Joining Sun Microsystems – 40 years ago (2022)

Original Article ↗ Hacker News Discussion ↗

Early Unix & Workstation Ecosystem

  • Commenters recall the early 1980s as a crowded field of small Unix-on-68k hardware vendors, many using similar recipes (680x0, Ethernet, Unisoft Unix).
  • By the mid–late 1990s, the landscape had consolidated to a handful of major Unix + proprietary-CPU stacks (Solaris/SPARC, AIX/POWER, HP-UX/PA-RISC, IRIX/MIPS, Tru64/Alpha).
  • Several people correct timeline conflations, emphasizing how different 1982 was from the 1990s in terms of chips, networking, and products.

Sun’s Rise, Unix’s Fate, and Linux

  • Sun’s rapid arc is admired: rough 68000 hardware in 1982, then Sun-2/3, SPARC, NFS/RPC, shared libraries, networked Unix, shaping the modern datacenter.
  • Many see Linux on commodity x86 as having cloned Sun’s environment at far lower cost, erasing Sun’s hardware advantage and turning Unix into background infrastructure.
  • Some argue Unix “won” by becoming a free, assumed standard (Linux/BSD), making commercial Unix a hard business; others stress that specialized commercial Unix still survives in niches.

ARM, SPARC, and Alternate Histories

  • Debate on whether early ARM or StrongARM were ever realistic Sun/SPARC substitutes. One side calls ARM an also-ran until much later; others point to embedded and WinCE success.
  • Speculation about alternate paths (Sun acquiring Acorn or Commodore, deals with Google, different strategy on OpenSolaris) is countered as largely revisionist or non-viable.

Sun’s Decline, Strategy, and Oracle

  • Multiple commenters blame strategic errors: unclear hardware vs software identity, failure to leverage software to sell hardware, late/weak x86 strategy, and missed opportunities around Java and Active Directory–like services.
  • Others cite specific missteps (cancelling Solaris x86 temporarily, not closing a Google/Solaris deal, closing professional services, monetization/lock-in attempts).
  • Oracle’s acquisition is viewed as a fire sale; opinions differ on whether IBM would have been a better buyer. Some argue Oracle now mainly milks legacy products, sacrificing mindshare.

Talent, Nepotism, and Career Eras

  • The article’s family connections spark a nuanced nepotism debate: some lament loss of extended-family “ladders,” others highlight fairness, corruption risks, and the role of friends/pro networks instead of relatives.
  • Commenters reflect on how rare and “gift-like” it felt to work at Sun, SGI, or similar companies in their heyday, contrasting that with today’s more commoditized, money-first industry.
  • Several note how early-career deep OS work (like porting Unix) is harder to replicate today, with higher abstraction layers and different opportunities, though not necessarily worse—just different.