1. Understanding gRPC, OpenAPI and REST and when to use them in API design (2020)

Total comment counts : 53

Summary

The article discusses the primary models for API design, focusing on RPC (Remote Procedure Call) and REST (Representational State Transfer), and how these models interact with HTTP protocols:

  1. RPC vs. REST:

    • RPC: Focuses on procedures where data is encapsulated within calls. gRPC, a modern RPC framework, uses HTTP/2 but abstracts the HTTP layer from the developer.
    • REST: Emphasizes on resources where behaviors are derived from manipulating these resources. True REST APIs are rare, as many APIs labeled as REST do not adhere strictly to its principles.

  2. HTTP as a Transport:

    • HTTP is widely used due to familiarity with its security measures and port usage (80 and 443). The article outlines three main approaches for API design using HTTP:
      • REST: Clients use URLs as-is without constructing them, adhering to the original intent of HTTP.
      • gRPC: Uses HTTP/2 but hides it behind RPC abstractions, making it easier for developers to focus on the service without worrying about HTTP specifics.
      • OpenAPI: Clients construct URLs from templates and parameters, which contrasts with RESTful principles but is popular for its clarity and ease of documentation.

  3. Hybrid Approaches:

    • Modern APIs often blend elements from both RPC and REST. For example, using gRPC for implementation while adopting REST-like resource orientation, allowing for standard HTTP interactions.

  4. API Design Choices:

    • The article explains that choosing between these models involves considering factors like the need for statelessness (REST), simplicity and efficiency (gRPC), or detailed specification and documentation (OpenAPI).

  5. Practical Implications:

    • The choice of API style affects how clients interact with the API, how the API is documented, and how it’s implemented. Each model has its advantages, with REST offering simplicity, gRPC providing performance with HTTP/2, and OpenAPI offering detailed specifications.

In summary, the article guides API designers through the landscape of current API design philosophies, highlighting how traditional models are evolving with new technologies and hybrid approaches, aiding developers in making informed decisions based on their project’s needs.

Top 1 Comment Summary

The article expresses a strong regret over learning about gRPC, a high-performance, open-source universal RPC framework. The author details numerous frustrations encountered while using gRPC:

  1. Complexity and Debugging: Despite claims that gRPC simplifies things by hiding internals, the author found themselves needing extensive debug logging to diagnose issues, particularly with inconsistent request failures.

  2. Configuration Overload: The system requires tuning many settings related to timeouts and retries, which are poorly named and documented, leading to significant time loss.

  3. Tooling and Compatibility Issues: Problems with tools like Maven plugins, debugging timeouts, and load balancers (LBs) not handling HTTP/2 well.

  4. Network and Security: Firewalls caused additional headaches, forcing a fallback to less optimal solutions like the Standard API.

  5. Documentation and Observability: The documentation is described as poor, and there were difficulties in integrating meaningful error messages into observability systems.

Overall, the author wishes they had never encountered gRPC due to the extensive time and effort required to manage its complexities and limitations.

Top 2 Comment Summary

The article discusses the author’s experiences with different API technologies:

  1. OpenAPI vs. REST: The author clarifies that OpenAPI is not an alternative to REST but a documentation tool for HTTP APIs. OpenAPI can describe RESTful APIs or other styles, providing a schema for tools to interpret how the API functions.

  2. gRPC and Protocol Buffers: gRPC is highlighted as an efficient RPC mechanism using Protocol Buffers (protobufs). However, it lacks the comprehensive ecosystem and middleware support found in HTTP libraries. The author mentions that while Google’s original RPC layer “stubby” wasn’t open-sourced, gRPC serves as a good alternative but requires more custom development for features like logging or authentication.

  3. Challenges with gRPC: A key issue with gRPC is its reliance on .proto files, which necessitates that clients and servers use compatible versions. This makes gRPC less flexible for debugging or spontaneous interactions compared to HTTP APIs, where one can interact with the API directly using tools like curl.

  4. Open Source Contribution: The author has developed and open-sourced a Go library (oapi-codegen) that generates clients and servers from OpenAPI specifications, aiming to facilitate API development.

In summary, the article critiques the distinctions often made between API technologies, explains the operational differences between gRPC and HTTP APIs, and shares a personal contribution to the developer community.

2. Tech takes the Pareto principle too far

Total comment counts : 25

Summary

The article discusses the differences between the game development and tech industries in terms of product development stages and completion:

  1. Vertical Slice in Game Development: Game developers create a ‘vertical slice’ to demonstrate to publishers and investors how the finished game will look, feel, and play, showcasing their ability to deliver a polished product.

  2. MVP in Tech Industry: In contrast, tech startups often present a ‘Minimum Viable Product’ (MVP), which is the least functional product that still has market value. This approach focuses on quickly proving a concept to attract funding, but it often does not guarantee the team’s ability to complete the full product.

  3. Pareto Principle: The article critiques the tech industry’s reliance on the Pareto Principle (the idea that 20% of the effort yields 80% of the results), suggesting that while this can get products to a functional state, it often neglects the critical final stages of development necessary for a complete, polished product.

  4. Cultural Impact: The culture of accepting and even celebrating unfinished tech products has led to a broader acceptance of products that are not fully developed. This includes game releases with day-one patches and additional content (DLC) that feels essential to the core experience.

  5. Unfinished Products: Many tech products remain incomplete due to early market entry or because the technology or methodology might not be capable of achieving full functionality. This is particularly evident in AI applications like self-driving cars and generative AI, where current methods might not be sufficient to reach the desired end quality or functionality.

The article concludes by reflecting on historical over-optimism in AI development, suggesting that while initial breakthroughs can be impressive, the final stages of refinement are often underestimated or unattainable with current approaches.

Top 1 Comment Summary

The article critiques the author’s perspective on product development methodologies, particularly distinguishing between the MVP (Minimum Viable Product) and Vertical Slice approaches:

  1. Purpose of MVP vs. Vertical Slice: The critic argues that while both concepts aim at product development, their objectives differ significantly. MVP is about validating if there’s a market need for a new software product (Product-Market Fit or PMF), whereas Vertical Slice focuses on ensuring the product can compete in an existing market by delivering a full-featured slice of the product.

  2. Contextual Differences: The critic points out that video games, which often use a Vertical Slice, are competing within a known market, unlike startups which might be creating demand for something new.

  3. Validation vs. Completion: The discussion highlights that the 80/20 rule in MVP development is about validating the product idea quickly rather than perfecting the product. The examples of Magic Leap and Humane are used to illustrate the pitfalls of spending too long on product development without market validation. These products, despite being technically advanced and fully developed, failed because they did not meet an existing market demand, showcasing the risk of not using lean methodologies to test market assumptions early on.

Overall, the critique underscores the importance of understanding the market context and the purpose behind different product development strategies, emphasizing quick validation over prolonged development when entering new or uncertain markets.

Top 2 Comment Summary

The article discusses the inappropriate application of rapid development cycles, typical in non-critical software, to critical systems like medical devices, drone controls, and power plant safety systems. The author points out that:

  • Critical software, such as that used in medical devices or power plants, requires a more cautious and thorough development process due to the potential risks involved.
  • There’s a growing trend where people mistakenly apply the fast-paced development strategies of consumer software to these critical areas, leading to potential safety and reliability issues.
  • The example given is Tesla’s self-driving technology, suggesting it might have been developed with an approach more suited to less critical applications.
  • There is concern expressed about AI development, where similar fast development cycles could lead to dangerous outcomes if the technology is used in potentially harmful applications.

3. Master the Art of the Product Manager ‘No’

Total comment counts : 32

Summary

The article you provided appears to be a brief message or title rather than a full article, making it difficult to summarize. It seems to be an inquiry or a greeting directed towards an email address or a username, possibly about an “AI Prompt Deck.” If you have more context or a longer piece of text, please provide it for a proper summary.

Top 1 Comment Summary

The article discusses various types of negative responses to ideas within a business or team meeting context:

  1. Outright Rejection: The idea is immediately dismissed as bad.

  2. Conditional Rejection: The idea is considered potentially bad but could be revisited if a stronger case is made.

  3. Incomplete Idea: The idea needs further development before a decision can be made on its merit.

  4. Backlog Limbo: The idea is seen as good but not urgent, remaining in a backlog unless prioritized.

  5. Future Consideration: The idea is good but not currently a priority; it will be actively considered for future work.

The article emphasizes the importance of recognizing these nuances rather than categorizing all these responses simply as “no,” which can help manage expectations and maintain productive discussions. It also suggests that the responses might need to be softened or rephrased in corporate settings to manage interpersonal dynamics and keep meetings focused.

Top 2 Comment Summary

The article references a satirical four-step crisis management strategy from the TV show “Yes, Prime Minister”:

  1. Denial - Initially, the situation is minimized, suggesting that there’s no real issue to worry about.
  2. Observation - Acknowledging that something might be happening, but advocating for a wait-and-see approach rather than taking immediate action.
  3. Deliberation - Recognizing the need for action but failing to implement any specific or effective measures.
  4. Regret - After the problem escalates, acknowledging that there might have been opportunities to act, but now it’s too late.

The text critiques this approach as a way of managing perceptions rather than addressing problems, ultimately leading to disillusionment and loss of morale among those who feel ignored or misled.

4. Lossless Compression of Vector IDs for Approximate Nearest Neighbor Search

Total comment counts : 3

Summary

The article discusses arXivLabs, a platform by arXiv that enables collaborators to create and share new features for the arXiv website. It highlights that both individuals and organizations involved with arXivLabs uphold values such as openness, community, excellence, and user data privacy. Additionally, it invites users with project ideas to contribute to arXiv’s community. There’s also a mention of an operational status update feature where users can receive notifications via email or Slack about arXiv’s status.

Top 1 Comment Summary

The article discusses how in ANN (Approximate Nearest Neighbor) search, the primary constraint is typically memory bandwidth rather than CPU processing power. This means that the CPU often has spare cycles which can be utilized for tasks like decompression without significantly impacting the search latency.

Top 2 Comment Summary

The article points out that a comparison (presumably in another context or article) has omitted a discussion on “roaring bitmaps,” a data structure detailed in an older paper. The author of the comment believes that this oversight is surprising, especially since the paper on roaring bitmaps is not new and should have been considered.

5. F-Droid’s Progress and What’s Coming in 2025

Total comment counts : 17

Summary

2024 Achievements and Reflections for F-Droid:

  • Community and Support: F-Droid highlighted the community’s role in its success, acknowledging the support from donations, grants, volunteers, and contributors.

  • Decentralization Efforts: Significant progress was made in decentralizing app distribution:

    • Continued development on making F-Droid a more robust, accessible platform, reducing Big Tech’s influence on app distribution.
    • Infrastructure upgrades included breaking out core client logic, adding support for IPFS and Filecoin for app distribution, enhancing client capabilities, and supporting iOS and PWAs.

  • Grant Completion and Future: The Filecoin Foundation grant ended in 2024, but the established tools and policies will continue to develop with ongoing financial support.

  • Expansion of Ecosystem Tools:

    • A major EU Horizon Europe grant allowed for the enhancement of tools like Repomaker, aiding developers in creating F-Droid compatible repositories.
    • F-Droid’s involvement in the Mobifree initiative, aiming to create a decentralized alternative to traditional app stores, was highlighted. F-Droid encourages project proposals for Mobifree with funding available.

  • Community Contributions: There was a notable increase in community contributions, including bug fixes, app updates, and new app additions.

  • Looking Forward to 2025:

    • F-Droid plans to contribute further to Mobifree, with tools set for pilot testing and subsequent public release.
    • Two new team members were announced: Hailey Still as project manager and UX designer, and another unspecified role to help scale efforts.

This summary reflects F-Droid’s commitment to decentralization, community involvement, and the expansion of its ecosystem to foster a more open-source, privacy-focused app distribution platform.

Top 1 Comment Summary

The article discusses the limitations of the F-Droid app repository, particularly its lack of a download counter. The author argues that sorting apps by update time or alphabetically isn’t helpful for users trying to choose the best app among several that offer similar functionalities. They suggest that a download counter could provide insight into an app’s popularity and reliability, helping users determine which apps are worth installing by showing if users keep the app installed after trying it or remove it soon after. While F-Droid avoids implementing a full rating and review system, a simple download counter could serve as a useful metric for app quality and user satisfaction.

Top 2 Comment Summary

The article expresses frustration over the persistent issues with the user interface for updating apps in F-Droid, noting that despite over a decade passing, this core functionality remains buggy.

6. Tailwind CSS v4.0

Total comment counts : 37

Summary

Summary of Tailwind CSS v4.0 Release:

  • New Features and Enhancements: Tailwind CSS v4.0 has been completely rewritten to enhance performance, offering over 3.5x faster full rebuilds and 8x faster incremental builds. It leverages modern CSS features for better optimization and reduced complexity.

  • Installation and Setup: The setup process is now more streamlined with fewer steps and less boilerplate. A new Vite plugin integration provides better performance than the previous PostCSS setup.

  • Configuration Changes: Configuration has shifted from JavaScript to CSS, simplifying project management by removing the need for a separate tailwind.config.js file. Customizations are now directly in CSS where Tailwind is imported.

  • Automatic Source Detection: The new version automatically detects and excludes files from scanning based on heuristics like .gitignore settings and file types, reducing the need for manual configuration.

  • CSS Imports: Tailwind now handles CSS file imports natively, eliminating the need for additional plugins like postcss-import.

  • CSS Variables: Design tokens are automatically converted into CSS variables, making them easily accessible at runtime for dynamic styling or animations.

  • Documentation and Upgrade Guide: Comprehensive documentation and an automated upgrade tool are available to assist in migrating from previous versions to v4.0.

  • Performance: Significant performance improvements, especially in incremental builds which can be over 100x faster when no new CSS compilation is needed.

Tailwind CSS v4.0 aims to be more performant, easier to configure, and more integrated with modern web development practices, encouraging developers to start using it in new or existing projects.

Top 1 Comment Summary

The article discusses the author’s changed perspective on Tailwind CSS, a utility-first CSS framework. Initially critical of Tailwind for its incompatibility with modern web development practices, the author now praises the significant improvements in Tailwind’s version 4. Key advancements include:

  • Native CSS Variables: Tailwind now supports accessing its theme through native CSS variables, allowing for more traditional CSS usage alongside Tailwind’s utility classes.
  • CSS-only Configuration: The ability to configure Tailwind using only CSS, which enhances its compatibility with existing projects and reduces its intrusive nature.

The author suggests that these updates make Tailwind more of an additive utility rather than a framework that overshadows or dictates project architecture. They speculate that the criticism Tailwind received might have influenced these changes or that the developers independently reached these beneficial updates. This shift potentially resolves many previous debates about Tailwind’s utility and application, allowing developers to focus more on development and productivity.

Top 2 Comment Summary

The article discusses the evolution of CSS and its increased user-friendliness due to consistent browser behavior. It highlights the simplicity of CSS as it requires no build step and helps keep HTML markup clean. However, the author questions the trend of using inline style attributes in HTML, which mixes presentation with structure, a practice traditionally avoided for maintaining separation of concerns. The author finds this approach confusing and sees little benefit in terms of efficiency or simplicity, suggesting that traditional CSS usage might still be preferable.

7. The Most Detailed Map of US Waters That You’ve Ever Seen (2023)

Total comment counts : 13

Summary

The article discusses the significance of water in shaping landscapes and human life, detailing how water systems are engineered across different regions of the U.S. It highlights the launch of the National Hydrography Dataset Plus High Resolution (NHDPlus High Resolution) by the US Geological Survey, which offers an unprecedented detailed mapping of the U.S. water bodies. This dataset includes:

  • Over 32 million features, providing a detailed view of rivers, streams, lakes, and wetlands.
  • A significant increase in detail from the previous medium resolution dataset, with nearly 25 million linear features and over 7 million area features.
  • Enhanced capabilities for creating maps, analyzing data, and visualizing water distribution through new cartography and integration with ArcGIS tools.

The article also touches on how human activities, like agricultural irrigation and land drainage, have altered natural water patterns, as seen in examples from Louisiana, California, and Missouri. The new dataset allows for advanced visualizations, including 3D animations and integration with other geographical data layers, enhancing the understanding of water’s role in various landscapes and its management.

Top 1 Comment Summary

The article describes a detailed model or depiction of a property that includes small features like a narrow creek and a seasonal puddle on a neighbor’s land which forms after heavy rains.

Top 2 Comment Summary

The article discusses an issue with a water map that inaccurately repeats outdated information about seasonal streams on a property. The author suggests that the map should be updated or validated using modern technology like satellite data to improve accuracy.

8. Most Influential Papers in Computer Science History

Total comment counts : 43

Summary

The article lists seven influential papers in the field of computer science and information technology, each chosen for their significant impact on today’s technological landscape:

  1. Alan Turing - His work in the 1930s introduced the concept of a “Turing Machine,” laying the theoretical foundation for what computers could do, defining computability and influencing all subsequent computing models.

  2. Claude Shannon - Developed information theory, providing the mathematical framework for data compression, error correction, and data transmission, fundamentally shaping how we communicate digitally.

  3. Edgar F. Codd - Created the relational model for database management, which led to SQL and relational databases, revolutionizing data storage and retrieval.

  4. Stephen A. Cook - His introduction of NP-completeness clarified the complexity of computational problems, impacting areas like algorithms, cryptography, and problem-solving efficiency.

  5. Vint Cerf and Robert Kahn - Developed TCP/IP protocols, enabling the creation of the internet by allowing different networks to communicate seamlessly.

  6. Tim Berners-Lee - Invented the World Wide Web, proposing a system where documents could be linked via hypertext, which has become a cornerstone of the internet’s utility and accessibility.

Each paper or concept has not only shaped its specific field but also has had widespread implications across technology, influencing everything from how we use our devices to how data is managed and transmitted globally. The article emphasizes that while this list is subjective, these papers have undeniably set the stage for modern digital advancements.

Top 1 Comment Summary

The article discusses several seminal papers in computer science:

  1. Communicating Sequential Processes by C.A.R. Hoare - This paper introduces concepts for structuring concurrent systems, focusing on how processes can communicate safely and efficiently.

  2. The Next 700 Programming Languages by Peter Landin - Landin argues about the potential for an infinite number of programming languages, each optimized for certain tasks, highlighting the diversity and specialization in programming language design.

  3. As We May Think by Vannevar Bush - This visionary essay from 1945 imagines a future where machines could extend human memory and thought processes, leading to concepts like hypertext and personal computing.

  4. Can Programming Be Liberated from the von Neumann Style? by John Backus - Backus critiques the traditional von Neumann architecture’s influence on programming languages and advocates for a functional programming approach which could lead to more efficient and less error-prone code.

Additionally, the article mentions an interesting course offered at Harvard:

  • Computer Science Classics - This course at Harvard is designed for advanced computer science students to gain a comprehensive understanding of the field’s history and development. It covers influential papers from the 1930s onwards, providing insights into the foundational ideas and evolution of computer science through its primary literature, offering a “replay” of the field’s growth in a condensed form.

This course aims not just to survey the field but to give students a lived experience of its development, fostering a deeper, more unified view of computer science.

Top 2 Comment Summary

The article discusses the criteria for selecting influential papers in computer science (CS). It questions whether the focus should be on papers that have shaped the theoretical aspects of CS, like Alan Turing’s foundational work, or those that have practical technological impacts, such as the Internet Protocol (IP) standard. The author points out that while Turing’s paper is crucial for theoretical CS, its absence might not have significantly altered technological development. Conversely, the IP standard, while essential for technology, lacks scientific depth, being more of a practical specification. This leads to a broader reflection on whether CS is being used as a broad term encompassing both theoretical science and practical technology.

9. Where is London’s most central sheep?

Total comment counts : 27

Summary

The article provided appears to be a blog navigation and index page rather than a traditional article with a narrative or thematic content. Here’s a summary:

  • Blog Navigation: The page includes links for navigating through blog posts, with options to view newer or older posts, return to the main page, and subscribe to updates via email, Twitter, and RSS.

  • External Links: It lists numerous blogs related to London life, culture, and exploration, such as “arseblog,” “Londonist,” “Spitalfields Life,” among others.

  • Archive: There’s an extensive archive listing posts by month and year, starting from January 2025 back to January 2002, indicating a long history of content.

  • Special Features: The blog offers several special features like:

    • Lists of things to do in and around London.
    • Waymarked walks in London.
    • An Inner London toilet map.
    • A 20-year blog series.
    • A tour of Britain.
    • Various thematic explorations like London’s lost rivers, Olympic park, and more.

  • Photography: Links to the author’s Flickr photostream.

  • Thematic Posts: Mention of unique or interesting posts like “The Seven Ages of Blog,” “My New Z470xi Mobile,” and thematic explorations like “London’s Lost Rivers” and “Unlost Rivers.”

  • Miscellaneous: There are references to various other topics and interests, including London Underground, pop culture, historical events, and technology, suggesting a broad range of subjects covered by the blog.

Overall, this blog seems to be a comprehensive, long-running personal or thematic blog focused on London, its history, culture, and surrounding areas, with a rich array of content for enthusiasts of urban exploration and history.

Top 1 Comment Summary

The article discusses a personal metric created by the author and his wife to evaluate the livability of cities, named “time to sheep.” This metric measures the travel time from the city center to the countryside where one can see sheep, suggesting that cities with a moderate distance to rural areas are more enjoyable. The couple found Bristol preferable over London based on this metric, implying that easy access to nature enhances urban living. Additionally, the beauty of Bristol itself is humorously acknowledged as another reason for their preference.

Top 2 Comment Summary

The article linked is from the blog “Diamond Geezer” and contains an apology or correction, referred to as an “errata,” from the author. The summary provided in the text snippet indicates that there has been a mistake or oversight in a previous post or publication, leading to this correction or apology. However, without further details from the actual content of the linked page, the summary remains focused on the acknowledgment of an error.

10. A Proper x86 Assembler in Haskell Using the Escardó-Oliva Functional

Total comment counts : 5

Summary

The article discusses the challenge of writing an X86 assembler in Haskell, focusing on the problem of calculating the size of jump instructions which can vary between 2 and 5 bytes depending on the distance to their target labels. Here are the key points:

  1. Problem Description: Assembling X86 instructions requires knowing the exact size of jump instructions, which isn’t possible in a single pass because the size depends on the distance to the label, which can only be known after all instructions are laid out.

  2. Functional Programming Approach: The author uses the Escardó-Oliva functional, a higher-order function from functional programming, to solve this problem. This functional computes optimal moves in a game-like scenario where each move represents choosing between a short (2 bytes) or near (5 bytes) jump encoding.

  3. Implementation Details:

    • Types and Data Structures: The article defines types like Label, JIx (Jump Index), X86 a (an instruction), Enc (encoding type), Move (a pair of jump index and its encoding), and others to handle the assembly process.
    • Function ij: Converts a list of X86 instructions into a list where jumps are indexed for lookup.
    • Game Theory Application: The assembly process is likened to a single-player game where the goal is to minimize the total size of the code. The optimalPlay function uses the Escardó-Oliva functional to determine the best encodings for jumps.

  4. Outcome and Moves: The p function calculates the total bytes needed based on the current moves (encodings of jumps), and εs provides selection functions for choosing the next move based on past moves.

  5. Label Offsets and Validation: The ixes function computes the offsets of labels and jumps assuming initially all jumps are short, with the understanding that invalid configurations will be pruned later. The valid function checks if the current configuration of encodings is consistent with the assembly rules.

In summary, the article demonstrates an innovative use of functional programming concepts to solve a practical problem in low-level programming, specifically by using game theory and optimization techniques to manage the complexities of X86 assembly code generation.

Top 1 Comment Summary

The article discusses the memory efficiency of different programming approaches, specifically focusing on:

  1. Multipass Assemblers: These are efficient in memory usage as they only require keeping a list of symbols in memory, rather than the entire program, which is beneficial even if not a major concern with today’s large memory capacities.

  2. Haskell Programming: The text hints at some drawbacks of Haskell in terms of memory usage. It mentions that while Haskell can offer concise and clear code (like a simplified version of quicksort), these implementations might not be as memory-efficient due to inherent properties like immutability and the use of linked lists, which can introduce additional O(n) complexities and constant factors in performance. This results in what might be called “not really a quicksort” when comparing to traditional implementations.

Top 2 Comment Summary

The article discusses the complexity of achieving what appears to be a single pass operation in programming. It suggests that even though it might look like a single pass, there’s likely an underlying mechanism involving:

  1. Transitional Internal State: There’s probably some internal state that changes as the operation progresses, requiring looping or iteration over this state.

  2. Lazy Evaluation: The process might use lazy evaluation, where computations are deferred until their results are needed, which can hide the multiple passes or iterations.

  3. State Preservation: To manage where to jump or continue in the code, the system must keep track of the current state or “where” it is in the process, which involves preserving execution states.

  4. Completion and Return: The operation calculates or processes data and then returns to fill in any gaps or complete the operation, implying multiple stages or passes in the background even if presented as one pass.

In summary, the article argues that what might be described as a “one pass” operation often involves complex state management and deferred computation, essentially making it more than just one pass when considering the internal mechanics.