Rise of the Wizards By Erik ``De Zeurkous'' Bogeholt, 2009. For many years I wondered what went wrong. Programs became bloated, operating systems that ran reliably for years stopped keeping up with the times, stalling, being kept adrift by a substance almost pure of hype. Of course, this didn't happen in the stroke of one night and one day; nor are these problems unknown elsewhere in our culture. I will attempt to describe what happened. 1. The end of the Real Programming era: This is partially described by Eric S. Raymond, hencefort `esr', in [1]. Let us suffice with the assumption that the Real Programming era ended around the time of the first hint of the pervasiveness of computer networks; this may not be very far from the truth. Around this point, the hackers, with their often superior grasp of high-level issues, succeeded the Real Programmers as the most prominent culture. Apart from a few prominent examples, the Real Programmers almost died out as a culture. They were replaced by the hackers. Thus ended the era of Real Programming. 2. The start of the Middle Area: This event is partially described by esr in [1] as well. Several UNIX sites connected to both the ARPANET and UUCP, in the process making the first pervasive link between the hacker and UNIX cultures; and both began to merge. 3. The First Period of the Middle Area: The First Period of the Middle Area of Computing was ultimately centered around the Berkeley Software Distribution of UNIX, henceforth `BSD', when the latter gained significant networking capability, and, more importantly, significant ability to communicate with the ARPANET. Slowly, the migration of hackers to UNIX and the migration of UNIX to the ARPANET began to take place. 4. The Second Period of the Middle Area: As the First UNIX Wars started to wreak heavoc, the UNIX and hacker/ARPANET cultures were about to be almost fully merged. 5. The Third Period of the Middle Area: As the Second phase of the UNIX Wars raged, the integration between the ARPANET and UNIX cultures was almost complete. Several pockets continued using ARPANET software on what was slowly transitioning into the Internet, as we now know it. The microcomputer hackers, a largely independent group of computer enthusiasts, started to gain an influence. 5. The Fourth Period of the Middle Area: As the UNIX Wars died out, a minor but significant number of disenchanted hackers left to colonize Windows NT. The remaining hackers split up in several camps. The Internet, now fully-formed, was beginning to get commercialized. 6. The Fifth Period of the Middle Area: The camps, each harboring their own version of Free UNIX, had started taking the shape of factions, and some infighting did occour. A cacophony of hackers and other people started once more leaving Windows for UNIX. This had a lasting effect on the level of bloat in the latter system. The infighting worsened, and with interference from the bloat, progress slowly stopped. The current status: At this moment, I believe the hacker methodology is no longer useful, just like the Real Programmer methodology is no longer useful in most cases. Just like when the Real Programmers disappeared due to their methodology no longer being technically economical (effort vs. performance) in most situations, the hacker methodology is similarly too bloated and too compact for the systems in use today, or at least our expectations of them. I firmly believe it is time to re-evaluate our strategy. While I cannot pinpoint the exact source and coinage, I propose `wizard' as the name for the type of person and methodology mostly endemic in UNIX nature. With consideration of our Real Programmer and Hacker heritage, which still teaches us things day by day, I believe we should move on. I do not believe in Worse is Better. In as far as that means `don't be anal about consistency', I can agree. In as far as it means `don't put consistency over ease of implementation', I can not. Many jobs seem difficult at first, but when one lets the mind flow freely, they will often be eclipsed by a somewhat convenient soution. That solution can be consistent; as long as one keeps that latter priority in mind. Completeness follows a similar path and story. When I say it is time for a `wizard' culture, I chiefly point at the ability to think and act high-level, without sacrificing consistency with the low-level parts. C and the original UNIX followed that path. They were carefully designed to be high-level, yet transparent. That is the path we should follow. That is our UNIX heritage. Hence, I propose to enter the following era: 7. The Sixth, and Last, period of the Middle Area. The hacker heritage we will keep, but it's heydays are over. We will write programs to be universal, without regard for technological economisms that do no longer hold true. Yet we will make and keep our tasks and programs well-defined; we should be, and make UNIX to be, an `universal pipe fitting' -- a great way to connect data and thoughts; yet with a personality, as we all have fairly unique personalities of our own. This will be the Rise of the Wizards. We will also expand into hardware, opening up the last vestiges of the proprietary market -- video, nontrivial networking -- and continue the current work to open them up with renewed vigor. We shall then start making computers physically like pipelines, like our software will be. Only that way we will step once more into the direction of a universal computer. After our accomplishments in that area has been made, we can ourselves retreat as Wizards. Thus we will end the Middle Area of Computing and evolve into a new stage of technology. What lies beyond, what culture we will evolve into, we will one day see. Into the future! --De Zeurkous.