Forgot your password?
Cyhist Apr. 1 1998 A
Document Actions
========================================================================= Date: Wed, 1 Apr 1998 00:11:04 -0500
Reply-To: dsenzig@execpc.com
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: Don Senzig <dsenzig@EXECPC.COM>
Subject: Re: An anonimous quotation
In-Reply-To: <980331112343.20805d6a@hobbes.orl.lmco.com> MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
On some of Padgett's reasoning...
On 31 Mar 98 at 11:23, A. Padgett Peterson P.E. Info wrote in part:
>Suspect is a hoax since there are several anomalies:
>
>>64kb of core is more memory than any competent programmer ever needs to do anything! We will design the "processor of the future (8088)" to
>
>1) Core memory was never used in a production PC (braces), SRAM & DRAM were common by the late 1970's. My ZX-81 still has one of the first
>6116s but 1108s were available considerably earlier.
My recollection of common usage of "core" was that many folks still used it to refer to the main memory even though solid state memory was the normally used type. SRAM and DRAM were certainly the memory of choice for personal computers from the Altair on. (or the Scelbi, the Mark 8 ...)
Consider that under Linux (and I assume Unix) you get a "core" dump file when an application crashes.
...trim...
>>address memory in several segments of 64kb so as to make the simultaneous loading of tasks easier for the OS. In addition, 10 tasks loaded simultaneously in memory is ridiculous, so the system architecture can be safely limited to 640kb (an absurdly large number) without fear of any user ever exceeding the system requirements!
>
>3) Again from memory so no guarentees but wasn't the first design a 500K/
>500K split and the "640k barrier" imposed by the video buffer at segment A000h (and the Lotus 1-2-3 habit of writing directly to it).
I remember a presentation by an Intel sales engineer about the time that the 808x processors were released who made a point of the segmentation being useful to multi-tasking as you could have several 64K segments, each for a seperate task. I believe that this was part of the sales pitch on the 808x pair.
I don't remember a 640K limit being part of that discussion. The talk did predate the PC.
But I do recall that the 640K limit being attributed to Microsoft (I believe that the attribution was to Bill himself) as that this made the PC 10 times the computer that a Z80 based one was.
...trim...
>>For
>>developers, the addition of a 10MB hard-disc will provide effectively infinite storage.
>
>Definately later, was not until BIOS revision of October 1982 added "Bios Extensions" which made hard disks supportble. 10 Mb was just the disk by the PC-XT but DEC and DG people were already used to thinking in terms of much larger disks.
But I don't believe that PC types were, to folks outside of the larger minis 10Mb seemed almost infinite. The PDP-11 (LSI-11) I worked with just before the PC came out had 3 RK-05s, each being 2.5Mb, it seemed like a pretty big space to me and the 3 other programmers that I worked with. The 14Mb disk on the Altos (Z80 muti-user, MP/M based) seemed pretty big too.
>Sounds like another "urban fable" to me.
I'm not so willing to dismiss it. Certainly could be but I think that there is a kernel of truth to it.
Don Senzig, Jr - dsenzig@execpc.com - http://www.execpc.com/~dsenzig
Is your biological clock year 2000 compliant?
______________________________________________________________________
Reply-To: dsenzig@execpc.com
Sender: "CYHIST Community Memory: Discussion list on the History of
Cyberspace" <CYHIST@MAELSTROM.STJOHNS.EDU> From: Don Senzig <dsenzig@EXECPC.COM>
Subject: Re: An anonimous quotation
In-Reply-To: <980331112343.20805d6a@hobbes.orl.lmco.com> MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT
______________________________________________________________________
Community Memory: Discussion List on the History of Cyberspace ______________________________________________________________________
On some of Padgett's reasoning...
On 31 Mar 98 at 11:23, A. Padgett Peterson P.E. Info wrote in part:
>Suspect is a hoax since there are several anomalies:
>
>>64kb of core is more memory than any competent programmer ever needs to do anything! We will design the "processor of the future (8088)" to
>
>1) Core memory was never used in a production PC (braces), SRAM & DRAM were common by the late 1970's. My ZX-81 still has one of the first
>6116s but 1108s were available considerably earlier.
My recollection of common usage of "core" was that many folks still used it to refer to the main memory even though solid state memory was the normally used type. SRAM and DRAM were certainly the memory of choice for personal computers from the Altair on. (or the Scelbi, the Mark 8 ...)
Consider that under Linux (and I assume Unix) you get a "core" dump file when an application crashes.
...trim...
>>address memory in several segments of 64kb so as to make the simultaneous loading of tasks easier for the OS. In addition, 10 tasks loaded simultaneously in memory is ridiculous, so the system architecture can be safely limited to 640kb (an absurdly large number) without fear of any user ever exceeding the system requirements!
>
>3) Again from memory so no guarentees but wasn't the first design a 500K/
>500K split and the "640k barrier" imposed by the video buffer at segment A000h (and the Lotus 1-2-3 habit of writing directly to it).
I remember a presentation by an Intel sales engineer about the time that the 808x processors were released who made a point of the segmentation being useful to multi-tasking as you could have several 64K segments, each for a seperate task. I believe that this was part of the sales pitch on the 808x pair.
I don't remember a 640K limit being part of that discussion. The talk did predate the PC.
But I do recall that the 640K limit being attributed to Microsoft (I believe that the attribution was to Bill himself) as that this made the PC 10 times the computer that a Z80 based one was.
...trim...
>>For
>>developers, the addition of a 10MB hard-disc will provide effectively infinite storage.
>
>Definately later, was not until BIOS revision of October 1982 added "Bios Extensions" which made hard disks supportble. 10 Mb was just the disk by the PC-XT but DEC and DG people were already used to thinking in terms of much larger disks.
But I don't believe that PC types were, to folks outside of the larger minis 10Mb seemed almost infinite. The PDP-11 (LSI-11) I worked with just before the PC came out had 3 RK-05s, each being 2.5Mb, it seemed like a pretty big space to me and the 3 other programmers that I worked with. The 14Mb disk on the Altos (Z80 muti-user, MP/M based) seemed pretty big too.
>Sounds like another "urban fable" to me.
I'm not so willing to dismiss it. Certainly could be but I think that there is a kernel of truth to it.
Don Senzig, Jr - dsenzig@execpc.com - http://www.execpc.com/~dsenzig
Is your biological clock year 2000 compliant?
______________________________________________________________________
