<link>https://mikelayuso.com/</link></image><item><title>Microdependency Hell: The Cost of Package Culture</h1> <article> <h1>Microdependency Hell: The Cost of Package Culture</h1> <p>Sun, 17 May 2026 00:00:00 +0000</p> <p>One thing I highlighted in my almost as an afterthought: the Rust binary ships with <em>no runtime dependencies</em>. Download it, run it, nothing to install. The more time I spend across C/C++, Rust, and Godot, the more that property looks less like a convenience and more like a design goal worth chasing on purpose, because the alternative has a cost.</p> <h2 id="the-incident">The incident</h2> <p>A perfect example in the JavaScript ecosystem is what happened on March 22, 2016. Azer Koçulu published a small npm package called <code>left-pad</code>, 17 lines that padded strings on the left. When Kik Messenger wanted the package name <code>kik</code> for their own use, npm sided with the company and transferred ownership away from Koçulu. In protest, he unpublished all 273 of his packages. When <code>left-pad</code> vanished, Babel and React Native broke. Thousands of projects at Facebook, PayPal, Netflix, Spotify and many others stopped building. Not because of a sophisticated attack. Because a justifiably angry maintainer deleted code he had written for free.</p> <h2 id="how-the-ecosystem-got-here">How the ecosystem got here</h2> <p>npm hosts over 4 million packages, not a measure of creativity, but of incentive. Each package is its own GitHub repo, its own resume line, and its own download counter. The ecosystem rewards fragmentation, because each successful package is a credential with a counter attached, and the ecosystem has never learned to care what the counter is counting. The result is packages like <code>is-odd</code>, which checks whether a number is odd, depending on <code>is-number</code>, depending on <code>kind-of</code>. Three packages, three maintainers, three supply chain entry points for a question you can answer in one line. A typical React or Next.js app pulls between 1,200 and 1,800 transitive dependencies. No human reads them. When I’m building something in Godot, I don’t <code>npm install physics</code>, the engine ships with it, one curated trust boundary maintained by a team. That contrast is hard to unsee.</p> <h2 id="the-real-risk">The real risk</h2> <p>Tools like Snyk scan for <em>known</em> vulnerabilities, by definition, ones that have already burned someone else. What they don’t catch: a package three levels deep maintained by someone who hasn’t logged in years, whose npm account gets taken over by an untrusted maintainer, who publishes a new version with a helpful postinstall script that reads your environment variables. This isn’t hypothetical. In 2018, <code>event-stream</code> was handed off to a stranger who volunteered to help, then used that access to exfiltrate Bitcoin private keys, downloaded 2 million times a week for two months before anyone noticed. In 2022, the maintainer of <code>colors</code> and <code>faker</code> pushed a self-sabotaging update as a protest. No hack, no zero-day. A maintainer in a bad mood, 20 million weekly downloads, production builds broken overnight. These aren’t bugs in the trust model. They <em>are</em> the trust model.</p> <h2 id="the-correction">The correction</h2> <p>Sindre Sorhus, the person more responsible than anyone for the small-modules culture, now publicly argues for fewer dependencies and bigger standard libraries. Node already added native fetch, a test runner, and SQLite without a single install in the last few years. The platform is starting to absorb what fragmentation spread. The instinct to keep surface area small, to understand what runs in your process, isn’t NIH (Not Invented Here) syndrome. It’s engineering. The JavaScript ecosystem spent 15 years treating it as optional. The supply chain incidents are the cost.</p> </article> <article> <h1>Shift - Pomodoro Game</h1> <p>Sat, 09 May 2026 00:00:00 +0000</p> <p>Shift turns your productivity sessions into a chill drive, helping you stay focused and build momentum without being intrusive.</p> </article> <article> <h1>From a Forgotten Python Script to a 2500x Faster Rust Tool: The Evolution of PNGToSVG</h1> <p>Sun, 19 Apr 2026 00:00:00 +0000</p> <p>I just wanted to share the story behind a tool I’ve been working on called . As you can probably guess, it converts PNG images into SVG vectors.</p> <p>The project started back in 2019 as a small Python script I wrote for a frontend job. Sometimes we needed to work with SVGs, and I just wanted a quick way to convert images locally without uploading them to some random remote service. When I switched jobs, I took that code, tossed it onto GitHub, and completely forgot about it.</p> <p>Fast forward five years. Out of nowhere, a contributor (“Kartik Nayak” on GitHub) made the first Rust implementation of the tool. A few months later, someone else (“Salman Sali”) jumped in and improved it by adding parallelization.</p> <p>This was late 2024. Seeing the community breathe new life into my old script blew my mind and gave me the perfect excuse to dive in and learn Rust myself. Since then, I’ve been rewriting, polishing, and tweaking the tool, focusing heavily on ease of use and performance.</p> <p>What used to be a sluggish script that took a couple of seconds to process a tiny 64x64 icon can now chew through 8000x8000 images in a fraction of that time. To put this performance leap into perspective, I ran a benchmark using a relatively complex ~900KB image.</p> <p>Here is how the evolution of the project looks:</p> <table> <thead> <tr> <th>Version</th> <th>Time (Seconds)</th> <th>Relative Performance</th> <th>Speedup Strategy</th> </tr> </thead> <tbody> <tr> <td><strong>Python (Legacy)</strong></td> <td>1,124.14s</td> <td>100% (Baseline)</td> <td>Original implementation</td> </tr> <tr> <td><strong>v0.4.0 / v0.4.1</strong></td> <td>~24.50s</td> <td>~2.18%</td> <td>Initial Rust Port (O(n) lookups)</td> </tr> <tr> <td><strong>v0.5.0</strong></td> <td>1.15s</td> <td>0.10%</td> <td>O(1) Hash-Set Optimization</td> </tr> <tr> <td><strong>v0.6.0</strong></td> <td>0.43s</td> <td>0.04%</td> <td>Memory Reuse & Direct IO</td> </tr> </tbody> </table> <p>Going from over 18 minutes in the original Python implementation to under half a second in Rust represents a massive <strong>~2580x speedup</strong>. But the numbers only tell half the story. Translating this project into Rust became an amazing learning exercise, and the technical journey went roughly like this:</p> <ul> <li><strong>Phase 1: The Rust Port (v0.4.x):</strong> Moving to Rust gave us a huge baseline boost (about 45x faster than Python), but the code still relied on linear searches for neighbor lookups, which struggled under the heavy weight of complex paths.</li> <li><strong>Phase 2: The Breakthrough (v0.5.0):</strong> As I learned more, I realized we could replace vector-based neighbor checks with O(1) Hash-Set lookups. This single change killed the quadratic scaling bottleneck and dropped processing time from 24 seconds to just 1.1 seconds—a 20x jump just within the Rust codebase.</li> <li><strong>Phase 3: Total Efficiency (v0.6.0):</strong> Finally, I focused on memory efficiency. Instead of constantly allocating new structures for every shape, we now reuse memory buffers, read pixel data directly via raw buffers, and use sorted edge lookups to trace shapes.</li> </ul> <p>Recently, I published the latest release (v0.6.1), which polishes a few rough edges and includes a shiny new (and very “programmer art”) icon, hahaha. Since then, I’ve received plenty of feedback from the community. Based on that, I’m already planning the improvements for v0.6.2, as well as a v0.7.0 release (which will include some breaking API changes, hence the version bump).</p> <p>Honestly, the best part of this whole experience hasn’t been the crazy performance gains. It’s the fact that randomly deciding to share a quick script allowed me to receive very useful community contributions, learn a completely new language, and actually build a fast, easy-to-use tool that actually helps me without interrupting my workflow with wait times in my own projects.</p> </article> <article> <h1>PNGToSVG - High-Performance CLI & Library</h1> <p>Wed, 20 Nov 2024 00:00:00 +0000</p> <p>PNGToSVG is a high-performance tool written in Rust that converts PNG raster images into SVG vectors. Originally developed as a Python prototype, it has been completely rewritten in Rust to maximize speed, portability, and ease of use.</p> <h2 id="features">Features</h2> <ul> <li><strong>⚡ Rust Performance</strong>: Significantly faster than the original Python implementation, processing images in milliseconds.</li> <li><strong>📦 Standalone Binary</strong>: No complex environments or dependencies required. Just download and run.</li> <li><strong>🖱️ Drag & Drop (Windows)</strong>: Designed for ease of use; just drag your PNG files onto the executable.</li> <li><strong>🛠️ CLI & Library</strong>: Fully scriptable for power users and easily integrated into other Rust applications.</li> </ul> </article> <article> <h1>Laser Metal Deposition (LMD) Process Monitoring: From 3D Visualization of Sensor Data Towards Anomaly Detection</h1> <p>Wed, 16 Oct 2024 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>Metal Additive Manufacturing (AM) allows producing geometrically complex metal components, unlocking new design possibilities and making it suitable to sectors such as healthcare, automotive and aerospace. AM processes are complex and require the use of many sensors to extract relevant process information for its monitoring and control. In the last years, many studies have applied advanced Deep Learning methods to extract knowledge from AM processes. However, these developments are specific to a particular setup, problem or defectology. Furthermore, they lack frameworks and pipelines to guide throughout their development, and do not include AI-related tools for data labelling, visualization, and AI model development and deployment. With the aim of simplifying the development and deployment of AM process monitoring systems, a dashboard-based framework that makes use of AI for anomaly detection and for feature extraction is presented in this study. The framework helps with development and deployment of monitoring systems by easing the incorporation of new sensors and the extraction of new features from captured data by end users. In this study, a Laser Metal Deposition (LMD) process is considered as the use case to show the usefulness of the developed framework.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>3D visualization framework for LMD process monitoring</li> <li>Sensor fusion techniques for comprehensive process data</li> <li>Machine learning approaches for anomaly detection</li> </ul> </article> <article> <h1>Advanced AI Architecture Through Asset Administration Shell and Industrial Data Spaces</h1> <p>Thu, 27 Jun 2024 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>This article develops an architecture for the implementation of Artificial Intelligence in the manufacturing value chain based on standard technologies and data spaces. The standards considered are IEC 63278 “Asset Administration Shell (AAS) for industrial applications” and DIN SPEC 27070:2020 – “Requirements and reference architecture of a security gateway for the exchange of industry data and services“ by IDSA. The architecture provides a data space that allows MONDRAGON industrial cooperatives to use data for the execution of advanced data analytics, Artificial Intelligence (AI) algorithms and interoperability between assets and IoT-platforms. The development of knowledge in this field allows, on the one hand, to optimise the consolidation of data as a strategic factor and, on the other hand, to increase collaboration between manufacturing companies, suppliers and technology providers. The article also explores specific Artificial Intelligence technologies with a wide application in industrial environments. In particular, the study has focused on research into Low/No Code, Explainability (XAI) tools and incremental learning algorithms. The contributions of this paper are summarised in 1) creating an IDS-AAS based architecture and data space that allows the exploitation of AI use cases, either by directly downloading models or by using AI as a service, 2) identifying useful AI tools for industry such as AutoML, No/Low code, XAI or incremental learning, 3) implementing a use case where different AI use alternatives are implemented.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Framework for AAS-based AI system integration</li> <li>Data sovereignty approaches using IDS principles</li> <li>Practical implementation guidelines for smart factories</li> </ul> </article> <article> <h1>Scrawled Seas</h1> <p>Wed, 01 May 2024 00:00:00 +0000</p> <p>Reach the finish line before your opponents, earn money, optimize your deck and win the pirate racing championship!</p> </article> <article> <h1>Block Lab</h1> <p>Sun, 15 Oct 2023 00:00:00 +0000</p> <p>Place blocks in the desired position by dragging and dropping to generate a solution, then press “Play” to run the simulation.</p> </article> <article> <h1>Demo-Magnet</h1> <p>Fri, 20 Jan 2023 00:00:00 +0000</p> <p>A short platformer prototype exploring magnet-based mechanics.</p> </article> <article> <h1>SALD</h1> <p>Mon, 15 Aug 2022 00:00:00 +0000</p> <p>SALD is an OpenGL wrapper and engine developed to simplify the process of prototyping graphical applications and learning modern OpenGL.</p> </article> <article> <h1>Hyperconnected Architecture for High Cognitive Production Plants</h1> <p>Tue, 21 Sep 2021 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>The HyperCOG project addresses the full digital transformation of process industry through an innovative Industrial Cyber-Physical System and Data Analytics. It is based on advanced technologies that enable the development of a hyperconnected network of digital nodes. The nodes can catch outstanding streams of data in real-time, which together with the high computing capabilities, provide sensing, knowledge and cognitive reasoning, making companies robust in the face of variant scenarios. The breaking-edge system proposed in this work is validated on productivity, environmental and replicability aspects on three use cases of three different sectors: steel, cement and chemical.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Architecture for seamless data integration across production systems</li> <li>Cognitive layer design for intelligent decision support</li> <li>Implementation guidelines for Industry 4.0 transition</li> </ul> </article> <article> <h1>A Novel Architecture for Cyber-Physical Production Systems in Industry 4.0</h1> <p>Mon, 23 Aug 2021 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>The Hyperconnected Architecture for High Cognitive Production Plants (HyperCOG) project aims at the process industry’s complete digital transformation through an advanced Industrial Cyber-Physical Infrastructure. It is based on advanced technologies that allow a hyperconnected network of digital nodes to be created improving the classic automation hierarchy of communication layers. The nodes will collect data streams in real-time, offering cognitive sensing and information along with high performance computing capabilities making the process industry businesses solid in different scenarios. The system is validated in three fields of the process industry: steel, cement and chemical where optimization in the use of energy and raw materials is obtained, among other benefits.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Framework for CPPS implementation in smart factories</li> <li>Integration patterns for legacy and modern manufacturing equipment</li> <li>Case studies demonstrating improved production efficiency</li> </ul> </article> <article> <h1>JMarkPad</h1> <p>Wed, 10 Mar 2021 00:00:00 +0000</p> <p>JMarkPad is a minimalistic markdown editor designed for simplicity and speed, featuring a real-time preview of your documents.</p> </article> <article> <h1>Wild Rum</h1> <p>Sun, 20 Sep 2020 00:00:00 +0000</p> <p>Developed for the IndieDevDay Game Jam 2020. Un hombre del salvaje oeste ha tomado más copas de las que debería… ¡Ayúdale a llegar sano y salvo!</p> </article> <article> <h1>Pirate King</h1> <p>Mon, 15 Jun 2020 00:00:00 +0000</p> <p>Pirate King is a university project exploring simultaneous-turn-based mechanics on a board game layout.</p> </article> <article> <h1>Sokoban Editor</h1> <p>Sun, 10 Nov 2019 00:00:00 +0000</p> <p>A classic Sokoban implementation featuring a robust level editor for creating and sharing puzzles.</p> </article> <article> <h1>Arrownaut</h1> <p>Tue, 28 May 2019 00:00:00 +0000</p> <p>Navigate through levels and defeat enemies using your trusty bow in this space-themed platformer.</p> </article> <article> <h1>A new method for surface crack detection by laser thermography based on Thermal Barrier effect</h1> <p>Sun, 31 Dec 2017 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>The aim of this work is to detect open surface cracks on metallic welds using laser thermography and posterior processing algorithms. In the last years the trend on welding processes has been to extend their use to applications with higher stress conditions. As a consequence, obtained welded regions are more prone to cracks. The types of cracks originated from welding processes are, in general, superficial and their detection is not straightforward since it detection is affected by their width. Traditional crack detection Non Destructive Techniques, such as Dye Penetrant or Magnetic Particles, are time consuming and require an experienced technician to correctly interpret the results. Furthermore, they foul the surface which needs to be cleaned after the test. Laser Thermography [5] [10] has several advantages over these traditional techniques: it is fast, non-contact and clean. In this paper, dynamic tests have been performed by scanning the weld maintaining the laser and camera motionless, and moving the piece linearly across them, so that the weld is scanned. The use of a scanning laser line to heat the surface allows the detection of the characteristic disruption that cracks generate in the natural diffusion of heat on the surface. Briefly, cracks block the heat front. This makes the temperature just before the crack to increase, while keeping the temperature of the region just after it at similar values for larger times than for other nearby regions. This thermal effect is known as “Thermal barrier” [11]. Here, a new method to detect the cracks based on this effect has been developed starting from two different approaches.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Novel laser thermography setup optimized for surface crack detection</li> <li>Thermal Barrier effect modeling for improved signal processing</li> <li>Experimental validation on various material types</li> </ul> </article> <article> <h1>Sticky Racer</h1> <p>Tue, 05 Dec 2017 00:00:00 +0000</p> <p>Your objective is to complete 3 laps as fast as possible but you will leave a trail of glue as you advance.</p> </article> <article> <h1>Spot Welding Monitoring System Based on Fuzzy Classification and Deep Learning</h1> <p>Fri, 30 Jun 2017 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>This work is a continuation of our previous work on the development of a monitoring system of a Spot Welding production line. Here we use the process information and photographs of more than 150,000 parts to improve the predictions of the previously developed fuzzy algorithm to predict the degradation state of the electrode. And, we present an alternative method based on deep-learning that aims at substituting the image analysis software developed by us to extract values associated with the quality level of the welded parts from photographs. The deep-learning algorithm learned here is applied to compress original photographs to a 15×15 pixels size image using an encoding / decoding model. Obtained compressed images are then used to predict quality parameters from a fuzzy rule-based classification algorithm. The results are promising and show that compressed images keep the relevant information from the original image that serve to directly determine the degree of the degradation of the electrode without requiring the use of previously developed image analysis software.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Hybrid fuzzy-deep learning architecture for weld quality classification</li> <li>Real-time monitoring capabilities for production environments</li> <li>Comprehensive validation across different welding scenarios</li> </ul> </article> <article> <h1>Electrode Degradation Analysis in Aluminium-Based Resistance Spot Welding</h1> <p>Thu, 30 Jun 2016 00:00:00 +0000</p> <h2 id="abstract">Abstract</h2> <p>This work presents the preliminary analysis done to determine the electrode degradation during a resistance spot welding manufacturing process of aluminium-based unions. This process represents a critical step in a high production rate manufacturing line, where currently identical welding parameters are used for the tens of thousands of dayly produced parts. During the welding process, the aluminium can melt locally, producing droplets that adhere to the electrode speeding its degradation that affects the quality of the joint. Degradation is hard to predict since it depends on part surface quality, electrode preparation and system set-up. In this work, an analysis of the degradation of the electrode associated with the observed and measured quality of the parts is carried out. With the final aim to advance toward a zero defect manufacturing scenario and increase the quality of the produced parts, a process data analysis based on fuzzy logic algorithms will be performed. An assessment of (i) produced part quality, (ii) the electrode degradation state, (iii) detection of different performance modes of the parameters and (iv) detection of events or process changes, will be done using image and production line data. This will allow to determine (i) when the electrode must be changed and (ii) the optimal combination of welding parameters to extend electrodes’ life assuring welding quality.</p> <h2 id="key-contributions">Key Contributions</h2> <ul> <li>Detailed characterization of electrode degradation patterns</li> <li>Analysis of process parameters affecting electrode lifespan</li> <li>Recommendations for electrode maintenance and replacement</li> </ul> </article> </main></body></html>