How we made a Ruby method 200x faster (5 minute read)
A performance regression in Campsite's API was identified through observability tools after code was refactored to transform rich text into plain text and Markdown. Profiling revealed that the bottleneck was the use of a certain node lookup. This was replaced with a more efficient HANDLERS_BY_NODE_NAMES hash lookup, which reduced processing time from 200ms to under 1ms.
|
The Practical Guide to Scaling Django (4 minute read)
Common performance bottlenecks in Django include database queries, template rendering, Python processing, cache misses, file I/O, and network latency. This guide details optimization techniques for each bottleneck, including database query optimization, caching strategies (view-level, fragment, and low-level API), asynchronous operations, background tasks, load balancing, and content delivery network (CDN) usage.
|
Optimizers: The Low-Key MVP (17 minute read)
DuckDB's query optimizer improves query performance, often by orders of magnitude, compared to both unoptimized and even hand-optimized queries. This improvement stems from several optimization rules, including filter pushdown, join order optimization, and TopN optimization, which automatically adapts to changing data conditions.
|
|
Good software development habits (5 minute read)
This developer writes about a few software development habits that have helped him write better software faster. They focus on writing small commits frequently, refactoring continuously, and deploying often. They also often write tests first to define API structures early.
|
You Can't Build Interactive Web Apps Except as Single Page Applications... And Other Myths (11 minute read)
Modern browsers, coupled with technologies like service workers and the Speculation Rules API, significantly improve Multi-Page-Application (MPA) performance, addressing issues like slow page transitions and offline capabilities. Specific techniques, such as caching strategies and Workbox for service worker automation, can be used to make MPAs more efficient. Interactive web apps aren't only for Single Page Applications (SPAs).
|
|
Fast Grid (GitHub Repo)
Fast Grid is a performant multithreaded web table that's capable of running at 120FPS while sorting, filtering, and scrolling simultaneously.
|
Literal (GitHub Repo)
Literal is a Ruby runtime type checker that provides types, structured objects, properties, and enums. Its built-in types include generic structures that enable some cool use cases like checking an array is all integers.
|
|
Why did Windows 95 setup use three operating systems? (5 minute read)
Windows 95's setup process utilized three operating systems (MS-DOS, Windows 3.1, and Windows 95) to handle upgrades from various starting points. Instead of creating three separate setup programs, developers cleverly designed a chained system using an MS-DOS program to install a minimal Windows 3.1 version, followed by a 16-bit Windows program for core setup tasks, and finally a 32-bit Windows 95 program for final configurations, which minimized code duplication.
|
YC is wrong about LLMs for chip design (8 minute read)
LLMs produce mediocre Verilog code and cannot design novel chip architectures, unlike skilled engineers. While LLMs can reduce chip design costs, benefiting large companies and potentially smaller startups by enabling leaner teams, they won't create significantly superior chips or open new markets.
|
|
|
Love TLDR? Tell your friends and get rewards!
|
|
Share your referral link below with friends to get free TLDR swag!
|
|
|
|
|
Track your referrals here.
|
|
Want to advertise in TLDR? 📰
|
If your company is interested in reaching an audience of web developers and engineering decision makers, you may want to advertise with us.
If you have any comments or feedback, just respond to this email!
Thanks for reading,
Priyam Mohanty, Jenny Xu & Ceora Ford
|
|
|
|
