Hear from CIOs, CTOs, and other C-level and senior execs on data and AI strategies at the Future of Work Summit this January 12, 2022. Learn more

Let theOSS Enterprise newsletterguide your open source journey!Sign up here.

Open source software (OSS) is never too far away from both acclaim and controversy, whether its a major security incident, a trademark tussle, or flying a helicopter on Mars.

Lets take a look back at some big OSS talking points of the year.

Security is always a major discussion point in the open source sphere, and 2021 was no different. The biggest story of the year was almost certainly thezero-day vulnerability found in the Apache logging library Log4j, which is used by countless companies across the consumer and enterprise realm from Apples iCloud to AWS and IBM.

Log4Shell, as the vulnerability is called, had existed since 2013, but was only discovered by Alibabas security staff in late November and publicly revealed two weeks later. It is considered particularly dangerous, given that it enables remote code execution (RCE), allowing hackers to gain access to remote systems and sensitive data. Log4Shell was elevated to near-celebrity status when it was awarded a CVSS (Common Vulnerability Scoring System) security rating of 10.

Although the Apache team issued a fix on December 6, Log4js ubiquity across cloud services, infrastructure, and everywhere in between, makes it difficult for every company to update their systems quickly enough on top of that, they might not even know that their software relies on Log4j in the first place. Needless to say, attackers began seeking to exploit Log4Shell in the wild and widened their scope to the ransomware realm.

There are many lessons to learn from this, as Reblazes open source program manager Justin Dorfman wrote in VentureBeat:

The incident exposes how a vulnerability in a seemingly simple bit of infrastructure code can threaten the security of banks, tech companies, governments, and pretty much any other kind of organization.

However, in the wake of the Log4j vulnerability, the usual argument reared its head, with countless people noting that it shone a light on the inherent security deficiencies of community-driven software. But others countered that by stating the main problem was that companies were happy to benefit from open source in the good times, not give anything back, and then point their finger at OSS when things go wrong.

Serving as a somewhat sobering reminder, one of the Log4j projects core maintainers Ralph Goers, who fixed the vulnerability has a full-time job elsewhere as a software architect. Goers works on Log4j and other open source projects in his spare time.

Above: LAS VEGAS, NEVADA NOVEMBER 30: Attendees arrive during AWS re:Invent 2021,

Image Credit: Noah Berger / Stringer via Getty

Arguably, one of the biggest talking points came at the turn of the new year, when Elastic revealed it was transitioning its database search engine Elasticsearch from an open source Apache 2.0 license to a duo of proprietary source available licenses. The move, ultimately, came as little surprise and was the culmination of years of headbutting between Elastic and Amazons cloud computing offshoot, Amazon Web Services (AWS).

As a fully open source project, any company had been free to do whatever they wanted with Elasticsearch including offering it as-a-service, as Amazon did when it launched the Amazon Elasticsearch Service way back in 2015. This kicked off a chain reaction of events that ultimately led Elastic to shift Elasticsearch and theKibana visualization dashboard to new licenses.

That Amazon had elected to use Elasticsearch in the name of its own managed service was one of the problems it was, in Elastics eyes, a clear trademark infringement, and it caused confusion in the market space about which Elasticsearch service was which. This is why Elastic filed a lawsuit against Amazon back in 2019, but lawsuits are not generally a swift process. Additionally, changing the license helped speed things up in terms of swaying Amazon away from the Elasticsearch brand. It worked, as just one week after Elastic announced the license switch, Amazon revealed it would begin work on an open source Elasticsearch fork, which would eventually ship under a completely new name OpenSearch.

Licensing kerfuffles were evident elsewhere in the open source sphere too.The Software Freedom Conservancy (SFC), whose sponsors include Google and Red Hat, sued Vizio, alleging that the smart TV maker breached two open source licenses by using and modifying software without making the derivative source code publicly available. Vizio is showing no signs of budging, though, and the case took a somewhat ugly turn when Vizio filed a request to remove the case from California State Court, seemingly based on the belief that consumers have no third-party beneficiary rights under copyleft.

Meanwhile, former U.S. president Donald Trumps upcoming social network Truth Social apparently violated Mastodons open source license, with Mastodon originally threatening a lawsuit. The crux of the problem was that Truth Socials terms-of-service claimed the code was entirely proprietary, and made no reference whatsoever to its Mastodon foundation moreover, the open source license stipulates that all derivative projects must also be made available under the same license.

While the social network is yet to formally launch, it appears it has gone some way toward meeting Mastodons licensing requirements it recently acknowledged that it was built upon Mastodon, and the developers uploaded a zip file of its source code. Whether that will be enough remains to be seen, but the eyes of the open source community will remain on Trumps company ahead of the official launch in 2022.

The issue of trademarks is by no means unique to AWS vs. Elastic. Just before the new year kicked off, Facebook asserted trademark ownership over the open source PrestoDB project. This caused a problem for PrestoSQL, a fork created by the original Presto creators when they left Facebook they were forced to change their projects name to Trino.

Fast-forward ten months to November, and live streaming software provider Streamlabs OBS had to drop OBS from its name after it was called out by the open source OBS project on which it is built. Similar to AWS vs. Elastic, avoiding brand confusion was central to this, with the OBS projects Twitter account revealing that some of its support volunteers had to deal with angry Streamlabs customers, who were apparently confused between the two entities.

Open source software is so pervasive, it has often been said that it is eating the world. But if the first-ever Martian helicopter flight is anything to go by, open source software is eating the entire solar system.

The historic achievement was made possible by an invisible team of open source developers from around the world, GitHubs former CEO Nat Friedman wrote. Some 12,000 developers contributed to open source projects used in the software that powered the helicopters maiden flight on the Red Planet and yet, most of these developers are not even aware that they helped make the first Martian helicopter flight possible, noted Friedman.

To mark the occasion, GitHub placed a Mars 2020 Helicopter Mission badge on the GitHub profile of every developer who had contributed to code that was used in the mission.

Linux was first released on September 17, 1991, the omnipresent open source operating system turned the grand old age of 30 this year.

Its impossible to overstate the importance of Linux across the technological spectrum. Android the worlds most widely used mobile operating system is based on a modified version of the Linux kernel. Today, Linux is used in everything from automobiles and air traffic control to medical devices, and is also widely employed in web servers, the most common being Apache. In fact, the growth of the web over the past 30 years has been fueled in large part by Linux and similar open source software.

Heres to the next 30 years of open source innovations.

Originally posted here:

2021: A year in open source - VentureBeat

To manage the complexities of todays cloud native infrastructure, theres an increasing need for IT monitoring technology, often referred to as observability. As core building blocks of this trend, two sophisticated technologies are rapidly developing: Monitoring-as-Code and Infrastructure-as-Code.

Lets dig into these emerging technologies and look at how they help monitor and manage complex cloud native systems.

Also see: Why Cloud Means Cloud Native

In the Infrastructure-as-Code model, there is still physical infrastructure, but it comes in the form of a descriptive model for software-defined compute, storage and networking resources that have been provisioned and subsequently delivered by cloud providers.

In IaC, there are no wires, plugs, hubs and routers; instead, there are connection topologies defined by source code files. So of course there are cables and wires, but they live in the cloud data center and code defines which ones we use, on what day, for what workload and at what level of impedance and so on.

As we now traverse into the cloud-native environment driven by Infrastructure-as-Code with all the containerized Kubernetes granularity that comes with it, we still need to keep our eyes on the central challenge and look out for unexpected turns and speed bumps. Which brings us to Monitoring-as-Code.

The always-on dashcam that is Monitoring-as-Code (MaC) plays a key role in managing complex systems. A close cousin to (but not wholly synonymous with) Observability-as-Code (OaC), MaC is now rapidly evolving from a simple data collection function to form part of the higher-tier observability strata that typify modern multi-cloud deployments.

In a world where DevOps, DevSecOps, and AIOps are all helping to create the comparatively new roles in Site Reliability Engineering (SRE), a new and more evolved use of monitoring will surely be needed as we move from static IT infrastructures to the dynamic models of IaC.

At its most basic level, Monitoring-as-Code (MaC) is an exercise in data collection via software agents, plugins and discrete elements of source code that work as exporters to track system-level data and direct it to a repository. This is MaC level 1.0, if you will.

To get to MaC 2.0 and beyond, we need to think about building a wider observability lifecycle and enabling it with intelligence, diagnosis and extended abilities to achieve levels of automated remediation with what we might call codified reactive logic. In simple terms, the foundation blocks of IT and cloud software are shifting.

One of the major tech trends for 2022 is going to be the integration of observability and pipelines as adjacent markets, said James Governor, co-founder of technology analyst house Red Monk. Testing is shifting left, but its also shifting right, into production. Software delivery pipelines are now a mission critical application. Were seeing observability players work more closely with companies providing CI/CD and software delivery tooling to make life easier for customers.

Governor says that so adjacent are the two work disciplines (or markets) of observability and software delivery that they are converging to enable what we can call progressive delivery.

Also see: AIOps Trends

As we move to progressive delivery, we can start focusing on use cases such as blue/green deployments, canarying, feature flags and dark launches, with progressive rollouts to specific user cohorts to limit the blast radius. Production then becomes part of the software delivery lifecycle, explained Governor.

He says that the as code revolution is driving all of this forward, enabling more sophisticated pipelines and extensions as organizations seek to modernize their app delivery, with better testing, more automation and observability baked in.

Current players in the Monitoring-as-Code space include Sumo Logic with its open source Sensu project. Described as an observability pipeline built to deliver Monitoring-as-Code on any cloud, Sensu promises to consolidate monitoring tools and fill gaps in observability across data silos with automated diagnosis and self-healing. In sum, it supports the full transept of modern IT from bare metal to Kubernetes.

Sumo Logic acquired the Sensu project back in June of 2021 and CEO Caleb Hailey and co-founder Sean Porter joined the new parent company in roles designed to further their work. The pair conceived Sensu while working on an internal project to replace an out-of-date Nagios infrastructure monitoring installation for a cloud information archiving company.

Without necessarily deriding their worth, Porter himself called out MaC 1.0 tools for what they were: software agents designed to transit system file information to configuration management tools that might include Puppet, Chef, Helm, Terraform, Ansible or other. He described this as an action akin to retrofitting traditional monitoring tools to the modern DevOps paradigm.

This all brings us full circle into another modern application platform implementation paradigm. The world of CI/CD can take MaC deep into its structure if we make sure we adopt the evolved MaC 2.0 at its core.

MaC 1.0 in CI/CD means using monitoring tools to test application and related services operational health and wellbeing. MaC 2.0 in CI/CD means using monitoring tools with codified reactive logic to build and test applications. And to automate their operational wellbeing in live production environments.

As the use of MaC 2.0 grows in contemporary CI/CD environments that make significant use of containerization through technologies like Kubernetes, along with artificial intelligence, the common packaging of containers can enable us to extend CI/CD into an even more formalized pipeline.

Sumo Logics Porter refers to these developments as the move to new new Infrastructure-as-Code, a term that clearly needs reinvention (neo-IaC, perhaps). Additionally, this also clearly applies to how we build CI/CD pipelines with more operational functions including monitoring and observability baked in.

Its important to create a more operationally intelligent, reactive and functional Continuous Integration/Continuous Delivery (CI/CD) pipeline that doesnt just do and view. It also sees, understands and reacts so this new IaC substrate can help build a system to serve highly demanding cloud services needed by so-called high-velocity, cloud native enterprises.

Monitoring-as-Code (MaC) solves a bane for many projects whereby unexpected issues during pre-production testing or deployments go undetected. We lose hours allowing failing tests to continue, then more time troubleshooting the problem, and of course, we miss the opportunity to investigate root cause at the point of failure, explained Seng Phung-Lu, AVP of site reliability engineering at TD Bank.

Phung-Lu says that with MaC tooling deployed alongside an application via a single unified pipeline, he and his team are able to catch any issues early and avoid (as he puts it himself) having to manually babysit the testing and CI/CD process.

The real value of these emerging technologies may be a new definition of what we mean by end-to-end monitoring.

Back in the day, end-to-end monitoring used to mean observability applied to everything from the device Operating System (OS) and application User Interface (UI) to the compute, storage, analytics functions and back again to the user. The new end-to-end monitoring is all of that, plus an ability to deliver monitoring across the entire breadth of infrastructure that all of the above runs on.

In the virtuous circle of MaC 2.0 inside an end-to-end neo-IaC environment, a new symbiotic relationship starts to develop; application metrics are collected and detected automatically in response to code changes and new deployments. Onward extension of the IT stack benefits from the learned mistakes (code execution failures) that have already been experienced and the stack itself draws upon its own evolutionary DNA.

About the Author:

Adrian Bridgewater is a leading journalist covering emerging technologies.

More here:

Why Cloud Native Needs Monitoring-as-Code and Infrastructure-as-Code | eWEEK - eWeek

One feature that users found useful in Windows was the ability to drag and drop files onto the taskbar, to open them in the pinned program. For instance, you could drag a text file from Explorer on to the Notepad icon on the taskbar to open it.

Microsoft removed this feature from Windows 11, along with quite a few options, not to mention the entire right-click menu's usability. Martin has a nice write-up that talks more about the taskbar's state in the latest operating system.

If you were using the Drag and Drop feature in prior versions of Windows, and want it back, in Windows 11, there is something that you can do. Like some other fixes that we have seen, this too requires a third-party solution.

The utility you will need for restoring the feature, is called Windows 11 Drag & Drop to the Taskbar (Fix). The name is kind of on the nose, but the program works quite well. Download the executable from GitHub. It is a portable software which creates a config file in the same folder. Run the program and an icon will appear on the system tray. This indicates that the application is active, it needs to be running in for the functionality that it restores.

To test if it works as intended, open any application that you'd like to open files in. E.g. Notepad. Minimize the window, and switch to Explorer. Drag a text document over Notepad's icon and mouse over the preview pane that appears, and click it. The file will be opened in the text editor.

Note: You need to have the Microsoft Visual C++ 2015-2022 Redistributable x64 installed on your computer, else the tool won't work.

When you drag something to the taskbar, the tool waits for a few milliseconds to see if the action is valid, and simulates the Win + T hotkey. This shortcut is normally used to switch between a program's windows in the taskbar, but the application uses it to open files.

The only downside to the program is that it cannot create be used to pin shortcuts to the taskbar, and according to the project's site, this function is not being worked on.

Windows 11 Drag & Drop to the Taskbar (Fix) is an open-source application, you can find the source code on the project's GitHub repository. Right-click on its taskbar icon and select the Configure option, to modify its settings, you'll see the config file open in a Notepad window. The program's documentation lists several options that you can add per your requirements, so give it a read. The tool works out of the box, but it never hurts to have more options. Users who are using a custom theme may want to look into the setting that checks the colors of the pixels, to prevent the tool from accidentally detecting a mouse action as a drag.

As for an official fix, Windows Latest reports that Drag and Drop for the Taskbar will be added in Windows 11 22H2, which could arrive in a few months.

What feature do you miss in Windows 11?

Summary

Article Name

How to restore drag and drop to the Taskbar in Windows 11

Description

Restore drag and drop to the Taskbar in Windows 11 with this user-friendly tool.

Author

Ashwin

Publisher

Ghacks Technology News

Logo

Go here to see the original:

Restore drag and drop to the Taskbar in Windows 11 with this open-source program - Ghacks Technology News

DUBAI, UAE, Dec. 29, 2021 /PRNewswire/ -- Joget, Inc., the open source no-code/low-code application platform company for faster and simpler digital transformation and its partner, Mokxa Technologies, concluded an insightful CIO summit at Atlantis, Dubai on December 16, 2021.

The MOJO (Mokxa and Joget) CIO co-hosted event gathered attendees from a wide range of industries including computer software, automotive, hospital and healthcare, financial services, government administration, education, construction, logistics and many more. Speakers at the event shared stories on their digital transformation journeys leveraging low-code and cloud native technologies to enable radically efficient enterprise evolution.

Project Management Institute, a US-based world leading professional association for project management with more than 600,000 global members, explained the current Citizen Development evolution and how it benefits IT by freeing up capacity, reducing maintenance costs and improving the transparency of application inventory.

Other customers shared their low-code success by streamlining the development and implementation of their internal and external applications on the Joget platform. These transformations resulted in more than $2 million in savings.

"It was one of the first in-person events we did in the MEA region and it was amazing to see and feel the energy in the room. We are witnessing a revolution in the enterprise software realm - pushing enterprises to move faster and create meaningful business transformations with limited resources. We are excited that we are leading this revolution around the globe," said Parminder Sethi, President and CEO at Mokxa Technologies.

"It was very energizing to hear from customers how they have leveraged Joget in innovative ways and the value they are creating with Joget for their organizations," said Raveesh Dewan, President and CEO at Joget, Inc. "At Joget we are always focused on how we can enable our customers to go faster with a simple execution and implementation approach."

To learn more about the Joget DX low-code platform, visit https://www.joget.com.

About Joget, Inc. Joget, Inc is the developer of the Joget open source no-code/low-code application platform. Joget empowers business users, non-coders or coders to create enterprise applications across industries and countries. With more than 200,000 downloads, 3,000 installations and 12,000 community users worldwide across various industries (including finance, manufacturing, IT, and more), Joget is a proven platform for a wide spectrum of organizations ranging from Fortune 500 companies to government agencies and small businesses.

About Mokxa Technologies, LLC Mokxais a global IT Products, Solutions and Consulting provider with expertise in process and technology transformations including concept, design, development, testing and data architecture. Mokxa offers multiple technological products and solutions in various industries such as Healthcare, M&A, High Tech, Insurance and Financial Services. Mokxa also specializes in products and services around Joget, a low-code/no-code and workflow automation platform with scores of implementations and support services worldwide. With a combination of hybrid thinking, domain experience and technology capabilities powered by innovation and passion, Mokxa strives to bring the best value to its customers with the agility and nimbleness of a startup but grounded in the realities of the customer situation.

ContactsMedia Relations[emailprotected] 1.888.60J.OGET (1.888.605.6438)

SOURCE Joget

Go here to see the original:

CIO Summit Concludes with Digital Transformation Insights from Joget Customers and Industry Thought Leaders - PRNewswire

A year after the SolarWinds compromise, federal agencies are once again entering the new year in the middle of a cybersecurity emergency, with IT offices racing to identify and patch instances of widely used software code on their networks.

But after 12 months of blocking and tackling, experts say the federal enterprise is better positioned to handle cybersecurity incidents and primed for more progress in 2022.

Agencies had until Dec. 23 to identify and patch instances of Log4j on their Internet-facing systems, according to an emergency directive issued by the Cybersecurity and Infrastructure Security Agency last week. The vulnerability in the widely used Apache open source logging software emerged earlier this month.

The directive also tells agencies to report all affected software applications to CISA by Dec. 28.

CISA is very pleased with the urgency with which agencies are addressing Log4j vulnerabilities, a CISA official told Federal News Network on-background. The official said CISA has hosted multiple calls attended by thousands of staff across civilian agencies, including chief information officers and chief information security officers, as well as IT operations and security operations center personnel.

Federal IT and Cybersecurity leaderships commitment to urgently addressing these vulnerabilities as a cohesive enterprise has been clear since the onset, the official said. We have no confirmed compromises across federal civilian networks relating to the Log4j vulnerability.

Efforts to address Log4j will bridge 2022 to a year that saw the creation of the National Cyber Directors office, the establishment of CISAs Joint Cyber Defense Collaborative, and efforts to implement President Joe Bidens cybersecurity executive order, among other federal developments.

Mark Montgomery, senior advisor to the recently shuttered Cyberspace Solarium Commission, said 2020 was a watershed year in cybersecurity, when lawmakers included 26 of the commissions recommendations in the annual defense bill, including the creation of the NCD.

2021 was much, much more of blocking and tackling, Montgomery said. More changes to CISA authorities, the establishment of a Response and Recovery Fund, building better public-private partnership programs in the Department of Defense and Department of Homeland Security. Those are smaller issues, but important issues that that help in in getting the cybersecurity agenda moving forward.

Chris Cummiskey, a consultant and former Department of Homeland Security under secretary for management, said it was a rebuilding year for federal cybersecurity after the Trump administration did away with the former White House national cybersecurity coordinator position.

Theres much better coordination now between the White House key cyber functions, the National Security Agency, the FBI, and CISA, Cummiskey said. When you have that coming together on all cylinders, its a much more effective response.

Bidens May cybersecurity executive order, spurred on by the SolarWinds breach, also set tight deadlines for agencies to begin improving their cyber practices, including through the shift to zero trust architectures. Chris DeRusha, federal chief information security officer, recently said that agencies are now entering the execution phase of the EO.

Cummiskey said past cyber executive actions have lacked sufficient accountability and performance tracking measures. But he believes officials like National Cyber Director Chris Inglis, CISA Director Jen Easterly and DeRusha will hold agencies accountable to the latest EO.

The difference this time is that youve got a lot of seasoned veterans in leadership roles that are helping other agencies kind of understand that if were going to move to zero trust architecture, or if were going to strengthen the enterprise, cybersecurity functions at agencies, its got to be more than just a reporting requirement through FITARA, Cummiskey said, referring to the Federal Information Technology Acquisition Reform Act that requires annual reports from agencies on IT inventories and plans.

The executive order put CISA in charge of multiple action items to improve federal cybersecurity, including the designation of critical software and the establishment of a Cyber Safety Review board.

CISA also saw continued support from Congress this year, especially funding. The American Rescue Plan added $650 million in emergency funds on top of CISAs $2 billion annual budget. Lawmakers are proposing increasing CISAs budget to $2.4 billion in fiscal year 2022.

However, in spite of a scourge of ransomware attacks, including incidents that shut down Colonial Pipeline and a major meatpacking plant, Congress could not come to an agreement on including cyber incident reporting requirements in legislation by the end of the year.

Tatyana Bolton, former cyber policy lead at CISA, said the continued policy of voluntary reporting for critical incidents leaves a major hole in the U.S. approach to cybersecurity. Bolton was also on the staff of the Solarium Commission and directs cyber policy for the R Street Institute.

The fact that we couldnt have that very simple provision into law is very unfortunate, and I think were going to see over the course of the next year how not having that tool in the toolbox for the federal government is going to be a weakness of our cyber strategy, Bolton said.

But she and Montgomery expect lawmakers will make another strong to pass reporting requirements in 2022.

Bolton also predicts Inglis and the National Cyber Directors office will make progress next year on efforts to introduce more resilience into the U.S. cybersecurity approach, taking a wider view of incidents like ransomware attacks and Log4j.

His efforts on resilience is focusing on the broader picture, she said. Its the forest for the trees.

Meanwhile, CISAs emergency directive shows the agency will continue to monitor the Log4j situation into the new year. CISA is planning to provide a report by Feb. 15 to both DHS Secretary Alejandro Mayorkas and the White House on identifying cross-agency status and outstanding issues, according to the directive.

View original post here:

After turbulent cyber year, agencies enter 2022 with fresh security crisis on hand - Federal News Network

A lot of people still believe that CryptoPunks is nothing more than a general Internet fad, but the concept behind this project is actually much more important.

On the face of it, CryptoPunks is simply a series of 8-bit, digital characters that have no real use. People buy them and use them as profile pictures on platforms like Twitter, messaging apps and more. They can sell these on other platforms too, and their price keeps appreciating.

However, each CryptoPunk avatar is linked to the blockchain and it can be linked to web3 apps. Which is exactly how the realistic avatars Mark Zuckerbergs Meta is imagining today will work in future.

And if you take all that into account, the fact that CryptoPunk 4156 was auctioned for $10.35 million doesnt sound too far-fetched. The project could, in theory, evolve to become something much bigger in future.

Read the original post:

Pak, Beeple, CryptoPunks, the Internets source code these are the 15 most expensive NFTs sold in 2021 - Business Insider India

In recent years, AI tools and frameworks have made the use of AI much friendlier for the IT sector

Artificial intelligence has been around for a very long time but it was considered to be extremely difficult. It was quite common for researchers and developers to shy away from studying or using it. However, in recent years, AI tools and frameworks have made it much friendlier for the IT sector. Artificial intelligence (AI) technologies are quickly transforming almost every sphere of our lives. From how we communicate to the means we use for transportation; we seem to be getting increasingly addicted to them. Here is the list of the top 10 open-source AI tools and frameworks to use in 2022.

Theano, an open-source Python library for deep learning, is also popular in the neural processing and data science communities. Its widely known for making it easy to implement complex neural networks by abstracting away the neural network components (such as the layers and hidden layers). Its often used to build and train AI models on graphics processing units (GPUs) and has been adopted by Facebook for both training and deploying AI applications.

It is an open-source library developed by Google Brain that is used to deal with complex datasets and execute high-volume numerical computations. Used by tech giants like Nvidia, Google, Intel and SAP, it allows developers to send counterfeit neural systems with massive datasets after setting up, and training them. Tensorflow certificate program follows self-paced learning in ML for foundational level along with the support of experienced Tensorflow programmers for this open-source software library.

PyTorch is an AI system created by Facebook. Its code is accessible on GitHub and at present has more than 22k stars. It has been picking up a great deal of energy since 2017 and is in a relentless reception development.

Caffe is a profound learning structure that prioritizes articulation, tempo, and quantifiable consistency. The Berkeley Vision and Learning Center (BVLC) and network contributors fund it. The Caffe Framework is needed for Googles DeepDream. This is a BSD-compliant C++ library with a Python interface.

Keras is a high-level AI API that can run on top of TensorFlow, Microsoft Cognitive Toolkit, and Theano. Its ease of use and focus on the developer experience make Keras the go-to for rapidly prototyping new apps. Many brands like Netflix, Uber, and Yelp, as well as smaller startups, have integrated Keras into their core products and services. Netflix, for example, has leveraged deep learning to predict customer churn, which is crucial as a subscription-based business.

The Microsoft Cognitive Toolkit (CNTK) is an open-source AI framework. CNTK can be included in projects as a library in various languages or used as a standalone machine-learning tool through its model description language called BrainScript. The commercial-grade toolkit is used by Skype, Bing, Cortana, and other brands with massive datasets that require a scalable and highly optimized machine learning platform.

It takes a different approach than the high-level application programming interfaces of libraries like fast.ai for PyTorch or Keras in TensorFlow. Instead of staying true to its mathematical and scientific computing roots to support coding patterns that are very similar to the equations you might read in a scientific paper describing a new ML technique. Flux.jl is used in many other Julia language machine learning projects, including DiffEqFlux.jl that we discussed earlier. For the interested beginner looking for an experience most similar to Autograd or JAX from Python, Zygote.jl, a Flux-based library for advanced automatic differentiation, is probably the best place to start.

It enables the trade-off of computation time for memory through a feature called forgetful backdrop, which is particularly useful for recurrent nets with exceptionally long sequences. Scalability was a primary consideration when developing this application (easy-to-use support for multi-GPU and multi-machine training). Numerous exciting things, such as the ability to effortlessly write custom layers in high-level languages.

RNN is an emerging framework for supervised learning and has an extremely flexible and intuitive interface. Its also suitable for designing algorithms for deep learning, which can be used to distinguish between like and dislike in data sets. RNN is the most popular deep learning framework for neural processing and natural language processing. The user community has been extraordinarily active and very helpful and the project is under active development.

Amazon recently open-sourced Amazon SageMaker Neo, a feature of its machine learning platform, as a service offering. The newly released Neo-AI project code will enable AI developers to train machine learning models and run them anywhere in the cloud. The Neo-AI project is optimized for edge computing devices and Internet of Things (IoT) sensors that need to make fast and low-latency predictions.

Share This ArticleDo the sharing thingy

Original post:

Top 10 Open-Source AI Tools and Frameworks to Use in 2022 - Analytics Insight

When you hear raytracing, you might think of complex dark algorithms that to stare too long at their source code invites the beginning of madness. And youre technically not far off from the truth, but [h3r2tic] put a small open-source ray tracing game demo up on GitHub. The actual rust code powering the game is relatively short (just four files), with the longest file being the physics file. But, of course, there is a small mountain of code under this sample in the form of libraries.

Kajiya, physx-rs, and dolly are the three libraries that make this little demo possible. Kajiya, in particular, is what makes raytracing possible as it uses the newer RTX features (so only more recent Nvidia and AMD cards are supported) and Vulkan bindings. But, of course, it isnt wholly ray-traced as we are still several years out from proper real-time raytracing. Nevertheless, the blend between raytracing and traditional rasterization looks incredible. The most important thing about this simple tiny sample isnt the game itself but what it stands for. It shows how easy it is to create a sample like this. Even just five years, creating a demo like this required massive effort and expertise.

Visually, it is stunning to look at. While the reflections are most apparent, the takeaway from this is the ease that real-time global illumination brings. A quick look through the code shows that there are very few lights in the scene, despite looking well lit with soft shadows. Traditional video games spend a significant amount of development time lighting a scene, placing additional lights, and tweaking them to make up for all the shortcuts that lighting has to take in a rasterized environment. As more and more games are built with raytracing in mind rather than tacked on at the end, we can ditch the small crumbling mountain of hacks that we are forced to use in games today and just rely on the rays to accurately light a scene.

If using a library for your raytracing seems too easy, perhaps youd like to take on the challenge of raytracing in excel. Video after the break.

Original post:

Whipping Together A Little Ray Tracer Racer - Hackaday