[varunnrao]

I think what ultimately led to Sun's downfall is a combination of what ESR [1] and joelonsoftware [2] have previously covered.

1. Sun didn't become the defacto desktop platform because they lost out to WinNT. So they lost out on the consumer market. 2. Custom server hardware and software makers like Sun and Silicon Graphics were the fashion till Google and later on Facebook came around and built their own data centers with consumer hardware and specialized software to overcome the inherent unreliability of that hardware. And anyway ever since web-based software became a thing your device is practically a console a la Chromebooks. So they lost the server market.

The only option left was to serve the high end HPC market like labs or even banks but that didn't make business sense since that's increasingly niche because those customers would eventually also want the effects of commoditization.

[1] - http://esr.ibiblio.org/?p=6279 [2] - https://www.joelonsoftware.com/2002/08/30/platforms/

[cryptonector]

The real losses were against Windows 2000 (specifically Active Directory) and to Linux.

The loss to Linux was greatly accelerated by Sun's failure to make a deal with Google for Google to use Solaris on their servers. The story I heard was that Scott wanted a server count for the license while Google believed server count was a top secret datum.

If Sun had made a deal with Google in 2002 and worked on OpenSolaris starting in 2001, then Linux might not have been quite the success it became.

[rbanffy]

It wasn’t Google’s investment that made Linux a viable OS for enterprise applications. Google using Solaris would have made little difference.

Active Directory was a huge win for Microsoft. We’ll see them milk that product for generations. Sun could have captured a part of that, but it’d need to compete against Microsoft when 99.9% of the clients using AD were Microsoft. I doubt they would succeed.

Another fun alt-history branch is the one Sun manages to sell thousands of Amigas as low-end Unix workstations, moving Unix down into the personal computer space, and saving Commodore.

Sadly, none of that happened and we live in the crappiest timeline.

[dboreham]

They didn't lose to NT. The loss in the consumer desktop market occurred in the DOS era.

[rbanffy]

NT ate the technical workstation space from below. Once NT was good enough on commodity hardware, they were toast.

Unless they went the Apple route and made “luxury workstations” average people would buy. Hindsight is always 20-20, so we now see all the things they could have done then to prevent now from happening.