I remember Scott McNealy saying he (and presumably his team) underestimated the impact and importance of 32 bit computing and concentrated on 64 bit. That'd be a strategic technical oversight. Maybe they wanted people to go to thin-clients. I suppose the key when you're that big is to work out when you do or can shape trends, and when you can and must follow them. Not sure Jonathan Schwartz corrected course, or by then too late.
That's not the reason I stopped using Suns. They annoyed me by switching to an aesthetically ugly unix. They priced their machines for their finance industry customers, then Intel + open source unix boxes caught up in performance for 1/4 the price.
Sun was huge on thin clients around the turn of the century, the last great battle between thin and thick clients. In the first dotcom bubble, Sun sold a zillion Enterprise-series (E400/E4000/etc. big boxes) servers and wanted to keep that flow going. The market evidently did not agree with their plans.
The perspective that is lost to time is that back then Sun was kind of like AWS nowadays WRT mindshare, with the addition of hardware business, which is key to my point: likely when McNealy said that they underestimated 32bit he was eliding the problematic future they saw in the UltraSparc 3, that it was an engineering dead-end (which begat the T series), and, well...capitalists profit from externalizing their liabilities. "We're still awesome, it was 32bit's fault."
Comments
I remember Scott McNealy saying he (and presumably his team) underestimated the impact and importance of 32 bit computing and concentrated on 64 bit. That'd be a strategic technical oversight. Maybe they wanted people to go to thin-clients. I suppose the key when you're that big is to work out when you do or can shape trends, and when you can and must follow them. Not sure Jonathan Schwartz corrected course, or by then too late.
That's not the reason I stopped using Suns. They annoyed me by switching to an aesthetically ugly unix. They priced their machines for their finance industry customers, then Intel + open source unix boxes caught up in performance for 1/4 the price.
Sun was huge on thin clients around the turn of the century, the last great battle between thin and thick clients. In the first dotcom bubble, Sun sold a zillion Enterprise-series (E400/E4000/etc. big boxes) servers and wanted to keep that flow going. The market evidently did not agree with their plans.
The perspective that is lost to time is that back then Sun was kind of like AWS nowadays WRT mindshare, with the addition of hardware business, which is key to my point: likely when McNealy said that they underestimated 32bit he was eliding the problematic future they saw in the UltraSparc 3, that it was an engineering dead-end (which begat the T series), and, well...capitalists profit from externalizing their liabilities. "We're still awesome, it was 32bit's fault."