The date on Mitch and Ray's spec is December 1984 and we designed most of the core parts of Notes soon after that in 1985.
Sorry to hear that many of you had to deal with our 1985 design choices all through the subsequent years. But I think many of the commenters here are maybe forgetting or not aware of what personal computing and networking were like 32 years ago: Windows 1.0 (beta), EGA graphics (640×350 w/16 colors), 640KB main memory (kilobytes not megabytes), Ethernet/TokenRing LANs, dial-up 9600 baud modems (kilobits/sec not megabits/sec or gigabits/sec). https://en.wikipedia.org/wiki/Windows_1.0
The Internet was just being formed out of the Arpanet and other research networks. Mail services and protocols like SMTP and POP3 didn't exist.
We worked with Ron Rivest to define BSAFE and build the first real commmercial product with public key crypto.
As far as the Notes database (again circa 1984): it seemed obvious to us, especially Ray, that relational was not the best model, but we took endless grief for not using a relational database.
In hindsight, we obviously could have made quite a few better decisions and evolved the design and code more quickly, but as one commenter pointed out, we were trying to maintain app and data compatibility across multiple client and server software and hardware environments that were continually evolving around us.
Sorry to hear that many of you had to deal with our 1985 design choices all through the subsequent years. But I think many of the commenters here are maybe forgetting or not aware of what personal computing and networking were like 32 years ago: Windows 1.0 (beta), EGA graphics (640×350 w/16 colors), 640KB main memory (kilobytes not megabytes), Ethernet/TokenRing LANs, dial-up 9600 baud modems (kilobits/sec not megabits/sec or gigabits/sec). https://en.wikipedia.org/wiki/Windows_1.0
The Internet was just being formed out of the Arpanet and other research networks. Mail services and protocols like SMTP and POP3 didn't exist.
We worked with Ron Rivest to define BSAFE and build the first real commmercial product with public key crypto.
As far as the Notes database (again circa 1984): it seemed obvious to us, especially Ray, that relational was not the best model, but we took endless grief for not using a relational database.
In hindsight, we obviously could have made quite a few better decisions and evolved the design and code more quickly, but as one commenter pointed out, we were trying to maintain app and data compatibility across multiple client and server software and hardware environments that were continually evolving around us.