Tech News

On Monday, Jean-Baptiste Kempf, the president of VideoLAN introduced a new AV1 decoder, named dav1d. This open source decoder aims to be small, fast, maintainable, correctly threaded, and cross-platform. What is AV1? AV1, short for AOMedia Video 1, is a new royalty-free video coding format designed for video transmissions over the internet. It is being developed by the Alliance for Open Media (AOMedia) composed of most of the important Web companies including YouTube, Google, Facebook, Netflix, Amazon, and Microsoft, among others. Back in September, the YouTube Developers account published a playlist of the first videos to receive AV1 transcodes. Technical details of dav1d The following are the technical details about dav1d: It is written in C99 without VLAs. This facilitates portability as some compilers still haven't implemented VLAs. It has asm in NASM/GAS syntax. It uses Meson/Ninja as build system. Meson and Ninja are open source build systems which are meant to be extremely fast. It supports x86, x64, ARMv7, ARMv8 architectures. It runs on Windows, Linux, macOS, Android, iOS making it cross-platform. It is released under BSD, so that it can be embedded anywhere, including non-open-source software; or even drivers, for hybrid decoders. Performance details of dav1d Currently, the reference decoder (libaom) for AV1 requires a lot of improvement as it is a research codebase. That is why, the VideoLAN and FFmpeg communities have come together to work on this new decoder, sponsored by the Alliance of Open Media. This project is partially funded by the Alliance for Open Media/AOM and is supported by TwoOrioles and VideoLabs. To show the performance improvements that dav1d comes with, the developers compared it with libaom: Smaller source code: The source code of dav1d is 1/10th of lines of code compared to libaom. Smaller binary size: Its weight is 1/3rd of the binary size of libaom. Smaller runtime memory footprint: The memory footprint of dav1d is 1/4th of memory footprint of libaom Stack usage is minimal: It uses a very limited amount of stack (about 1 KB). Faster than libaom 1.0.0, slower than libaom HEAD: Depending on the threads conditions it is approximately faster than libaom 1.0.0, but slower than libaom HEAD. You can find the source code of dav1d at VideoLAN website. Read more about dav1d at Jean-Baptiste Kempf’s website and also watch this presentation at Video Dev Days 2018. YouTube starts testing AV1 video codec format, launches AV1 Beta Playlist A new Video-to-Video Synthesis model uses Artificial Intelligence to create photorealistic videos Facebook Watch is now available world-wide challenging video streaming rivals, YouTube, Twitch, and more

Last week, California’s Governor Jerry Brown passed a bill giving rise to a new law that will ban automated accounts, more commonly known as bots, from pretending to be real people in pursuit of selling products or influencing elections. The bill was approved on September 28 and will be effective from July 1, 2019. As per the California Senate, "This bill would, with certain exceptions, make it unlawful for any person to use a bot to communicate or interact with another person in California online with the intent to mislead the other person about its artificial identity for the purpose of knowingly deceiving the person about the content of the communication in order to incentivise a purchase or sale of goods or services in a commercial transaction or to influence a vote in an election.” The law will assist in tackling social media manipulation to determine foreign interference. Bots caused major issues during the 2016 U.S. Presidential elections and have ever since grown to be a menace that platforms like Twitter have been trying to combat. The 2016 U.S. Presidential elections saw Russian-controlled bots playing an active role in manipulating opinions, retweeting Donald Trump's tweets 470,000 times, and Hillary Clinton's fewer than 50,000 times. The main aim of this effort is to target bots that spread misinformation. Twitter said that it took down 9.9 million potentially spammy or automated accounts per week in May and has placed warnings on suspicious accounts. Twitter has also announced an update in its work on its "election integrity" project, ahead of the US mid-term elections in November. These include updating its rules regarding fake accounts and sharing stolen information. It said it would now take into account stock avatar photos and copied profile bios in determining whether an account is genuine. Robert Hertzberg, a state senator from California who pushed for the new law forcing bots to disclose their lack of humanity, told The New York Times he was the subject of a bot attack over a bail reform bill. So he decided to fight bots with bots by launching @Bot_Hertzberg in January. As per California law, the account discloses its automated nature. “*I AM A BOT.*” states the account’s Twitter profile. “Automated accounts like mine are made to misinform & exploit users. But unlike most bots, I'm transparent about being a bot.” To know more about this California Senate ’s bill in detail, check out the Senate bill. Sentiment Analysis of the 2017 US elections on Twitter Facebook, Twitter takes down hundreds of fake accounts with ties to Russia and Iran, suspected to influence the US midterm elections DCLeaks and Guccifer 2.0: How hackers used social engineering to manipulate the 2016 U.S. elections

On Monday, Opera Touch was made available for iPhone users that aims to challenge Apple’s default browser, Safari. It particularly targets the newly launched Apple phones, iPhone XS and XS Max and the upcoming iPhone XR. Opera Touch is your on-the-go browser which helps you instantly search the web with one hand. It comes with Flow to make sharing links, images, videos or notes much easier. Instant searching Opera Touch opens up in search mode to help you instantly find things on the web. You can search using text input, voice, or a QR code. Also, the core elements of the interface are located at the bottom of the screen making them easy to reach. Seamless browsing experience with Flow You can use Opera Touch on iPhone together with your Opera computer browser for a seamless web browsing experience across devices. The plus point here is that you do not need to create an account to connect your iPhone with your computer browser. Simply, start the Opera computer browser and scan the QR code displayed there with Opera Touch and you are good to go. After connecting, you can send links, videos, and notes to yourself with a single click through Flow and they will be displayed across all your Flow-enabled devices. Your on-the-go browser It makes searching on-the-go easier as you can conveniently explore the internet using just one hand. The Fast Action Button present in the bottom middle of the screen provides the browser’s key functions, including access to your most recent tabs and search. You can hold and swipe the button open to switch to your most recent tabs, reload or close the page or send the current tab to your computer in Flow. Safe and secure browsing The data shared with Flow is fully end-to-end encrypted. Its opt-in ad blocker, blocks intrusive ads, making web pages load faster. Also, the browser comes with Opera’s cryptojacking protection that reduces the risk of your mobile getting overheated when you browse the web. Learns from your browsing patterns The browser will learn from your browsing pattern giving you a more personalized browsing experience. It will automatically add your favorite sites to the browser’s home screen. Read more about Opera Touch on Opera’s official website and to download the browser head over to the App Store. Vivaldi 2.0, a hyper-customizable browser, releases with Vivaldi Sync, Resizable Tab Tiling, and more! Firefox Reality 1.0, a browser for mixed reality, is now available on Viveport, Oculus, and Daydream Firefox Nightly browser: Debugging your app is now fun with Mozilla’s new ‘time travel’ feature

Last week, Google project zero criticized Ubuntu and Debian developers for not merging kernel security fixes fast enough and leaving users exposed in the meantime. The kernel community clarified yesterday on how it is making attempts to reduce and control the bugs in the Linux ecosystem by testing and kernel hardening. They acknowledge that there is not a lot the kernel community can do to eliminate bugs as bugs are part and parcel of software development. But they are focusing on testing to find them. Now there is a security team in the kernel community made up of kernel developers who are well versed with kernel core concepts. Linux Kernel developer Kroah Hartman said: “A bug is a bug. We don’t know if a bug is a security bug or not. There is a famous bug that I fixed and then three years later Red Hat realized it was a security hole”. In addition to fixing bugs, the kernel community will contribute to hardening to the kernel. Kernel hardening enables additional kernel-level security mechanisms to improve the security of the system. Linux Kernel Developer Kees Cook and others have made huge efforts to take hardening features that have been traditionally outside of the kernel and merge them for the kernel. Cook provides a summary of all the new hardening features added with every kernel released. Hardening the kernel is not enough, new features need to be enabled to take advantage of them which is not happening. A stable kernel is released every week at the official Kernel website. Then, companies pick one to support for a longer period of time for enabling device manufacturers to take advantage of it. However, Hartman observed that barring Google Pixel, most Android phones don’t include the additional hardening features, making all those phones vulnerable. He added that companies should enable these features. Hartman stated: “I went out and bought all the top of the line phones based on kernel 4.4 to see which one actually updated. I found only one company that updated their kernel,” he said.  “I'm working through the whole supply chain trying to solve that problem because it's a tough problem. There are many different groups involved -- the SoC manufacturers, the carriers, and so on. The point is that they have to push the kernel that we create out to people.” However, the big vendors like Red Hat and SUSE keep the kernel updated which have these features. The kernel community is also working with Intel to mitigate Meltdown and Spectre attacks. Intel changed its approach in how they work with the kernel community after these vulnerabilities were discovered. The bright side to this is that the Intel vulnerabilities proved that things are getting better for the kernel community. More testing is being done, patches are being made and efforts are put to make the kernel as bug-free as possible. To know more, visit the Linux Blog. Introducing Wasmjit: A kernel mode WebAssembly runtime for Linux Linux programmers opposed to new Code of Conduct threaten to pull code from project Linus Torvalds is sorry for his ‘hurtful behavior’, is taking ‘a break (from the Linux community) to get help’

Splunk have kicked off their .conf 2018 conference with a bang. Splunk announced plans to update their premium enterprise products - Splunk Enterprise and Splunk Cloud - powered with machine learning and analytics capabilities. These products will be initially made available for public use through beta programs, and will allow Splunk customers to work with large-scale data and extract useful insights out of them. The upcoming versions of Splunk Enterprise and Splunk Cloud will include the following features: Splunk Business Flow - This feature will give customers the ability to make smarter business decisions by understanding relevant trends and process flows. Splunk Data Stream Processor - This feature will let users evaluate, transform and perform predictive analytics on streaming data. Splunk Mobile & Splunk Cloud Gateway - This new feature will allow Splunk users to interact with a whole suite of Splunk products from their mobile devices. Leverage the power of Natural Language Processing and query Splunk through voice and text commands. Build effective cloud-native services with Splunk Developer Cloud. Splunk Data Fabric Search will allow Splunk users to perform large-scale search of real-time data. Use Augmented Reality (AR) to interact with data such as QR codes, UPC scanning, etc. In his keynote, Merritt also announced plans for Splunk MLTK, a Splunk-specific application to run AI use-cases. This tool will be able to interact with key open source AI libraries such as Tensorflow and Spark’s MLlib. With these capabilities, businesses will be able to perform large-scale analytics by harnessing AI in their Splunk applications and related projects. Splunk Chief Executive Doug Merritt was of the opinion that businesses are finding it hard to keep track of the ERP data inside their data centers. “If we really want to be successful, we’ve got to tap into this sea of data around the world, outside of our walls,” he said. “We are in the midst of the data revolution, and these product updates ensure the Splunk platform evolves as our world does to deliver business outcomes no matter the organization, team or dataset”. Read more Why should enterprises use Splunk? Splunk leverages AI in its monitoring tools Splunk Industrial Asset Intelligence (Splunk IAI) targets Industrial IoT marketplace

The Sourcegraph team announced that they’re making Sourcegraph, a self-hosted code search and navigation engine, available as open source, earlier this week. “We opened up Sourcegraph to bring code search and intelligence to more developers and developer ecosystems—and to help us realize the Sourcegraph master plan,” writes Quinn Slack on the announcement page. This Sourcegraph master plan involves making basic code intelligence ubiquitous (for every language, and in every editor, code host, etc.). It wants to focus on making code review continuous and intelligent. Additionally, they also hope to increase the amount and quality of open-source code. Sourcegraph comprises the following features: Instant Code Search: Fast global code search with a hybrid backend. This combines a trigram index with in-memory streaming. You can search in files and diffs in your code by just using simple terms, regular expressions, and other filters. Code intelligence: It offers code intelligence for many languages using the Language Server Protocol. It also makes browsing code and finding references on your code easier. Data Center: Once you grow to hundreds or thousands of users and repositories, you can graduate from the single-server deployment to a highly scalable cluster using the Sourcegraph Data Center. Integrations: It offers Integration with third-party developer tools via the Sourcegraph Extension API. Organizations that are already using the Sourcegraph navigation engine can upgrade to Sourcegraph Enterprise (previously called Data Center) to get a hand on features that large organizations need such as single sign-on, backups, and recovery, cluster deployment, etc. However, these additional features that come with the Enterprise edition are paid and not open source. “We're also excited about what this means for Sourcegraph as a company. All of our customers, many with hundreds or thousands of developers using Sourcegraph internally every day, started out with a single developer spinning up a Sourcegraph instance and sharing it with their team. Being open-source makes it even easier to start using Sourcegraph in that way”, explained the announcement page. For more information, check out the official announcement. Facebook open sources LogDevice, a distributed data store for logs Kong 1.0 launches: the only open source API platform specifically built for microservices, cloud, and serverless Uber’s Marmaray, an Open Source Data Ingestion and Dispersal Framework for Apache Hadoop

The Anaconda team announced the release of Anaconda Distribution 5.3.0 in a blog post yesterday. Harnessing the speed of Python, learning and performing data science and machine learning is all the more easy in this new update. Here is the list of new features in Anaconda 5.3.0 #1 Utilising Python’s speed Anaconda Distribution 5.3 is compiled with Python 3.7, in addition to Python 2.7 Anaconda installers and Python 3.6 Anaconda metapackages. This will ensure the new update takes full advantage of Python’s speed and feature improvements. #2 Better CPU performance Users deploying TensorFlow can make use of the Intel Math Kernel Library 2019 for Deep Neural Networks (MKL 2019) included in this upgrade. These Python binary packages will ensure high CPU performance. #3 Better Reliability The team has improved the reliability by capturing and storing package metadata for installed packages. The additional metadata is used by the package cache to efficiently manage the environment while storing the patched metadata used by the conda solver. #4 New packages added There are over 230 packages which have been updated or added by the team. #5 Work in Progress on the casting bug The team is working on the casting bug in NumPy with Python 3.7 and the patch is in progress until NumPy is updated. To know more about this release, you can head over to the full release notes for the Distribution. Arm releases free Cortex-M processor cores for FPGAs, includes measures to combat FOSSi threat PyTorch 1.0 preview release is production ready with torch.jit, c10d distributed library, C++ API

Mozilla announced a major update to Firefox Focus, a free and open-source privacy-focused browser by Mozilla for Android, iOS smartphones and tablets, yesterday. Firefox Focus allows you to stay “focused” by automatically blocking ads and trackers. The new updates explore new features, revamped design, and GeckoView Technology behind Firefox Focus for Android. New Features There are three new features added in Firefox Focus, namely, home screen tips, search suggestions, and Siri shortcut for iOS users. Home Screen tips It’s good news for people who like to keep tabs on feature releases. Now, you can see the core functionalities of Firefox Focus on the start screen. This will provide an overview of the whole range of possibilities that your privacy browser needs to offer. More than that, this new addition will not interrupt the usage at all and will get automatically refreshed after every click on the Erase button. All you need to do is just open the browser and a bunch of helpful feature recommendations will be presented in your preferred language on the start screen (Android). However, for iOS users, this feature is currently available only in the English language. Search Suggestions Search suggestions make the entire searching process more convenient. You can activate the feature by opening the app and select settings > “Search” > select the checkbox “Get search suggestions”. Now, keeping in mind the privacy of the Firefox Focus users, you don’t have to share what you’re typing in the address bar with your search provider. The search suggestions feature has been turned “off” by default, so you can choose whether or not you want to turn it on. Siri shortcuts for iOS users Siri shortcuts have been added to the updated Firefox Focus. You can now set and open a favorite website, erase and open Firefox, as well as erase in the background using Siri shortcuts. Mozilla’s aim with this was to improve the Firefox Focus experience for iOS users. Revamped Design Visual design for Firefox Focus browser has been completely optimized for the recently released Android Pie. New icons, a customized URL bar, and a simplified settings menu make Firefox Focus easier to use. It also offers a consistent user experience. New engine for Firefox Android users Focus has been using Android’s built-in WebView all this time, but it has some limitations. WebView is not designed for building browsers. To add next-generation privacy features to Focus, Focus developers require deep access to the browser internals. This is why, they have opted for their own engine, Gecko. Firefox Focus is now based on GeckoView, Mozilla’s own mobile engine, making it perfect for use in Android apps. GeckoView will help them leverage all of their Firefox expertise to build more compelling, safe, and robust online experiences. GeckoView will enable Mozilla to implement unique privacy-enhancing features in the future, such as reducing the potential of a third-party collection. For more information, check out the official Mozilla website. Mozilla releases Firefox 62.0 with better scrolling on Android, a dark theme on macOS, and more Firefox Nightly browser: Debugging your app is now fun with Mozilla’s new ‘time travel’ feature Firefox Nightly’s Secure DNS Experimental Results out

At the Xilinx Developer Forum in San Jose, Arm announced its collaboration with Xilinx, the market leader in FPGAs. The collaboration plans to bring the benefits of Arm Cortex-M processors to FPGA through the Arm DesignStart program, thus providing scalability and a standardized processor architecture across the Xilinx portfolio. Users can expect a fast, completely no-cost access to soft processor IP, while taking advantage of the easy design integration with Xilinx tools and comprehensive software development solutions to accelerate success on FPGA. These processors will enable embedded developers to design and innovate confidently, while benefiting from simplified software development and superior code density. In addition, products can be easily scaled on these processors, thanks to the support of the broadest technology ecosystem of software, tools, and services provided by the team. Arm for FPGA comes with the following benefits: #1 Maximum choice and flexibility Users can obtain an easy and instant access to Cortex-M1 and Cortex-M3 soft processor IP for FPGA integration with Xilinx products. They will not be charged any license fee or royalties for this access. #2 Reduced software costs The processor focuses on reducing software costs while obtaining maximum reuse of software across an entire OEM’s product portfolio on a standardized CPU architecture, scaling from single board computers through to FPGAs. #3 Ease of design The team has ensured an easy integration with Xilinx system and peripheral IP through Vivado Design Suite. They use a drag-and-drop design approach to create FPGA systems with Cortex-M processors. The extensive software ecosystem and knowledge base of others designing on Arm, will ultimately result in reducing the time to market for these processors #4 Measures to Combat FOSSi (free and open source silicon) threat Arm Cortex-M1 includes a mandatory license agreement, which contains clauses against reverse-engineering. The clause also prevents the use of these cores for comparative benchmarking. These clauses will assist Arm to enable IP holds up against the latest and greatest FOSSi equivalents (like RISC-V) when running on the same FPGAs. This is not the first time that Arm has raised its voice against FOSSi threats. Earlier this year, they had also launched an aggressive marketing campaign specifically targeting RISC-V. The Arm and Xilinx collaboration will enable developers to take advantage of the benefits of heterogeneous computing on a single processor architecture. To know more about this news, head over to Arm’s official blog. Meet ‘Foreshadow’: The L1 Terminal Fault in Intel’s chips SpectreRSB targets CPU return stack buffer, found on Intel, AMD, and Arm chipsets Qualcomm announces a new chipset for standalone AR/VR headsets at Augmented World Expo

Yesterday, Amazon announced that it will be increasing the minimum wage for all of its US employees - full-time, part-time, temporary (including those hired by agencies), and seasonal employees. They raised the wages of employees in the UK as well from £8 to £9.50 an hour. The $15 minimum wage will affect more a total of more than 250,000 Amazon employees and 100,000 seasonal temporary employees who will be hired across US this holiday season. This new policy is effective from November 1. Amazon’s move to increase wages comes after recent pressure from campaigns criticizing Amazon’s pay policies. US Senator Bernie Sanders publicly accused Amazon of underpaying its workers and introduced a bill called Stop BEZOS act last month. The bill also known as ‘Stop Bad Employers by Zeroing Out Subsidies’ was aimed to force Amazon and other large companies to increase the living wages paid to workers. Currently, an Amazon Warehouse Worker earns $13 on average an hour. The new wages will add $2 an hour, which may not look like much but will add about $4,000 to yearly wages of employees who were making $13 till now. Amazon currently offers restricted stock units (RSU) which cannot be used by employees. So, keeping in mind employee preferences, the RSU stock grant program in Amazon will be replaced by a direct stock purchase plan for hourly employees and customer service employees. Amazon said that they will also work towards getting Congressional support for increasing the federal minimum wage. The current rate, $7.25, was set nearly a decade ago and is low. Amazon is phasing out the incentive pay component and $15 will be an effective minimum wage which won’t require any additional targets to be met. The wage policy will affect Amazon’s hourly operations and customer service employees who also see an increase, including those already getting paid $15. This does not affect the current benefits Amazon offers its employees which include: A comprehensive healthcare plan including medical, dental, and vision cover Almost 20 weeks of paid parental leaves Matching 401k of employees Paying 95% of associates’ tuition in advance for courses in high-demand fields, irrespective of whether those jobs are at Amazon or not Training hourly associates in critical job skills like resume writing, effective communication, and basic computer operating knowledge Jeff Bezos, Amazon CEO stated: “We listened to our critics, thought hard about what we wanted to do, and decided we want to lead. We’re excited about this change and encourage our competitors and other large employers to join us.” To know more visit the Amazon website. IBM faces age discrimination lawsuit after laying off thousands of older workers, Bloomberg reports Microsoft demands its suppliers, contractors to offer employees paid family leave, or risk doing business with them It’s Day 1 for Amazon Devices: Amazon expands its Echo device lineup, previews Alexa Presentation Language and more

Google announced a new experimental game streaming service, namely, Project Stream, earlier this week. Google calls this project a “technical test” and has partnered up with Ubisoft, one of the most popular video game publishers, to stream their upcoming Assassin’s Creed Odyssey via Project Stream on Chrome. “We’ve been working on Project Stream, a technical test to solve some of the biggest challenges of streaming. For this test, we’re going to push the limits with one of the most demanding applications for streaming—a blockbuster video game,” writes Catherine Hsiao on the announcement blog post. Google points out that their major goal with Project Stream is for it to effectively stream AAA game titles. This is because the Google team is inspired by the technology that goes behind AAA video games. Additionally, working with a AAA game title is more challenging as opposed to working with a game that comprises less intense graphics. “Every pixel is powered by an array of real-time rendering technology, artistry, visual effects, animation, simulation, physics, and dynamics. We’re inspired by the game creators who spend years crafting these amazing worlds, adventures, and experiences, and we’re building technology that we hope will support and empower that creativity,” states the post.   Project Stream  With Project Stream, Google is working to ensure that latency stays minimal and the graphics of the game are not compromised when using its streaming service. “The idea of streaming such graphically-rich content that requires near-instant interaction between the game controller and the graphics on the screen poses a number of challenges.  When streaming TV or movies, consumers are comfortable with a few seconds of buffering at the start, but streaming high-quality games requires latency measured in milliseconds, with no graphics degradation,” adds Google. Google has made limited spaces available for users to try Project Stream, starting October 5. If you want to participate then you can apply on Project Stream’s official website. Participation is only open for the U.S. residents who are 17 years or older. For more information, check out the official announcement. Google Project Zero discovers a cache invalidation bug in Linux memory management, Ubuntu and Debian remain vulnerable Google announces new Artificial Intelligence features for Google Search on its 20th birthday Google announces the Beta version of Cloud Source Repositories

On Monday, the Kubernetes team announced that DigitalOcean, which was available in Early Access, is now accessible as Limited Availability. DigitalOcean simplifies the container deployment process that accompanies plain Kubernetes and offers Kubernetes container hosting services. Incorporating DigitalOcean’s trademark simplicity and ease of use, they aim to reduce the headache involved in setting up, managing and securing Kubernetes clusters. DigitalOcean incidentally are also the people behind Hacktoberfest which runs all of October in partnership with GitHub to promote open source contribution. The Early Access availability was well received by users who commented on the simplicity of configuring and provisioning a cluster. They appreciated that deploying and running containerized services consumed hardly any time. Users also brought to light issues and feedback that was utilized to increase reliability and resolve a number of bugs, thus improving user experience in the limited availability of DigitalOcean Kubernetes. The team also notes that during early access, they had a limited set of free hardware resources for users to deploy to. This restricted the total number of users they could provide access to. In the Limited Availability phase, the team hopes to open up access to anyone who requests it. That being said, the Limited Availability will be a paid product. Why should users consider DigitalOcean Kubernetes? Each customer has their own Dedicated Managed Cluster. This provides security and isolation for their containerized applications with access to the full Kubernetes API. DigitalOcean products provide storage for any amount of data.   Cloud Firewalls make it easy to manage network traffic in and out of the Kubernetes cluster. DigitalOcean provides cluster security scanning capabilities to alert users of flaws and vulnerabilities. In typical Kubernetes environments; metrics, logs, and events can be lost if nodes are spun down. To help developers learn from the performance of past environments, DigitalOcean stores this information separately from the node indefinitely. To know more about these features, head over to their official blog page. Some benefits for users of Limited Availability: Users will be able to provision Droplet workers in many more of regions with full support. To test out their containers in an orchestrated environment, they can start with a single node cluster using a $5/mo Droplet. As they scale their applications, users can add worker pools of various Droplet sizes, attach persistent storage using DigitalOcean Block Storage for $0.10/GB per month, and expose Kubernetes services with a public IP using $10/mo Load Balancers. This is a highly available service designed to protect against application or hardware failures while spreading traffic across available resources. Looks like users are really excited about this upgrade: Source: DigitalOcen Blog Users that have already signed up for Early Access, will receive an email shortly with details about how to get started. To know more about this news, head over to DigitalOcean’s Blog post. Kubernetes 1.12 released with general availability of Kubelet TLS Bootstrap, support for Azure VMSS Nvidia GPUs offer Kubernetes for accelerated deployments of Artificial Intelligence workloads  

Peer-to-peer file services like BitTorrent were seeing a decline in traffic over the last decade, with the increasing popularity of affordable streaming services like Netflix. The number of piracy streaming options fell and was said to be accounted for the services provided by the on-demand video services and strict anti-piracy laws. However, recent reports published by Sandvine, who has been keeping a close eye on the file-sharing traffic, suggested that the traffic for streaming file-services is now growing. The report also noted that BitTorrent was earlier evidently losing out on market share but is now emerging to be the leading file-sharing place with 97% share. Over the past few years, Netflix had emerged as the leader in the streaming market and was single-handedly responsible for hosting a wide variety of content. While this was happening, the market observed a decrease in upstream and downstream file sharing of content, popularly called the torrents. This went on to suggest that the streaming market was rising and piracy downloading was reducing. However, Netflix is no more the single most popular streaming services that exist. Other video streaming services like Amazon Prime videos, Hulu, and others have fragmented the market with popular content playing over multiple services. This seems to have re-introduced the traditional practice of file-sharing over the internet and BitTorrent has gained the market. Sandvine's report further stated the following in its report: “More sources than ever are producing ‘exclusive’ content available on a single streaming or broadcast service – think Game of Thrones for HBO, House of Cards for Netflix, The Handmaid’s Tale for Hulu, or Jack Ryan for Amazon. To get access to all of these services, it gets very expensive for a consumer, so they subscribe to one or two and pirate the rest.” File sharing involves uploading and downloading a part of a file over a peer-to-peer network. Reportedly, file sharing makes up for 3% of all download and 22% of all upload traffic. Here, BitTorrent enjoys the lion's share with a massive 97% of all upstream file-sharing traffic. Evidently, fragmentation in the subscription video-on-demand market is playing the major role. To know more about this in detail, check out the discussion thread on Hacker News. The cryptocurrency-based firm, Tron acquires BitTorrent Facebook Watch is now available world-wide challenging video streaming rivals, YouTube, Twitch, and more Implementing fault-tolerance in Spark Streaming data processing applications with Apache Kafka

Last week, Atlassian software company released their updated ‘Atlassian Software License Agreement’ and their ‘Cloud Terms of Service’, which would be effective from the 1st of November, 2018. Just as any general agreement, this one too mentions the scope, the users authorized, use of software, and so on. However, it has set up certain restrictions based on the performance of its software. As per the new agreement, benchmarking of Atlassian software is forbidden. Restrictions on benchmarking As per the discussion on the Atlassian Developer Community, Andy Broker highlighted two clauses from the restriction section have been highlighted, which includes: (i) publicly disseminate information regarding the performance of the Software Andy Broker, a marketplace vendor explains this clause as, “This sounds very much like the nonsense clause that Intel were derided for, regarding the performance of their CPU’s. Intel backtracked after being lambasted by the world, I can’t really understand how these points got into new Atlassian terms, surely the terms have had a technical review? Just… why, given all the DC testing being done ongoing, this is an area where data we gathered may be interesting to prospective customers.” (j) encourage or assist any third party to do any of the foregoing. Andy Broker further adds, “So, we can’t guide/help a customer understand how to even measure performance to determine if they have a performance issue in the “Software”, e.g. generating a performance baseline before an ‘app’ is involved? The result of this would appear sub-optimal for Customer and Vendors alike, the “Software” performance just becomes 3rd party App performance that we cannot ‘explain’ or ‘show’ to customers.” Why Atlassian decided to forbid benchmarking? As per a discussion thread on Hacker News, many users have stated their views on why Atlassian planned to forbid benchmarking on its software. According to a comment, “If the company bans benchmarking, then the product is slow.” A user also stated that Atlassian software is slow, has annoying UX, and it is very inconsistent. This may be because most of its software is built using JIRA which has a Java backend and is not Node based. Jira cannot be rebuilt from scratch only slowly abstracted and broken up into smaller pieces. Also, around 3 years ago Atlassian forked behind the firewall for two distinct products, multi-tenant cloud and traditional to get into the cloud sector with a view to attracting more potential customers, thus increasing growth. A user also stated, “Cloud was full buy into AWS, taking a behind the firewall product and making it multi-tenanted cloud-based is a huge job. A monolith service now becomes highly distributed so latency's obviously mounted up due to the many services to service interactions.” The user further added, “some things which are heavily multi-threaded and built in statically compiled languages had to be built in single threaded Node.js because everyone is using Node.js and your language is now banned. It's not surprising there are noticeable performance differences.” Another user has suggested, “the better way for a company to handle this concern is to proactively run and release benchmarks including commentary on the results, together with everything necessary for anyone to reproduce their results.” The user added that they can even fund a trustworthy neutral third party to perform benchmarking with proper disclosure of the funding. To read the entire discussion in detail, head over to Hacker News. Atlassian acquires OpsGenie, launches Jira Ops to make incident response more powerful Atlassian sells Hipchat IP to Slack Atlassian open sources Escalator, a Kubernetes autoscaler project

An open-source libre GPU project is under the works by Luke Kenneth Casson Leighton. He is the hardware engineer who developed the EOMA68, an earth-friendly computer. The project already has access to $250k USD in funding. The basic idea for this "libre GPU" is to use a RISC-V processor. The GPU will be mostly software-based. It will leverage the LLVM compiler infrastructure and utilize a software-based Vulkan renderer to emit code and run on the RISC-V processor. The Vulkan implementation will be used for writing in the Rust programming language. The project's current road-map has details only on the software side of figuring out the RISC-V LLVM back-end state. Work is being done on writing a user-space graphics driver, implementing the necessary bits for the proposed RISC-V extensions like "Simple-V". While doing this, they will start figuring out the hardware design and the rest of the project. The road-map is quite simplified for the arduous task at hand. The website notes: “Once you've been through the "Extension Proposal Process" with Simple-V, it need never be done again, not for one single parallel / vector / SIMD instruction, ever again.” This process will include creating a fixed-function 3D "FP to ARGB" custom instruction, a custom extension with special 3D pipelines. With Simple-V, there is no need to worry about about how those operations would be parallelised. This is not a new concept, it's borrowed directly from videocore-iv. videocore-iv calls it "virtual parallelism". It's an enormous effort on both the software and hardware ends to come up with a RISC-V, Rust, LLVM, and Vulkan open-source combined project. It is difficult even with the funding considering it is a software based GPU. It is worth noting that the EOMA68 project was started by Luke in 2016 and raised over $227k USD from crowdfunding participants and hasn't shipped yet. To know more about this project, visit the libre risc-v website. NVIDIA leads the AI hardware race. But which of its GPUs should you use for deep learning? AMD ROCm GPUs now support TensorFlow v1.8, a major milestone for AMD’s deep learning plans PyTorch-based HyperLearn Statsmodels aims to implement a faster and leaner GPU Sklearn