Forgot your password?
Cyhist Feb 11 1997 F
Document Actions
Date: Tue, 11 Feb 1997 17:02:01 -0500
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
From: "David S. Bennahum" <davidsol@panix.com>
Subject: CM> Software complexity, microprocessors.
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
One effect the microprocessor has had on software, as we've discussed, was to increase by an amazing amount the size and complexity of what could be stored and manipulated in memory. The 4004 processor was famously billed as containing the same computing capacity as ENIAC. I am not sure how many ENIACS would be needed to match the latest Pentium processor. But that aside, one effect is that today the number of people who can author a single piece of commercial software, like a word processor, or an O/,S, is bascially zero. I doubt any of the top ten selling consumer software items have a single person who can say they coded the program, or even how the program works.
So, given this state of things, it became clear that, sometime ago, maybe in the 80s, the last human actually wrote an entire O/S. Does anyone know the last time an O/S, for say a "major" platform (digital watches don't count), was written by essentially one person? It's a tricky question, because software, even 30 years ago, was written in teams, but the point is, back then I think you could read the code, and figure out how things fit together. Today a print-out of Windows 95 would be beyond any one person's ability to comprehend. Anyway, any thoughts on when the system "flipped" to becoming too huge for anyone to read?
That programs are now so large, and authored by so many people, has changed the architecture of software. When people talk of programs that program themselves, or teaching computers to teach themselves, it seems partly a reflection that, with the relentless velocity of Moore's Law, this is, de facto, the ultimate solution, since people can no longer keep pace, and write software that takes advantage of the microprocessor's remarkable capacity.
It is chilling to think that the number of programmers, in the way A. Padgett Peterson described it in his previous post, has only doubled, while the amount of software has increased far more than that, in the past three decades since the invention of the microprocessor.
db
______________________________________________________________________
Reply-To: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@SJUVM.STJOHNS.EDU>
From: "David S. Bennahum" <davidsol@panix.com>
Subject: CM> Software complexity, microprocessors.
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
One effect the microprocessor has had on software, as we've discussed, was to increase by an amazing amount the size and complexity of what could be stored and manipulated in memory. The 4004 processor was famously billed as containing the same computing capacity as ENIAC. I am not sure how many ENIACS would be needed to match the latest Pentium processor. But that aside, one effect is that today the number of people who can author a single piece of commercial software, like a word processor, or an O/,S, is bascially zero. I doubt any of the top ten selling consumer software items have a single person who can say they coded the program, or even how the program works.
So, given this state of things, it became clear that, sometime ago, maybe in the 80s, the last human actually wrote an entire O/S. Does anyone know the last time an O/S, for say a "major" platform (digital watches don't count), was written by essentially one person? It's a tricky question, because software, even 30 years ago, was written in teams, but the point is, back then I think you could read the code, and figure out how things fit together. Today a print-out of Windows 95 would be beyond any one person's ability to comprehend. Anyway, any thoughts on when the system "flipped" to becoming too huge for anyone to read?
That programs are now so large, and authored by so many people, has changed the architecture of software. When people talk of programs that program themselves, or teaching computers to teach themselves, it seems partly a reflection that, with the relentless velocity of Moore's Law, this is, de facto, the ultimate solution, since people can no longer keep pace, and write software that takes advantage of the microprocessor's remarkable capacity.
It is chilling to think that the number of programmers, in the way A. Padgett Peterson described it in his previous post, has only doubled, while the amount of software has increased far more than that, in the past three decades since the invention of the microprocessor.
db
______________________________________________________________________
