cloud if needed."
Customer Demand
There is demand for stable, scalable databases, said Ron Pacheco, director of product management for Red Hat's platform business unit -- though he could not speak to any specific level of demand that might be driving Microsoft's strategy.
"Customers are openly embracing the open source development model, even for their own internally developed applications," he told LinuxInsider, "as it leverages the collective intelligence of a large community of mutually interested parties that yields very fast innovation."
Microsoft first announced plans to offer SQL on the Linux platform early last year, noting that it was important to make the database available across multiple platforms. Enterprise customers have gravitated toward Linux not only for its relatively lower costs, but also for its fewer security risks.
"Linux is the dominant operating system of the cloud and of open source tools and platforms," said Doug Henschen, principal analyst at Constellation Research.
Microsoft needed to make Azure attractive to cloud developers across platforms, he pointed out.
"Microsoft had to fill that gap," Henschen told LinuxInsider, "and indeed, it recently introduced a MySQL cloud service -- the open source database -- that runs on Linux."
There are several takeaways from Microsoft's announcement of SQL Server 2017 RC1, suggested Al Gillen GVP for software development and open source at IDC.
"Microsoft is truly supporting a mixed open source/Windows support model," he told LinuxInsider.
Further, "SQL Server on Linux is a competitive alternative to Oracle on Linux, giving customers that want an enterprise-quality database that is commercially supported on Linux two choices," Gillen pointed out.
Clearly, he said, "there is room to grow the SQL Server business into the Linux space."
Subscribe to:
Posts (Atom)