Blog

  • Efficient Long-Term Trend Analysis in Presto Using Datelists

    Efficient Long-Term Trend Analysis in Presto Using Datelists

    Data analytics teams often find themselves delving into long-term trend analysis to uncover valuable insights and track patterns over time. Whether it’s examining Week over Week (WoW), Month over Month (MoM), or Year over Year (YoY) trends, this type of analysis typically involves storing data across multiple years. However, this approach can quickly consume a significant amount of storage space and make querying across years of partitions both inefficient and costly. To compound the challenge, performing user attribute cuts on such a vast dataset can further complicate the process.

    In response to these issues, a more streamlined and cost-effective solution can be implemented using datelists. By leveraging datelists in Presto, data analytics teams can significantly enhance the efficiency and effectiveness of their long-term trend analysis processes. Datelists offer a structured approach to managing dates and can greatly simplify the handling of time-based data.

    One key advantage of using datelists is the reduction in storage requirements. Rather than storing data across multiple partitions spanning several years, datelists enable teams to maintain a centralized list of dates. This centralized approach not only optimizes storage utilization but also streamlines the querying process by eliminating the need to scan through numerous partitions.

    Moreover, datelists facilitate seamless user attribute cuts, enabling teams to perform targeted analyses with ease. By associating user attributes with specific dates in the datelist, analysts can efficiently filter and segment data based on their requirements. This level of granularity empowers teams to gain deeper insights into trends and patterns, facilitating more informed decision-making.

    In practical terms, implementing datelists in Presto involves creating a dedicated table that stores a comprehensive list of dates. This table can then be linked to the primary dataset, allowing for efficient date-based queries and analyses. By indexing the datelist table appropriately, teams can further enhance query performance and accelerate data retrieval processes.

    Furthermore, the versatility of datelists extends beyond traditional trend analysis scenarios. Teams can leverage datelists for various time-based analyses, such as seasonality studies, holiday impact assessments, and trend forecasting. This flexibility underscores the value of datelists as a fundamental tool in the data analytics toolkit.

    In conclusion, incorporating datelists into Presto workflows offers a practical and efficient solution for conducting long-term trend analysis. By centralizing date management, optimizing storage utilization, and facilitating targeted user attribute cuts, datelists empower data analytics teams to unlock valuable insights from historical data with ease. Embracing datelists as a core component of data analysis processes can drive efficiency, cost savings, and actionable intelligence, making it a valuable asset for modern analytics environments.

  • Attackers Hijack Google Advertiser Accounts to Spread Malware

    In a daring move that strikes at the core of Google’s operations, attackers have been hijacking Google advertiser accounts to spread malware through malicious advertisements. This sophisticated form of malvertising poses a significant threat to users and businesses alike, highlighting the need for heightened cybersecurity measures in the digital advertising realm.

    The concept of malvertising involves using online advertising to distribute malicious software. By infiltrating legitimate ad networks and platforms, cybercriminals can reach a wide audience with their nefarious payloads. In the case of Google advertiser accounts being compromised, the attackers gain access to a vast network of potential victims, leveraging the trust associated with Google’s brand to deceive users into clicking on dangerous ads.

    The implications of this form of attack are far-reaching. Not only does it undermine the integrity of online advertising, but it also puts users at risk of falling victim to various cyber threats, including ransomware, spyware, and phishing scams. For businesses whose ads are being used as vehicles for malware distribution, the damage to their reputation and the potential legal ramifications can be severe.

    Google, being at the forefront of digital advertising, recognizes the gravity of the situation. The tech giant has acknowledged the issue and is swiftly taking action to mitigate the impact of these malicious activities. By addressing the vulnerabilities in its advertiser accounts and enhancing its security protocols, Google aims to safeguard its users and maintain the trust that is essential for a thriving online ecosystem.

    As IT and development professionals, it is crucial to stay vigilant against evolving cybersecurity threats such as malvertising. Implementing robust security measures, conducting regular audits of online advertising practices, and educating users about the risks of clicking on suspicious ads are all vital steps in combating this pervasive threat.

    In conclusion, the hijacking of Google advertiser accounts to disseminate malware is a stark reminder of the ongoing battle against cybercrime in the digital age. By staying informed, proactive, and collaborative in our approach to cybersecurity, we can collectively defend against such insidious attacks and preserve the integrity of the online environment. Let us remain steadfast in our commitment to securing the digital landscape for all users.

  • Attackers Hijack Google Advertiser Accounts to Spread Malware

    In the ever-evolving landscape of cybersecurity threats, a recent development has sent shockwaves through the digital advertising realm. Attackers have been hijacking Google advertiser accounts to disseminate malware—a particularly audacious form of malvertising that strikes at the core of Google’s operations. This alarming trend has not gone unnoticed by cybersecurity experts, who highlight the severity of the issue and the urgent need for robust countermeasures.

    Malvertising, a portmanteau of “malicious advertising,” involves the use of online ads to spread malware. By infiltrating legitimate ad networks, cybercriminals can reach a vast audience of unsuspecting users. In the case of hijacked Google advertiser accounts, the impact is magnified due to the tech giant’s prominent position in the digital advertising ecosystem. This not only poses a significant threat to users but also undermines the trust and integrity of online advertising platforms.

    Researchers closely monitoring these malicious activities have sounded the alarm, emphasizing the brazen nature of the attacks. By leveraging Google’s own infrastructure, attackers can cloak their nefarious intentions and bypass traditional security measures. This sophisticated approach underscores the evolving tactics employed by cybercriminals to maximize their reach and impact.

    Google, for its part, has acknowledged the gravity of the situation and is taking swift action to mitigate the threat. The tech giant has reassured users and advertisers that it is actively working to address the issue and strengthen its defenses against such attacks. Given Google’s central role in the online advertising ecosystem, its proactive response is crucial in safeguarding users and preserving trust in digital advertising channels.

    The implications of attackers hijacking Google advertiser accounts to spread malware are far-reaching. Beyond the immediate risks to users’ devices and data, such incidents erode confidence in the online advertising industry as a whole. Advertisers, already grappling with ad fraud and brand safety concerns, now face the added challenge of ensuring the security of their advertising channels.

    In response to this emerging threat landscape, a multi-faceted approach is essential to bolster cybersecurity defenses. Advertisers must remain vigilant and implement stringent verification processes to prevent unauthorized access to their accounts. Likewise, users are advised to exercise caution when interacting with online ads and to maintain up-to-date security software to mitigate the risk of malware infections.

    As the cybersecurity landscape continues to evolve, collaboration and information sharing among industry stakeholders are paramount. By staying informed about emerging threats and adopting best practices in cybersecurity, advertisers, technology companies, and users can collectively combat malicious activities such as the hijacking of Google advertiser accounts for malware distribution.

    In conclusion, the recent incidents of attackers hijacking Google advertiser accounts to spread malware serve as a stark reminder of the persistent challenges in cybersecurity. As technology advances and threat actors become more sophisticated, proactive measures and a collective effort are crucial in safeguarding digital ecosystems. By remaining vigilant, informed, and adaptable, we can effectively defend against malicious activities and uphold the integrity of online platforms.

  • Terraform Gets AI Boost in New Cloud Management Platform

    In the fast-paced realm of cloud management, staying ahead of the curve is paramount. Recently, Terraform has taken a significant leap forward with an AI boost in its new cloud management platform. This development is poised to revolutionize how enterprises interact with the cloud, offering unprecedented levels of efficiency and automation.

    Aharon Twizer, a visionary in cloud technology for over a decade, has spearheaded this transformative initiative. By integrating artificial intelligence into Terraform’s capabilities, Twizer aims to streamline processes, optimize resource utilization, and enhance overall performance for enterprise users.

    Imagine a scenario where routine tasks, such as provisioning resources or managing configurations, are handled autonomously by Terraform’s AI-powered features. This not only frees up valuable human resources but also reduces the margin of error, ensuring seamless operations in complex cloud environments.

    One of the key advantages of this AI boost is its ability to adapt and learn from user interactions over time. By leveraging machine learning algorithms, Terraform can proactively suggest optimizations, predict potential issues, and customize solutions based on each enterprise’s unique requirements.

    For instance, Terraform’s AI capabilities could analyze historical usage patterns and recommend cost-effective strategies to optimize cloud spending. This proactive approach not only saves time and effort but also contributes to significant cost savings for organizations operating in the cloud.

    Moreover, the integration of AI in Terraform opens up new possibilities for predictive analytics and trend forecasting. By harnessing the power of data insights, enterprises can make informed decisions, anticipate future needs, and align their cloud strategies with business objectives more effectively.

    In practical terms, this AI-enhanced cloud management platform could revolutionize the way IT teams operate. Tasks that previously required manual intervention and oversight can now be automated, allowing teams to focus on strategic initiatives and innovation, rather than mundane operational tasks.

    As the digital landscape continues to evolve, technologies like Terraform’s AI-powered platform are essential for organizations looking to stay competitive and agile in the cloud. By embracing these innovations, enterprises can unlock new levels of efficiency, scalability, and performance in their cloud operations.

    In conclusion, Terraform’s AI boost in its new cloud management platform represents a significant milestone in the evolution of cloud technology. By combining the power of artificial intelligence with robust infrastructure management capabilities, Terraform is poised to empower enterprises to navigate the complexities of the cloud with confidence and efficiency. It’s an exciting time for cloud technology, and Terraform is leading the way towards a more intelligent and automated future.

  • Terraform Gets AI Boost in New Cloud Management Platform

    In the ever-evolving realm of cloud management, Terraform has taken a significant leap forward with an AI boost in its latest platform. Aharon Twizer, a key player in cloud innovation, has dedicated a decade to simplifying the cloud experience for enterprises.

    This advancement signifies a monumental shift in how organizations interact with cloud environments. By integrating artificial intelligence into Terraform, tasks such as provisioning and managing infrastructure become more efficient and intuitive. Imagine deploying resources with precision and speed, all while leveraging intelligent automation to optimize performance and cost-effectiveness.

    With this AI enhancement, Terraform empowers users to navigate complex cloud ecosystems with ease. From automating routine operations to predicting and addressing potential issues proactively, the platform offers a seamless experience that aligns with the dynamic nature of modern IT environments.

    Furthermore, the fusion of AI capabilities with Terraform opens doors to innovative possibilities. Machine learning algorithms can analyze usage patterns, anticipate scaling needs, and suggest optimizations, enabling organizations to stay ahead of the curve in a competitive digital landscape.

    The implications of this AI integration extend beyond operational efficiency. By harnessing the power of artificial intelligence, Terraform not only streamlines cloud management but also paves the way for strategic decision-making based on data-driven insights. This synergy between AI and cloud technology heralds a new era of agility and intelligence in IT infrastructure management.

    In essence, Terraform’s AI boost represents a paradigm shift in cloud management, propelling enterprises towards a future where automation, intelligence, and scalability converge seamlessly. As organizations strive to stay at the forefront of technological innovation, embracing AI-driven solutions like Terraform becomes not just a choice but a strategic imperative in unlocking the full potential of the cloud.

    To learn more about the transformative impact of AI in Terraform and how it can revolutionize your cloud management strategy, visit The New Stack for in-depth coverage of this groundbreaking development.

  • Kafka vs NATS: A Comparison for Message Processing

    In a distributed architecture, the way systems communicate is paramount to the entire infrastructure’s success. The efficiency, scalability, and dependability of the architecture hinge greatly on how events, messages, and data are transmitted and stored.

    Two prominent tools in the realm of streaming and messaging, Kafka and NATS, offer distinct architectures and performance attributes tailored to specific needs. Understanding the nuances of these platforms is crucial in choosing the right fit for your projects. Let’s delve into a comparison of NATS and Kafka, shedding light on their features and real-world applications.

    Kafka: The Robust Data Streaming Platform

    Kafka, known for its robust data streaming capabilities, is a distributed event streaming platform designed for high performance and fault tolerance. It excels in scenarios where durability and reliability are paramount, making it ideal for use cases requiring persistent storage of messages.

    One of Kafka’s key strengths lies in its ability to handle high message throughput while ensuring data integrity and fault tolerance. Its distributed nature allows for seamless scalability across multiple nodes, enabling horizontal growth as the volume of data increases. Kafka’s architecture, based on topics and partitions, facilitates efficient data organization and retrieval.

    NATS: The Lightweight Messaging System

    Contrasting Kafka’s comprehensive feature set, NATS shines as a lightweight and high-performance messaging system. NATS prioritizes simplicity and speed, making it an excellent choice for scenarios demanding low latency and high message delivery rates.

    NATS operates on a publish-subscribe model, where publishers distribute messages to subscribers efficiently. This streamlined approach minimizes overhead and ensures rapid message propagation within the system. NATS’ simplicity is a boon for applications requiring real-time communication and quick message processing.

    Choosing the Right Tool for the Job

    When deciding between Kafka and NATS for your messaging needs, consider the specific requirements of your project. If durability, fault tolerance, and scalability are paramount, Kafka proves to be a robust choice. Its ability to handle large volumes of data while ensuring data integrity makes it well-suited for applications like log aggregation, stream processing, and data integration.

    On the other hand, if your focus is on low latency, high throughput, and simplicity, NATS emerges as a compelling option. Applications that prioritize real-time communication, IoT data streaming, and microservices architectures can benefit significantly from NATS’ lightweight design and rapid message delivery capabilities.

    Real-World Applications

    At work, I’ve encountered scenarios where Kafka’s persistence and fault tolerance were instrumental in handling critical data streams. For instance, in a real-time analytics platform, Kafka’s durability ensured that no data was lost, even during system failures. Additionally, Kafka’s support for stream processing frameworks like Apache Flink enabled seamless data processing at scale.

    Conversely, in projects requiring rapid response times and minimal latency, NATS proved to be a game-changer. Its efficient message delivery mechanism allowed for instant updates in collaborative applications, where real-time data synchronization was paramount. NATS’ lightweight footprint and ease of use made it an ideal choice for microservices communication within our architecture.

    Conclusion

    In the realm of message processing, the choice between Kafka and NATS boils down to the specific needs of your project. While Kafka excels in durability and scalability, NATS stands out for its speed and simplicity. By assessing your project requirements, you can select the tool that aligns best with your goals, ensuring optimal performance and efficiency in your messaging infrastructure.

    In conclusion, whether you opt for Kafka’s robustness or NATS’ agility, both platforms offer unique strengths that cater to diverse messaging scenarios. Understanding the intricacies of each tool empowers you to make informed decisions, driving the success of your projects in the ever-evolving landscape of message processing.

  • Kafka vs NATS: A Comparison for Message Processing

    In a distributed architecture, communication between systems serves as the backbone of the entire infrastructure. The efficiency, scalability, and dependability of the infrastructure hinge greatly on how events, messages, and data are shared and stored.

    Two prominent contenders in the realm of streaming and messaging are Kafka and NATS. While both are stalwarts in the field, they diverge significantly in terms of architecture and performance attributes. Understanding these distinctions is key to determining the optimal choice for specific use cases.

    Kafka: The Powerhouse of Persistence

    Kafka, an Apache Software Foundation project, is renowned for its durability and fault tolerance. Its robust architecture enables the retention of vast amounts of data over extended periods. This makes Kafka an ideal candidate for scenarios requiring durable message storage and replay capabilities, such as log aggregation and analytics processing.

    One of Kafka’s standout features is its distributed commit log design. Messages are written to disk sequentially, ensuring high throughput and low latency. This design choice makes Kafka a solid option for use cases demanding high availability, such as real-time analytics and monitoring systems.

    NATS: The Lightweight Contender

    On the other end of the spectrum lies NATS, a lightweight and high-performance messaging system. NATS prides itself on its simplicity and speed, making it an excellent choice for scenarios where low latency and high message throughput are paramount, such as IoT applications and real-time data processing.

    NATS operates on a publish-subscribe model, where messages are quickly broadcast to multiple subscribers. This design lends itself well to scenarios requiring rapid message dissemination and real-time communication, such as chat applications and telemetry systems.

    Choosing the Right Tool for the Job

    When deciding between Kafka and NATS for message processing, it’s essential to consider the specific requirements of your use case. While Kafka excels in scenarios demanding durability and fault tolerance, NATS shines in environments prioritizing speed and simplicity.

    For instance, if you’re building a system that requires long-term data retention and the ability to replay messages, Kafka would be the preferred choice. Conversely, if your focus is on real-time communication and low latency, NATS would offer a more streamlined solution.

    Real-World Applications

    In my experience, I’ve leveraged Kafka for projects that necessitated persistent message storage, such as building a real-time analytics platform for processing large volumes of streaming data. The durability and scalability of Kafka proved invaluable in handling the continuous influx of data while ensuring fault tolerance.

    Conversely, when working on IoT applications where rapid message processing was critical, I turned to NATS for its lightweight nature and exceptional performance. NATS’ ability to swiftly deliver messages to multiple subscribers in real-time significantly enhanced the responsiveness of the IoT ecosystem.

    Conclusion

    In the realm of message processing, choosing between Kafka and NATS boils down to the specific requirements of your use case. While Kafka offers robust persistence and fault tolerance, NATS excels in speed and simplicity. By aligning your choice with the unique demands of your project, you can harness the full potential of these powerful messaging tools for optimal performance and scalability.

  • The Hidden Cost of Open Source Waste

    In the fast-paced world of software development, open source tools have become indispensable, fueling creativity, fostering collaboration, and accelerating project timelines. The allure of utilizing open source solutions lies in their cost-effectiveness and the vibrant community support they offer. Developers often leverage open source libraries, frameworks, and platforms to streamline their work, tapping into a vast pool of resources without reinventing the wheel.

    However, beneath the surface of this seemingly utopian landscape lies a hidden cost that can quietly accumulate over time: open source waste. This concept refers to the inefficiencies, redundancies, and unused resources that can result from the unchecked proliferation of open source components within a project or organization. While the initial adoption of open source software may seem like a cost-saving measure, the long-term implications of mismanagement can lead to significant resource drain and operational challenges.

    One of the primary contributors to open source waste is the lack of visibility and control over the various components integrated into a software project. With the ease of access to a plethora of open source offerings, developers may unwittingly introduce redundant or overlapping tools, leading to bloated codebases and decreased system performance. Additionally, the dynamic nature of open source ecosystems means that projects can quickly become outdated or unsupported, further exacerbating the problem.

    Consider a scenario where a development team incorporates multiple open source libraries to handle similar functionalities within a software application. Without proper oversight and governance mechanisms in place, these libraries may evolve independently, resulting in compatibility issues, security vulnerabilities, and a maintenance nightmare. In this case, what initially appeared as a cost-effective solution now translates into wasted time, effort, and resources spent untangling the web of dependencies.

    Furthermore, the licensing complexities associated with open source software can pose legal risks if not managed diligently. Failure to comply with licensing terms or to track the usage of third-party components can expose organizations to potential lawsuits, fines, or reputational damage. The absence of a robust open source management strategy leaves companies vulnerable to compliance issues, especially as regulatory requirements around software usage continue to evolve.

    To mitigate the hidden cost of open source waste, organizations must adopt a proactive approach to managing their dependencies effectively. Implementing tools that provide visibility into the open source components used, tracking their usage, monitoring for security vulnerabilities, and ensuring compliance with licensing requirements are essential steps in mitigating risks. Regular audits of the software stack, conducting code reviews, and promoting a culture of awareness around open source best practices can help prevent waste accumulation and maintain a lean, efficient development process.

    In conclusion, while open source software remains a cornerstone of modern software development, its unchecked proliferation can lead to unforeseen costs and complexities. By acknowledging the hidden cost of open source waste and taking proactive measures to address it, organizations can harness the benefits of open source while safeguarding against potential pitfalls. Embracing a strategic approach to open source management not only enhances operational efficiency but also fosters a culture of responsibility and sustainability within the development community.

  • The Hidden Cost of Open Source Waste

    Open Source Waste

    Open source software has been a game-changer in the tech world, fostering collaboration, accelerating development cycles, and driving innovation across industries. Its benefits are undeniable, offering cost-effective solutions and a vast pool of resources for developers worldwide. However, amidst the euphoria of open source adoption, there lies a hidden cost that often goes unnoticed: open source waste.

    While the allure of free software components is enticing, the management of open source usage within organizations can lead to significant inefficiencies and increased costs if not handled strategically. The proliferation of open source libraries, frameworks, and tools has made it challenging for teams to track, monitor, and maintain all the dependencies within their projects effectively. This lack of visibility can result in what is known as “dependency sprawl,” where redundant, outdated, or unused components accumulate over time, bloating the software stack and slowing down development processes.

    Consider a scenario where a development team integrates a popular open source library into their project to expedite coding tasks. Initially, this library serves its purpose efficiently, contributing to faster development and time-to-market. However, as the project evolves, other team members may introduce newer libraries or frameworks to address specific requirements, inadvertently neglecting to remove the original library. This accumulation of dependencies not only clutters the project but also poses security risks, as outdated components may contain vulnerabilities that hackers can exploit.

    Moreover, the lack of a centralized management system for open source components can lead to duplication of efforts, with different teams within an organization unknowingly using similar or identical libraries for separate projects. This redundancy not only consumes valuable resources but also hampers collaboration and knowledge sharing across teams. As a result, developers spend more time resolving compatibility issues, debugging code, and addressing conflicts between conflicting dependencies, diverting their focus from actual software development tasks.

    To mitigate the hidden costs of open source waste, organizations must implement robust strategies for managing their software supply chain effectively. This includes establishing clear policies for evaluating, approving, and monitoring the use of open source components, as well as implementing automated tools for dependency tracking and vulnerability scanning. By maintaining an inventory of all dependencies, version numbers, and licenses within their projects, teams can identify and eliminate redundant or obsolete components, optimizing their software stack for efficiency and security.

    Furthermore, fostering a culture of awareness and education around open source best practices can help mitigate the risks associated with dependency sprawl. Encouraging developers to stay informed about the latest updates, security patches, and deprecated libraries can empower them to make informed decisions when selecting and incorporating open source components into their projects. Additionally, promoting cross-team collaboration and knowledge sharing can prevent duplication of efforts and promote reusability of code across projects, reducing the overall impact of open source waste on organizational resources.

    In conclusion, while open source software offers invaluable benefits to developers and organizations, the hidden costs of mismanaged dependencies and inefficient usage can outweigh the initial advantages. By recognizing the implications of open source waste and implementing proactive measures to mitigate its risks, teams can harness the power of open source innovation while maintaining control over their software development processes. Embracing transparency, accountability, and continuous improvement in open source management is key to unlocking the full potential of collaborative software development in the digital age.

  • Heterogeneity of Computing Environments Using Cross-Compilation

    Title: Enhancing Application Portability: Exploring Cross-Compilation in Diverse Computing Environments

    In the realm of software development, the increasing prevalence of open-source solutions has revolutionized the way complex systems are built and deployed. As organizations strive to create applications that can seamlessly operate across a spectrum of hardware platforms, the concept of cross-compilation emerges as a vital tool in their arsenal.

    Cross-compilation is particularly essential in the domain of embedded systems, where the goal is to execute applications on specialized hardware such as ARM and PowerPC boards. These environments are often characterized by resource constraints, making direct compilation impractical. Here, developers harness the familiar x86 architecture as a host and employ target-specific toolchains to produce binaries optimized for the intended hardware.

    By adopting cross-compilation practices, developers can transcend the limitations imposed by diverse hardware architectures, ensuring the portability and efficiency of their applications across a multitude of platforms. This approach not only streamlines the development process but also enhances the overall user experience by enabling software to reach a broader audience without necessitating extensive rewrites or modifications.

    One exemplary tool that facilitates cross-compilation is Buildroot, a popular framework that simplifies the process of generating embedded Linux systems. By leveraging Buildroot’s capabilities, developers can create custom toolchains and configurations tailored to the unique requirements of their target hardware, thereby optimizing performance and compatibility.

    Moreover, the utilization of cross-compilation techniques promotes code reusability and modularity, enabling developers to focus on enhancing functionality rather than grappling with hardware-specific intricacies. This level of abstraction empowers teams to iterate swiftly, adapt to evolving technologies, and deliver robust applications that transcend traditional hardware boundaries.

    In essence, the heterogeneity of computing environments necessitates innovative approaches like cross-compilation to bridge the gap between diverse hardware architectures and software applications. By embracing this methodology, developers can future-proof their projects, enhance scalability, and unlock new opportunities for innovation in an ever-evolving technological landscape.

    To delve deeper into the intricacies of cross-compilation and its impact on software development, you can explore insightful resources like the DZone article on cross-compilation with Buildroot. This comprehensive guide sheds light on best practices, tools, and methodologies essential for mastering the art of cross-compilation and harnessing its transformative potential in your projects.

    In conclusion, the adoption of cross-compilation techniques represents a paradigm shift in the development paradigm, offering developers a powerful means to navigate the complexities of heterogeneous computing environments with agility and efficiency. By integrating cross-compilation into their workflow, organizations can elevate their software development practices, optimize resource utilization, and unlock new avenues for innovation in the digital landscape.