Static

Things That Turbo Pascal Is Smaller Than (2011)

First reported by Prog21.dadgum ·

The signal ●○○○ Compiled by AI from Prog21.dadgum and Hacker News
Why you might care

If you are developing for resource-constrained environments, you can now leverage decades-old techniques for extreme optimization.

What happened

In 2011, James Hague highlighted the astonishingly small size of Turbo Pascal 3.02 for MS-DOS, released in September 1986. The entire integrated development environment and compiler package measured just 39,731 bytes. Hague compared this to various contemporary digital artifacts, demonstrating how much software bloat had occurred in 25 years. For instance, the minified jQuery 1.6 library was over twice the size at 90,518 bytes, the Yahoo.com homepage exceeded 219,000 bytes, and even a single image of the iPhone 4S on Apple's website was larger at 190,157 bytes. Other comparisons included the zlib.h header file from the Mac OS X Lion SDK (80,504 bytes), the OS X Lion 'touch' command (44,016 bytes), and the Erlang R14B02 parser's compiled code (286,324 bytes). This comparison underscored the dramatic increase in file sizes for software and web content in the intervening decades.

What it means

The stark contrast between Turbo Pascal's compact size and the growing file sizes of 2011 web pages and libraries illustrates a significant trend in software development: the tradeoff between functionality and size. Modern applications and web content often prioritize feature richness and user experience, leading to increased resource consumption. This phenomenon suggests a market where ease of development and broad accessibility often outweigh the need for minimal footprint, a shift driven by generally abundant hardware resources and network bandwidth.

This comparison serves as a historical marker, highlighting how far computing has come in terms of both capability and resource demands. For developers, it raises questions about current software bloat and the potential for rediscovered optimization techniques. It also implies that the lessons learned from creating highly efficient software in the past might offer valuable insights for future development, particularly in areas where performance and size are critical constraints, such as embedded systems or low-bandwidth applications.

AI-written summary. May contain errors.

Things