Developer Finds Silent Java I/O Bug After Refactoring Bare-Metal HTTP Server

·1 views

A developer building an HTTP server in Java from raw sockets — without any frameworks — encountered a subtle bug after refactoring the initial single-method code into separate classes. The server accepted connections and ran without throwing any exceptions, yet browsers received no response and hung indefinitely. The root cause was a missing flush() and close() call on the PrintWriter after the refactor, meaning buffered response data was never actually sent to the client. The incident highlighted how Java's layered I/O constructors — wrapping OutputStreamWriter in BufferedWriter inside PrintWriter — exist to manage the high cost of kernel-level I/O operations through buffering. The author is documenting the project as an ongoing series to explore low-level networking concepts in Java.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

How C# Reflection Eliminates Repetitive Code With Runtime Type Inspection

Reflection is a C# feature available since .NET Framework 1.0 in 2002 that allows code to inspect and interact with types, methods, and properties at runtime without prior compile-time knowledge. It resides in the System.Reflection namespace and is widely used, often indirectly, by .NET developers. A practical illustration involves a dynamic email template engine, where without reflection, developers must write a separate merge method for every email type, leading to code duplication, poor scalability, and hard-to-catch silent bugs. Using reflection, a single generic Render() method can inspect any object's properties at runtime and replace matching placeholders in a template, eliminating the need for type-specific methods entirely. This approach means new email types can be added without modifying the rendering engine, making the codebase easier to maintain and extend.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Indian developer builds lightweight image editor APIC to fill gap between Photoshop and GIMP

A 13-year-old solo developer from India has created APIC (Advanced Image Processing Center), a free desktop image processing tool, after finding existing options either too costly or too bulky. Photoshop's subscription cost and GIMP's 200MB install size left no practical middle ground for tasks like batch converting 50 images. APIC supports format conversion across PNG, JPG, WEBP, BMP, TIFF, and ICO, along with compression, cropping, resizing, and batch processing. The tool ships as a single executable file of just 95MB and includes drag-and-drop support and a system-wide image search feature. It is available for free download on GitHub.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Why SQL Mastery Requires a Mindset Shift From Loops to Set-Based Thinking

A technical perspective published on DEV Community argues that SQL is far more than a query language — it is a distinct way of reasoning about data that demands declarative, set-based thinking rather than imperative, loop-driven logic. The article contends that developers who avoid SQL by hiding it behind ORMs often encounter serious performance problems that force a reckoning with the language. Key principles highlighted include the importance of schema design, the nuanced role of indexes as access-path hints rather than universal speed boosters, and reading execution plans before attempting optimizations. The piece also warns that concurrency issues such as isolation levels and transactions are frequent sources of data corruption when misunderstood. Ultimately, the author concludes that abstracting away SQL does not reduce system complexity — it merely relocates it to places that are harder to diagnose and fix.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How Human Memory Works: Multiple Systems, Each With a Distinct Role

Memory is not a single system but a collection of parallel systems — sensory, working, and long-term — each handling different types of information across different timescales. Working memory holds roughly seven items for seconds to minutes, while long-term memory stores information for years with virtually unlimited capacity. Research by Ebbinghaus showed that forgetting is steepest immediately after learning, but spaced repetition — reviewing material at intervals of one day, one week, and one month — significantly slows that decay. Encoding quality matters more than sheer repetition, as connecting new information to existing knowledge creates stronger, more retrievable memories. Context also plays a key role: recall improves when the environment and cues at retrieval match those present during the original learning.

0 comments Read more at DEV Community