Engineering Decisions Behind Windows 3.1 Setup
Raymond Chen, a seasoned engineer at Microsoft, recently shed light on a long-standing question regarding the company’s choice to utilize a miniature version of Windows 3.1 for the installation of Windows 95, rather than opting for a compact version of Windows 95 itself. The rationale behind this decision is rooted in a blend of engineering practicality, user experience considerations, and the constraints of the time.
As we approach the 30th anniversary of Windows 95, Chen elaborated on the intricacies of the operating system’s setup process. When users initiated the installation from MS-DOS, a streamlined version of Windows 3.1 was deployed. This 16-bit application facilitated the copying of files and the migration of settings, paving the way for the full-fledged Windows 95 experience.
If users began their journey with Windows 3.1, they would directly engage with the 16-bit application. The question arises: why not leap directly into a miniaturized Windows 95 to expedite the setup process within a fully 32-bit environment?
Chen acknowledged the possibility but pointed out several challenges from both engineering and user experience perspectives. From an engineering standpoint, the path of utilizing Windows 3.1 was the more straightforward option; Microsoft had a ready-to-go miniature version of this operating system. In contrast, Windows 95 was plagued by delays, making the development of a variant solely for setup purposes a daunting task that would likely face significant resistance from the engineering team.
While some resourceful individuals in later years have managed to compress Windows 95 to a mere 5 MB, this still represents a larger footprint compared to the leaner Windows 3.1 that Microsoft ultimately employed.
The quest to minimize operating system size continues today, as evidenced by projects like Nano11, which aim to eliminate unnecessary bloat from Microsoft’s current flagship offerings.
Another critical factor in the decision-making process was user experience. Even if Microsoft had successfully created a smaller version of Windows 95, the installation would still necessitate at least two floppy disks, whereas Windows 3.1 required only one. Additionally, the installation process would involve two reboots: the first to boot into the mini-Windows 95 and a second to transition into the full Windows 95 environment.
Chen humorously noted that this approach violated the “only one reboot” principle, a sentiment that resonates with anyone who has experienced the frustrations of multiple restarts during updates today. However, in those earlier days, Microsoft prioritized user experience over the relentless push of new technologies.
Ultimately, the decision to leverage Windows 3.1 for the installation process was a calculated one. It minimized engineering costs and streamlined the user experience by reducing the number of necessary reboots. As Chen reflects on this 30-year-old decision, it serves as a reminder of a time when user satisfaction was at the forefront of Microsoft’s design philosophy.
