About:

Will Keleher is a software engineer passionate about ed-tech, writing, D&D, and punctuation, with a background in teaching English in rural China.

Website:

Specializations:

Interests:

Performance Site-reliability Team effectiveness Ed-tech Writing Dungeons & Dragons Fantasy novels 3d-printing Wordplay Punctuation

Outgoing Links:

Subscribe to RSS:
Optimizing SQL queries can significantly reduce database load by employing strategies like early exits and ratcheting, enhancing performance without complex changes.
The post argues that using literal and optional parameters in TypeScript can obscure code clarity, advocating for named parameters to enhance readability and maintainability.
MySQL's query planner struggles with `ORDER BY` and `LIMIT` for multiple `gid` values, leading to inefficiencies that can be mitigated using a `UNION` strategy.
Efficient SQL query performance relies on understanding index usage, selectivity, and the benefits of compound indexes over index merging.
Using request_uuid in database tables simplifies tracking changes and debugging, enhancing overall system legibility and efficiency.
Effective request logging is essential for debugging and analysis, with strategies to minimize costs while capturing valuable data insights.
The post highlights the importance of slow query logs in database optimization and the need for metadata to trace query origins for effective performance improvements.
The post examines the struggle to convey authenticity in digital communication while addressing the limitations of formats and the nuances of language.
The post explores the tension between personal writing style and technical limitations, highlighting the author's emotional investment in their craft.
Effective code review of LLM-generated code requires trunk-based development, local file checks, and reliable diff viewers, while recognizing LLMs' limitations in providing accurate feedback.
Caution is advised when using LLMs for code generation, as writing code enhances understanding and critical thinking, which LLMs may inhibit.
The author argues against using the 'find' command for file searches, recommending globbing and the 'fd' utility as more efficient alternatives.
Automating CSS and code migrations through scripting can significantly enhance efficiency and maintainability, especially when combined with AI tools.
Practical shell scripting techniques and patterns are shared, focusing on string manipulation, error handling, and improving script readability and efficiency.
This post provides insights into enhancing shell command usability through various techniques and encourages experimentation for learning shell scripting.
Categorizing TODO comments in code can enhance their utility and manageability, making tasks more searchable and scriptable within a codebase.
Essential Lodash functions can significantly improve code readability and performance in functional programming by simplifying data organization and transformation.
The author explores the joy of customizing their personal website with CSS techniques, including animations and pseudo-elements, while sharing practical tips and personal reflections.
The post details the author's attempts to create an image of a seven-legged spider using AI, revealing both advancements and limitations in the technology.
The author explores the challenges of generating a spider silhouette with a missing leg using AI tools, revealing their limitations in image generation.
Writers can enhance their productivity in Google Docs by using advanced find-and-replace with regular expressions, pasting without formatting, and leveraging Markdown for efficient text formatting.