https://iamrajiv.github.io Software engineer building backend systems, developer platforms, and infrastructure in Go. en-US rajivperfect007@gmail.com (Rajiv Ranjan Singh) https://iamrajiv.github.io/writing/google-summer-of-code-2026/ https://iamrajiv.github.io/writing/google-summer-of-code-2026/ Mon, 26 Jan 2026 00:00:00 GMT How GSoC selection actually works, why the organization name matters far less than the project idea, and what separates a proposal that gets read from one that gets skimmed. https://iamrajiv.github.io/writing/making-sense-of-the-green-tea-garbage-collector/ https://iamrajiv.github.io/writing/making-sense-of-the-green-tea-garbage-collector/ Sun, 30 Nov 2025 00:00:00 GMT Go's new collector scans memory a block at a time instead of chasing individual pointers. The reason that helps says something about where performance actually goes on modern hardware. https://iamrajiv.github.io/writing/decoupling-web-worker-instantiation/ https://iamrajiv.github.io/writing/decoupling-web-worker-instantiation/ Wed, 23 Jul 2025 00:00:00 GMT A library that constructs its own web worker will break in someone else's bundler. Letting the caller construct it instead fixes every case at once, and the reasoning generalises. https://iamrajiv.github.io/writing/demystifying-remote-procedure-calls/ https://iamrajiv.github.io/writing/demystifying-remote-procedure-calls/ Fri, 04 Apr 2025 00:00:00 GMT RPC makes a network call look like a function call. That abstraction is why it is productive and also where its problems come from — both are worth understanding before you pick it. https://iamrajiv.github.io/writing/ten-habits-of-standout-developers/ https://iamrajiv.github.io/writing/ten-habits-of-standout-developers/ Sat, 22 Feb 2025 00:00:00 GMT Ten things I have seen separate the engineers I learned the most from — none of which are about knowing more than everyone else. https://iamrajiv.github.io/writing/profiling-webassembly-with-pprof-and-wzprof/ https://iamrajiv.github.io/writing/profiling-webassembly-with-pprof-and-wzprof/ Sat, 28 Sep 2024 00:00:00 GMT Compiling to WebAssembly usually costs you your profiler. wzprof instruments the runtime instead of the guest, which gets pprof back — for any source language. https://iamrajiv.github.io/writing/exploring-go-1-22-release/ https://iamrajiv.github.io/writing/exploring-go-1-22-release/ Thu, 22 Feb 2024 00:00:00 GMT Per-iteration loop variables, routing patterns in the standard library, and math/rand/v2 — the release that quietly fixed Go's most expensive design mistake. https://iamrajiv.github.io/writing/sliding-window-algorithmic-mental-model/ https://iamrajiv.github.io/writing/sliding-window-algorithmic-mental-model/ Sat, 10 Feb 2024 00:00:00 GMT Recognising a sliding window problem is most of the work. The pattern is a small set of signals in the question, and two variants that cover nearly all of them. https://iamrajiv.github.io/writing/api-gateways-load-balancers-reverse-proxies/ https://iamrajiv.github.io/writing/api-gateways-load-balancers-reverse-proxies/ Sun, 06 Aug 2023 00:00:00 GMT All three are one address that forwards traffic somewhere else. What separates them is not the software — it is what sits behind the address, and one diagram covers the whole distinction. https://iamrajiv.github.io/writing/higher-order-functions-in-go/ https://iamrajiv.github.io/writing/higher-order-functions-in-go/ Tue, 23 May 2023 00:00:00 GMT Collapsing repetitive entry/exit logging into one line per method with a function that returns a function — and the one-character mistake that silently breaks it. https://iamrajiv.github.io/writing/optimizing-performance-with-inlining/ https://iamrajiv.github.io/writing/optimizing-performance-with-inlining/ Tue, 23 May 2023 00:00:00 GMT Inlining removes a fixed cost of about a nanosecond per call. Whether that matters depends entirely on what the function body costs — and a benchmark that ignores this will mislead you. https://iamrajiv.github.io/writing/demystifying-dangling-pointers/ https://iamrajiv.github.io/writing/demystifying-dangling-pointers/ Sat, 20 May 2023 00:00:00 GMT A pointer to freed memory usually keeps working for a while, which is what makes it dangerous. The four ways C++ produces them, and the tools that catch what discipline misses. https://iamrajiv.github.io/writing/demystifying-columnar-storage-model/ https://iamrajiv.github.io/writing/demystifying-columnar-storage-model/ Thu, 11 May 2023 00:00:00 GMT Storing a table by column instead of by row changes what compresses well and what queries cost. A Go benchmark shows the effect, and also shows why the usual demonstration of it is misleading. https://iamrajiv.github.io/writing/getting-started-with-google-summer-of-code/ https://iamrajiv.github.io/writing/getting-started-with-google-summer-of-code/ Sat, 06 May 2023 00:00:00 GMT An introduction to open source and Google Summer of Code — what the program is actually for, how to pick a project, and what makes a proposal worth reading. https://iamrajiv.github.io/writing/improving-software-performance-with-buffers/ https://iamrajiv.github.io/writing/improving-software-performance-with-buffers/ Sat, 06 May 2023 00:00:00 GMT Writing 100,000 lines to a file takes 9ms unbuffered and 3.5ms buffered. The reason is system calls, and the same reasoning applies far beyond file I/O. https://iamrajiv.github.io/writing/go-resources/ https://iamrajiv.github.io/writing/go-resources/ Sat, 05 Mar 2022 00:00:00 GMT The resources I'd hand someone learning Go, ordered the way I'd actually work through them rather than dumped as a list. https://iamrajiv.github.io/writing/theoretical-runtime-analysis-of-big-o-notation/ https://iamrajiv.github.io/writing/theoretical-runtime-analysis-of-big-o-notation/ Sun, 27 Feb 2022 00:00:00 GMT A loop bound by a problem constraint is technically O(1). Working out why that is true and why it is useless says something real about what Big O measures. https://iamrajiv.github.io/writing/measure-elapsed-time-of-cpp-program/ https://iamrajiv.github.io/writing/measure-elapsed-time-of-cpp-program/ Tue, 22 Feb 2022 00:00:00 GMT Timing a block of C++ with <chrono>, why steady_clock is the right clock for the job, and how to read the same duration in four different units. https://iamrajiv.github.io/writing/linux-foundation-mentorship-2021/ https://iamrajiv.github.io/writing/linux-foundation-mentorship-2021/ Wed, 29 Dec 2021 00:00:00 GMT Twelve weeks on moja global's FLINT UI — the frontend for a carbon-accounting engine — building out a component library, CI, and the infrastructure around it. https://iamrajiv.github.io/writing/google-season-of-docs-2021/ https://iamrajiv.github.io/writing/google-season-of-docs-2021/ Thu, 02 Dec 2021 00:00:00 GMT A second season, this time with Wechaty — improving the gRPC and OpenAPI ecosystem, and learning how a distributed community actually keeps a project moving. https://iamrajiv.github.io/writing/kharagpur-winter-of-code-2020/ https://iamrajiv.github.io/writing/kharagpur-winter-of-code-2020/ Sun, 03 Jan 2021 00:00:00 GMT Five weeks contributing to three open-source projects, all of it CI and tooling — and why that turned out to be a good place for a beginner to start. https://iamrajiv.github.io/writing/google-season-of-docs-2020/ https://iamrajiv.github.io/writing/google-season-of-docs-2020/ Thu, 10 Dec 2020 00:00:00 GMT Rebuilding the gRPC-Gateway documentation site over three months — a new theme, a tutorials section written from scratch, and what I learned about documenting something by using it. https://iamrajiv.github.io/writing/an-efficient-algorithm-for-generating-set-partitions/ https://iamrajiv.github.io/writing/an-efficient-algorithm-for-generating-set-partitions/ Thu, 30 Jan 2020 00:00:00 GMT Generating every partition of an integer takes exponential time because there are exponentially many. Counting them takes quadratic time — and the difference is worth understanding. https://iamrajiv.github.io/writing/finding-nth-fibonacci-number-in-logarithmic-time/ https://iamrajiv.github.io/writing/finding-nth-fibonacci-number-in-logarithmic-time/ Sun, 26 Jan 2020 00:00:00 GMT A 2x2 matrix identity turns Fibonacci into exponentiation, and fast exponentiation takes it from linear to logarithmic time. https://iamrajiv.github.io/writing/monty-hall-problem/ https://iamrajiv.github.io/writing/monty-hall-problem/ Thu, 18 Jul 2019 00:00:00 GMT Why switching doors wins two times out of three, worked through by enumeration and then by Bayes' theorem — including the variant where the answer really is 50–50. https://iamrajiv.github.io/writing/discriminant-of-a-conic-section/ https://iamrajiv.github.io/writing/discriminant-of-a-conic-section/ Thu, 24 Jan 2019 00:00:00 GMT Two numbers — a determinant and a discriminant — tell you which conic a second-degree equation draws, without plotting a single point. https://iamrajiv.github.io/writing/initial-commit/ https://iamrajiv.github.io/writing/initial-commit/ Thu, 20 Sep 2018 00:00:00 GMT Why I started writing here, and what I intend to put in it.