| Holy wars of the past - how did they turn out? [message #405149] |
Thu, 04 February 2021 12:15  |
|
|
Originally posted by: Thomas Koenig
The proverbial computer holy war is probably big vs. little endian,
which has been pretty much decided in favor of little endian
by default or by Intel, although TCP/IP is big-endian.
Machine language vs. those CPU-time-wasting assemblers - made
obsolescent by high-level programming languages.
High-level vs. assembler: Hardly anybody does assembler any more.
Structured programming vs. goto - structured programming won.
RISC vs. CISC: The really complex CISC-architectures died out.
The difference is now less important with superscalar architectures.
VMS vs. Unix - decided by DEC's fate, and by Linux.
DECNET vs. TCP/IP: See above.
Emacs vs. vi: vim has led to a resurgence of vi, and many people
are using this even on Windows.
Everybody vs. Fortran: Hating FORTRAN become the very definition
of a computer scientist. They didn't notice that, since 1991,
it has become quite a modern programming language.
Others?
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405151 is a reply to message #405149] |
Thu, 04 February 2021 12:44  |
|
|
Originally posted by: Jim Jackson
On 2021-02-04, Thomas Koenig <tkoenig@netcologne.de> wrote:
> The proverbial computer holy war is probably big vs. little endian,
> which has been pretty much decided in favor of little endian
> by default or by Intel, although TCP/IP is big-endian.
>
> Machine language vs. those CPU-time-wasting assemblers - made
> obsolescent by high-level programming languages.
>
> High-level vs. assembler: Hardly anybody does assembler any more.
>
> Structured programming vs. goto - structured programming won.
>
> RISC vs. CISC: The really complex CISC-architectures died out.
> The difference is now less important with superscalar architectures.
>
> VMS vs. Unix - decided by DEC's fate, and by Linux.
>
> DECNET vs. TCP/IP: See above.
>
> Emacs vs. vi: vim has led to a resurgence of vi, and many people
> are using this even on Windows.
>
> Everybody vs. Fortran: Hating FORTRAN become the very definition
> of a computer scientist. They didn't notice that, since 1991,
> it has become quite a modern programming language.
>
Ethernet v. IBM Token Ring - did you know that ethernet could never work
as it would collapse under moderately heavy loads?
|
|
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405154 is a reply to message #405153] |
Thu, 04 February 2021 13:47 |
|
|
Originally posted by: Jim Jackson
On 2021-02-04, Ahem A Rivet's Shot <steveo@eircom.net> wrote:
> On Thu, 4 Feb 2021 17:44:34 -0000 (UTC)
> Jim Jackson <jj@franjam.org.uk> wrote:
>
>> Ethernet v. IBM Token Ring - did you know that ethernet could never work
>> as it would collapse under moderately heavy loads?
>
> Ethernet over thicknet or co-ax did.
>
Well not in my experience, which was extensive. The vampire taps were a
B*, as were fitting the N-type plugs, but thin/cheaper net sorted that.
Of cpourse as with any network sizing and planning and knowing what you
doing always helped.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405155 is a reply to message #405149] |
Thu, 04 February 2021 14:07 |
Anssi Saari
Messages: 327
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Thomas Koenig <tkoenig@netcologne.de> writes:
> RISC vs. CISC: The really complex CISC-architectures died out.
> The difference is now less important with superscalar architectures.
Might this be coming back with Apple's M1 showing how AArch64 can run
circles around x86-64 at lower power? Might get interesting if something
like that starts eating Intel's Xeon business. Or all x86 business.
> Emacs vs. vi: vim has led to a resurgence of vi, and many people
> are using this even on Windows.
Has this been decided then? What about Emacs in Windows? Actually I
probably would've abandoned Emacs if it weren't for the VHDL and ORG
modes.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405156 is a reply to message #405155] |
Thu, 04 February 2021 14:19 |
scott
Messages: 4237
Registered: February 2012
Karma: 0
|
Senior Member |
|
|
Anssi Saari <as@sci.fi> writes:
> Thomas Koenig <tkoenig@netcologne.de> writes:
>
>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>> are using this even on Windows.
>
> Has this been decided then? What about Emacs in Windows? Actually I
> probably would've abandoned Emacs if it weren't for the VHDL and ORG
> modes.
It's still around 50-50 here. The Verilog
types tend to use emacs (and csh), and the software folks vim (and bash/ksh).
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405158 is a reply to message #405151] |
Thu, 04 February 2021 15:34 |
|
|
Originally posted by: J. Clarke
On Thu, 4 Feb 2021 17:44:34 -0000 (UTC), Jim Jackson
<jj@franjam.org.uk> wrote:
> On 2021-02-04, Thomas Koenig <tkoenig@netcologne.de> wrote:
>> The proverbial computer holy war is probably big vs. little endian,
>> which has been pretty much decided in favor of little endian
>> by default or by Intel, although TCP/IP is big-endian.
>>
>> Machine language vs. those CPU-time-wasting assemblers - made
>> obsolescent by high-level programming languages.
>>
>> High-level vs. assembler: Hardly anybody does assembler any more.
>>
>> Structured programming vs. goto - structured programming won.
>>
>> RISC vs. CISC: The really complex CISC-architectures died out.
What do you consider to be a "really complex CISC-architecture"?
>> The difference is now less important with superscalar architectures.
>>
>> VMS vs. Unix - decided by DEC's fate, and by Linux.
>>
>> DECNET vs. TCP/IP: See above.
>>
>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>> are using this even on Windows.
>>
>> Everybody vs. Fortran: Hating FORTRAN become the very definition
>> of a computer scientist. They didn't notice that, since 1991,
>> it has become quite a modern programming language.
>>
>
> Ethernet v. IBM Token Ring - did you know that ethernet could never work
> as it would collapse under moderately heavy loads?
|
|
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405162 is a reply to message #405154] |
Thu, 04 February 2021 16:49 |
|
|
Originally posted by: gareth evans
On 04/02/2021 18:47, Jim Jackson wrote:
> On 2021-02-04, Ahem A Rivet's Shot <steveo@eircom.net> wrote:
>> On Thu, 4 Feb 2021 17:44:34 -0000 (UTC)
>> Jim Jackson <jj@franjam.org.uk> wrote:
>>
>>> Ethernet v. IBM Token Ring - did you know that ethernet could never work
>>> as it would collapse under moderately heavy loads?
>>
>> Ethernet over thicknet or co-ax did.
>>
>
> Well not in my experience, which was extensive. The vampire taps were a
> B*, as were fitting the N-type plugs, but thin/cheaper net sorted that.
The N plugs were named by a mathematician because there are N different
ways of assembling them, and all are wrong!
Actually N after the designer Neill, just as the C type were named
after Concelman, and they joined forces to create the BNC,
Bayonet-Neill-Concelman
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405163 is a reply to message #405149] |
Thu, 04 February 2021 16:50 |
Freddy1X
Messages: 61
Registered: August 2012
Karma: 0
|
Member |
|
|
Thomas Koenig wrote:
> The proverbial computer holy war is probably big vs. little endian,
> which has been pretty much decided in favor of little endian
> by default or by Intel, although TCP/IP is big-endian.
>
> Machine language vs. those CPU-time-wasting assemblers - made
> obsolescent by high-level programming languages.
>
> High-level vs. assembler: Hardly anybody does assembler any more.
>
> Structured programming vs. goto - structured programming won.
>
> RISC vs. CISC: The really complex CISC-architectures died out.
> The difference is now less important with superscalar architectures.
>
> VMS vs. Unix - decided by DEC's fate, and by Linux.
>
> DECNET vs. TCP/IP: See above.
>
> Emacs vs. vi: vim has led to a resurgence of vi, and many people
> are using this even on Windows.
>
> Everybody vs. Fortran: Hating FORTRAN become the very definition
> of a computer scientist. They didn't notice that, since 1991,
> it has become quite a modern programming language.
>
> Others?
PDA Vs. smartphone.
There is mainframe Vs. PC/distributed Vs. laptop Vs. pocket thingy.
Freddy,
pick three of tha above,
--
Not for sale to minors.
/|>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\|
/| I may be demented \|
/| but I'm not crazy! \|
/|<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\|
* SPAyM trap: there is no X in my address *
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405164 is a reply to message #405155] |
Thu, 04 February 2021 16:51 |
|
|
Originally posted by: gareth evans
On 04/02/2021 19:07, Anssi Saari wrote:
> Thomas Koenig <tkoenig@netcologne.de> writes:
>
>> RISC vs. CISC: The really complex CISC-architectures died out.
>> The difference is now less important with superscalar architectures.
>
> Might this be coming back with Apple's M1 showing how AArch64 can run
> circles around x86-64 at lower power?
I can't get my head round assembler on the ARM architecture when there
are no operations on memory locations, with all data firstly having
to be moved into registers.
Too much exposure to PDP11 and X86, I suppose.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405165 is a reply to message #405149] |
Thu, 04 February 2021 17:14 |
John Levine
Messages: 1405
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
In article <rvha47$pap$1@newsreader4.netcologne.de> you write:
> Machine language vs. those CPU-time-wasting assemblers - made
> obsolescent by high-level programming languages.
Actually, made obsolescent by random access main memory. As we recall
from the Story of Mel, if your main memory is a drum, your program
runs a lot faster if it's laid out to minimize rotational delay. And
also if you know where each instruction is, you can save a few words
by using the instructions as constants. That lived on as long as the
PDP-11 where the Unix C compiler had a peephole optimization to turn
OP #N,N(R)
into
OP (PC),N(R)
to use the same instruction word as both the constant N and the index offset N.
> High-level vs. assembler: Hardly anybody does assembler any more.
FORTRAN did that, but producing shockingly good code. It did stuff
that's still pretty advanced, like nested loop merging.
> RISC vs. CISC: The really complex CISC-architectures died out.
> The difference is now less important with superscalar architectures.
Technology did that. CISC made sense when RAM was expensive, microcode
ROM was faster than RAM, and caches were too expensive for anything but
high end mainframes. That was then.
--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly
|
|
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405170 is a reply to message #405163] |
Thu, 04 February 2021 18:39 |
|
|
Originally posted by: J. Clarke
On Thu, 04 Feb 2021 16:50:05 -0500, Freddy1X <freddy1X@indyX.netX>
wrote:
> Thomas Koenig wrote:
>
>> The proverbial computer holy war is probably big vs. little endian,
>> which has been pretty much decided in favor of little endian
>> by default or by Intel, although TCP/IP is big-endian.
>>
>> Machine language vs. those CPU-time-wasting assemblers - made
>> obsolescent by high-level programming languages.
>>
>> High-level vs. assembler: Hardly anybody does assembler any more.
>>
>> Structured programming vs. goto - structured programming won.
>>
>> RISC vs. CISC: The really complex CISC-architectures died out.
>> The difference is now less important with superscalar architectures.
>>
>> VMS vs. Unix - decided by DEC's fate, and by Linux.
>>
>> DECNET vs. TCP/IP: See above.
>>
>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>> are using this even on Windows.
>>
>> Everybody vs. Fortran: Hating FORTRAN become the very definition
>> of a computer scientist. They didn't notice that, since 1991,
>> it has become quite a modern programming language.
>>
>> Others?
>
> PDA Vs. smartphone.
I'm not sure that was ever a "holy war". Smartphones were a natural
evolution of PDAs.
> There is mainframe Vs. PC/distributed Vs. laptop Vs. pocket thingy.
To some extent that's still going.
> Freddy,
> pick three of tha above,
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405171 is a reply to message #405165] |
Thu, 04 February 2021 18:40 |
|
|
Originally posted by: J. Clarke
On Thu, 4 Feb 2021 22:14:45 -0000 (UTC), John Levine <johnl@taugh.com>
wrote:
> In article <rvha47$pap$1@newsreader4.netcologne.de> you write:
>> Machine language vs. those CPU-time-wasting assemblers - made
>> obsolescent by high-level programming languages.
>
> Actually, made obsolescent by random access main memory. As we recall
> from the Story of Mel, if your main memory is a drum, your program
> runs a lot faster if it's laid out to minimize rotational delay. And
> also if you know where each instruction is, you can save a few words
> by using the instructions as constants. That lived on as long as the
> PDP-11 where the Unix C compiler had a peephole optimization to turn
>
> OP #N,N(R)
>
> into
>
> OP (PC),N(R)
>
> to use the same instruction word as both the constant N and the index offset N.
>
>> High-level vs. assembler: Hardly anybody does assembler any more.
>
> FORTRAN did that, but producing shockingly good code. It did stuff
> that's still pretty advanced, like nested loop merging.
>
>> RISC vs. CISC: The really complex CISC-architectures died out.
>> The difference is now less important with superscalar architectures.
>
> Technology did that. CISC made sense when RAM was expensive, microcode
> ROM was faster than RAM, and caches were too expensive for anything but
> high end mainframes. That was then.
So you're saying Intel is not CISC?
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405172 is a reply to message #405149] |
Thu, 04 February 2021 19:14 |
Dan Espen
Messages: 3867
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Thomas Koenig <tkoenig@netcologne.de> writes:
> RISC vs. CISC: The really complex CISC-architectures died out.
> The difference is now less important with superscalar architectures.
So I'm thinking, wait a minute, isn't X86 CISC? I've seen some of it's
instruction set, sure looks CISCy to me. Here's what Google says:
x86 is definitely CISC, but one of the first things a modern x86 CPU
does with an instruction stream is convert it into a different
instruction set that it uses internally, which is (but doesn't have to
be) more RISC-like. Effectively, they appear as CISC to the outside
world, but are RISC under the hood.Nov 28, 2018
So I'm guessing the other big CISC player is Z/Series. Probably the
same cop-out applies.
I started out programming an IBM 14xx which took CISC to a whole new
level since even memory addressing was in decimal. A thoroughly
pleasant environment to work in. So, big CISC fan here.
--
Dan Espen
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405173 is a reply to message #405169] |
Thu, 04 February 2021 19:18 |
scott
Messages: 4237
Registered: February 2012
Karma: 0
|
Senior Member |
|
|
J. Clarke <jclarke.873638@gmail.com> writes:
> On Thu, 4 Feb 2021 21:08:25 -0000 (UTC), John Levine <johnl@taugh.com>
> wrote:
>
>> In article <u2mo1gl05lpufm1u3td97upt5vvqv7p6pc@4ax.com>,
>> J. Clarke <jclarke.873638@gmail.com> wrote:
>>>> > RISC vs. CISC: The really complex CISC-architectures died out.
>>>
>>> What do you consider to be a "really complex CISC-architecture"?
>>
>> The usual example is VAX.
>>
>> I'd say IBM zSeries is pretty CISC but it has a unique niche..
>
> You might want to compare those to Intel.
>
> The instruction set reference for the VAX is a single chapter with 141
> pages. The instruction set reference for Intel is three volumes with
> more than 500 pages each.
And the architecture reference manual for ARMv8 (soi disant RISC processor) is
pushing 9000 pages, 2200 pages are the instruction sets, and that doesn't
include the scalable vector extension instructions (hundreds of instructions).
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405175 is a reply to message #405171] |
Thu, 04 February 2021 22:04 |
John Levine
Messages: 1405
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
In article <6f1p1gtq5f3hsu36gmjfl08af7tajgdkh2@4ax.com>,
J. Clarke <jclarke.873638@gmail.com> wrote:
>> Technology did that. CISC made sense when RAM was expensive, microcode
>> ROM was faster than RAM, and caches were too expensive for anything but
>> high end mainframes. That was then.
>
> So you're saying Intel is not CISC?
I think that x86 tells us that with enough thrust pigs can indeed fly.
Your point about the enormously long X86 architecture manual is a good one but it feels
to me less like Intel is making the instruction set more complex and more like it's
adding coprocessors that are started by instructions. Dunno how useful a distinction
it is but I think you can make a distinction between the VAX approach of making every
instruction and every address mode utterly general and the x86 adding features to speed
up graphics arithmetic or enable nested virtualization.
--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405176 is a reply to message #405172] |
Thu, 04 February 2021 22:14 |
John Levine
Messages: 1405
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
In article <rvi2kc$k4r$1@dont-email.me>,
Dan Espen <dan1espen@gmail.com> wrote:
> x86 is definitely CISC, but one of the first things a modern x86 CPU
> does with an instruction stream is convert it into a different
> instruction set that it uses internally, which is (but doesn't have to
> be) more RISC-like. Effectively, they appear as CISC to the outside
> world, but are RISC under the hood.Nov 28, 2018
>
> So I'm guessing the other big CISC player is Z/Series. Probably the
> same cop-out applies.
It's even more like that than you imagine.
The complex zSeries instructions are implemented in millicode, which is vertical microcode
consisting of simpler zSeries instructions implemented in hardware.
https://www.cmg.org/wp-content/uploads/2016/08/the-what-and- why-of-system-z-millicode.pdf
https://www.researchgate.net/publication/224103049_Millicode _in_an_IBM_zSeries_processor
--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405183 is a reply to message #405149] |
Fri, 05 February 2021 07:18 |
|
|
Originally posted by: gareth evans
On 05/02/2021 01:07, Andreas Kohlbach wrote:
> On Thu, 4 Feb 2021 21:51:32 +0000, gareth evans wrote:
>>
>> On 04/02/2021 19:07, Anssi Saari wrote:
>>> Thomas Koenig <tkoenig@netcologne.de> writes:
>>>
>>>> RISC vs. CISC: The really complex CISC-architectures died out.
>>>> The difference is now less important with superscalar architectures.
>>>
>>> Might this be coming back with Apple's M1 showing how AArch64 can run
>>> circles around x86-64 at lower power?
>>
>> I can't get my head round assembler on the ARM architecture when there
>> are no operations on memory locations, with all data firstly having
>> to be moved into registers.
>
> Recently I also had a look into ARM Assembly. But...
>
>> Too much exposure to PDP11 and X86, I suppose.
>
> ... yes, that might be a problem. I had a look into 6502, Z80 and M68000
> before. That kind id spoils the learning of something different for me.
>
Many moons ago, I did a 3D Space Invaders game for the Oric. This was a
6502 machine code exercise saving back to a cassette recorder.
My approach to dealing with the inevitable bugs in the software was to
occasionally sprinkle a succession of three NOPs into the machine code
so that bugs might be resolved with a 3-byte jump to replace those NOPs.
1983; Ye Gods! 38 years ago!
What progress there has been in those 38 years because going back 38
years prior to then takes us to 1945 with its Williams Tubes and its
mercury delay lines, not to mention 100s of valves (tubes to the Yanks)
that had increased reliability if you never switched them off!
|
|
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405185 is a reply to message #405184] |
Fri, 05 February 2021 08:11 |
|
|
Originally posted by: gareth evans
On 05/02/2021 12:39, Thomas Koenig wrote:
> Andreas Kohlbach <ank@spamfence.net> schrieb:
>
>> ... yes, that might be a problem. I had a look into 6502, Z80 and M68000
>> before. That kind id spoils the learning of something different for me.
>
> I did a bit of assembler on a 6502 (a C64), and some on a Z80.
>
> When I first laid hands on a handbook for the 68000, I thought "This is
> not assembler, this is a high-level language."
>
IMHO the 68000 showed its descent from the PDP11 but with the
lamentable loss of the extra level of indirect addressing
on every address mode, indirect anyway, auto increment,
auto decrement and indexed.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405187 is a reply to message #405165] |
Fri, 05 February 2021 08:35 |
|
|
Originally posted by: Thomas Koenig
John Levine <johnl@taugh.com> schrieb:
> CISC made sense when RAM was expensive, microcode
> ROM was faster than RAM, and caches were too expensive for anything but
> high end mainframes. That was then.
Interestingly enough, the first PDP-11 with cache was introdued
1975, the Data General Eclipse had a cache with its introduction
in 1974.
Seems like an idea whose time had come for minicomputers in the
mid-1970's, when CISC was in full swing.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405194 is a reply to message #405187] |
Fri, 05 February 2021 12:27 |
Elliott Roper
Messages: 129
Registered: October 2012
Karma: 0
|
Senior Member |
|
|
On 5 Feb 2021 at 13:35:15 GMT, "Thomas Koenig" <tkoenig@netcologne.de> wrote:
> John Levine <johnl@taugh.com> schrieb:
>
>> CISC made sense when RAM was expensive, microcode
>> ROM was faster than RAM, and caches were too expensive for anything but
>> high end mainframes. That was then.
>
> Interestingly enough, the first PDP-11 with cache was introdued
> 1975, the Data General Eclipse had a cache with its introduction
> in 1974.
>
> Seems like an idea whose time had come for minicomputers in the
> mid-1970's, when CISC was in full swing.
It missed the bus (pun intended)
the J-11 (aka 11/73) Q-bus main memory was near as dammit quick as the 11/70's
cache.
--
To de-mung my e-mail address:- fsnospam$elliott$$
PGP Fingerprint: 1A96 3CF7 637F 896B C810 E199 7E5C A9E4 8E59 E248
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405196 is a reply to message #405184] |
Fri, 05 February 2021 13:37 |
Peter Flass
Messages: 8375
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
Thomas Koenig <tkoenig@netcologne.de> wrote:
> Andreas Kohlbach <ank@spamfence.net> schrieb:
>
>> ... yes, that might be a problem. I had a look into 6502, Z80 and M68000
>> before. That kind id spoils the learning of something different for me.
>
> I did a bit of assembler on a 6502 (a C64), and some on a Z80.
>
> When I first laid hands on a handbook for the 68000, I thought "This is
> not assembler, this is a high-level language."
>
Wonderful architectuere.
--
Pete
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405197 is a reply to message #405183] |
Fri, 05 February 2021 14:03 |
Charlie Gibbs
Messages: 5313
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 2021-02-05, gareth evans <headstone255@yahoo.com> wrote:
> Many moons ago, I did a 3D Space Invaders game for the Oric. This was a
> 6502 machine code exercise saving back to a cassette recorder.
>
> My approach to dealing with the inevitable bugs in the software was to
> occasionally sprinkle a succession of three NOPs into the machine code
> so that bugs might be resolved with a 3-byte jump to replace those NOPs.
Still, if you had to jump out somewhere other than where those three NOPs
were, you'd still have to do some fancy footwork. I didn't bother; I'd
just stuff in the jump, put a copy of the overlaid instruction in the
patch area, then continue with the patch code.
> 1983; Ye Gods! 38 years ago!
Even longer for me. Where does the time go?
> What progress there has been in those 38 years because going back 38
> years prior to then takes us to 1945 with its Williams Tubes and its
> mercury delay lines, not to mention 100s of valves (tubes to the Yanks)
> that had increased reliability if you never switched them off!
Nowadays it's hard drives that are more reliable if you never switch
them off. I run my boxes 24/7 and have hardly ever had to replace
a drive before the whole machine is superseded after 10 years or so.
--
/~\ Charlie Gibbs | "Some of you may die,
\ / <cgibbs@kltpzyxm.invalid> | but it's a sacrifice
X I'm really at ac.dekanfrus | I'm willing to make."
/ \ if you read it the right way. | -- Lord Farquaad (Shrek)
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405198 is a reply to message #405183] |
Fri, 05 February 2021 14:23 |
Joe Pfeiffer
Messages: 764
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
gareth evans <headstone255@yahoo.com> writes:
>
> Many moons ago, I did a 3D Space Invaders game for the Oric. This was a
> 6502 machine code exercise saving back to a cassette recorder.
>
> My approach to dealing with the inevitable bugs in the software was to
> occasionally sprinkle a succession of three NOPs into the machine code
> so that bugs might be resolved with a 3-byte jump to replace those NOPs.
>
> 1983; Ye Gods! 38 years ago!
>
> What progress there has been in those 38 years because going back 38
> years prior to then takes us to 1945 with its Williams Tubes and its
> mercury delay lines, not to mention 100s of valves (tubes to the Yanks)
> that had increased reliability if you never switched them off!
Ah, yes. Just a couple of years later I was doing some programming on a
preproduction Fairchild F9450 (Mil-Std 1750A architecture). I
discovered that occasionally specific extra high order bits were set in
the PC when an interrupt took place -- hardware bug. I wound up
inserting jumps at the addresses the PC wound up at, back to where it
was supposed to be.
I had to do extensive bowdlerizing of the comments in my code before
turning it over to them. Then the other programmer on the project took
the wrong floppy with him... fortunately they had a good sense of humor
about it.
At the time, yield was being measured in wafers per chip. I don't know
if they ever did get it to something useable, or if it ever went into
actual production.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405200 is a reply to message #405196] |
Fri, 05 February 2021 14:36 |
Charlie Gibbs
Messages: 5313
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 2021-02-05, Peter Flass <peter_flass@yahoo.com> wrote:
> Thomas Koenig <tkoenig@netcologne.de> wrote:
>
>> Andreas Kohlbach <ank@spamfence.net> schrieb:
>>
>>> ... yes, that might be a problem. I had a look into 6502, Z80 and M68000
>>> before. That kind id spoils the learning of something different for me.
>>
>> I did a bit of assembler on a 6502 (a C64), and some on a Z80.
>>
>> When I first laid hands on a handbook for the 68000, I thought "This is
>> not assembler, this is a high-level language."
>
> Wonderful architectuere.
Yes, and one more example of how it's better to be first than best.
As someone once quipped, "It's a good thing iAPX432 didn't catch on.
Otherwise, a truly horrible Intel architecture might have taken over
the world."
--
/~\ Charlie Gibbs | "Some of you may die,
\ / <cgibbs@kltpzyxm.invalid> | but it's a sacrifice
X I'm really at ac.dekanfrus | I'm willing to make."
/ \ if you read it the right way. | -- Lord Farquaad (Shrek)
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405201 is a reply to message #405151] |
Fri, 05 February 2021 15:35 |
Anne & Lynn Wheel
Messages: 3156
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Jim Jackson <jj@franjam.org.uk> writes:
> Ethernet v. IBM Token Ring - did you know that ethernet could never work
> as it would collapse under moderately heavy loads?
IBM E&S center in Dallas turned out report comparing 16mbit token-ring
to ethernet ... my only explanation was that they used prototype 3mbit
ethernet before standard that included listen-before-transmit.
New IBM Almaden research was provisioned with extensive cat5 assuming
16mbit token-ring ... but found that ethernet had higher aggregate
network throughput, lower latency, and ethernet cards had higher
per card throughput
ACM SIGCOMM 8/16-19, 1988, V18N4 had extensive analysis of 10mbit
ethernet ... they demonstration 30station lan with low-level device
driver code in constant loop transmitting minimum size packets with
effective LAN throughput dropping off to 8mbit/sec. These were $69
10mbit ethernet cards.
$899 16mbit T/R microchannel cards were significantly kneecapped ... IBM
PC/RT had done their own at-bus 4mbit token-ring card and found it had
higher per card throughput than microchannel 16mbit cards (and were
forbidden from doing their own 16mbit T/R microchannel cards). The
communication group was fiercely fighting off client/server and
distributed computing and 16mbit T/R microchannel cards had design point
of 300+ stations sharing same LAN doing dumb terminal emulation.
Claimed motivation for development of T/R and CAT5 ... was that
the 3270 dumb terminals was 3270 coax was starting to exceed
bldg load weights ... i.e. each 3270 dumb terminal had coax
that ran from machine room for each 3270 (cable trays were
becoming massive) ... communication group was trying
to address the problem with CAT5 T/R lans going to IBM/PCs
emulating 3270 dumb terminals.
--
virtualization experience starting Jan1968, online at home since Mar1970
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405202 is a reply to message #405200] |
Fri, 05 February 2021 15:39 |
|
|
Originally posted by: maus
On 2021-02-05, Charlie Gibbs <cgibbs@kltpzyxm.invalid> wrote:
> On 2021-02-05, Peter Flass <peter_flass@yahoo.com> wrote:
>
>> Thomas Koenig <tkoenig@netcologne.de> wrote:
>>
>>> Andreas Kohlbach <ank@spamfence.net> schrieb:
>>>
>>>> ... yes, that might be a problem. I had a look into 6502, Z80 and M68000
>>>> before. That kind id spoils the learning of something different for me.
>>>
>>> I did a bit of assembler on a 6502 (a C64), and some on a Z80.
>>>
>>> When I first laid hands on a handbook for the 68000, I thought "This is
>>> not assembler, this is a high-level language."
>>
>> Wonderful architectuere.
>
> Yes, and one more example of how it's better to be first than best.
> As someone once quipped, "It's a good thing iAPX432 didn't catch on.
> Otherwise, a truly horrible Intel architecture might have taken over
> the world."
>
Somewhat similiar, after coming out of a film years ago, another child
said, "Arent we lucky that the Germans and Japs did not win the war."
--
greymausg@mail.com
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405203 is a reply to message #405201] |
Fri, 05 February 2021 16:27 |
|
|
Originally posted by: Bob Eager
On Fri, 05 Feb 2021 10:35:46 -1000, Anne & Lynn Wheeler wrote:
> IBM E&S center in Dallas turned out report comparing 16mbit token-ring
> to ethernet ... my only explanation was that they used prototype 3mbit
> ethernet before standard that included listen-before-transmit.
The university where I worked had their first ring network in 1978. It
ended up with several linked rings until it was decommissioned in favour
of ethernet in about 1992.
It worked well, but it wasn't IBM.
--
Using UNIX since v6 (1975)...
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405207 is a reply to message #405151] |
Fri, 05 February 2021 16:41 |
Jorgen Grahn
Messages: 606
Registered: March 2012
Karma: 0
|
Senior Member |
|
|
On Thu, 2021-02-04, Jim Jackson wrote:
....
> Ethernet v. IBM Token Ring - did you know that ethernet could never work
> as it would collapse under moderately heavy loads?
Solved by redefining what "Ethernet" means, right?
/Jorgen
--
// Jorgen Grahn <grahn@ Oo o. . .
\X/ snipabacken.se> O o .
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405208 is a reply to message #405156] |
Fri, 05 February 2021 16:54 |
Jorgen Grahn
Messages: 606
Registered: March 2012
Karma: 0
|
Senior Member |
|
|
On Thu, 2021-02-04, Scott Lurndal wrote:
> Anssi Saari <as@sci.fi> writes:
>> Thomas Koenig <tkoenig@netcologne.de> writes:
>
>>
>>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>>> are using this even on Windows.
>>
>> Has this been decided then? What about Emacs in Windows? Actually I
>> probably would've abandoned Emacs if it weren't for the VHDL and ORG
>> modes.
>
> It's still around 50-50 here. The Verilog
> types tend to use emacs (and csh), and the software folks vim (and bash/ksh).
Around here, very few use vim, and I've not met a fellow Emacs user
for many years.
/Jorgen
--
// Jorgen Grahn <grahn@ Oo o. . .
\X/ snipabacken.se> O o .
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405209 is a reply to message #405201] |
Fri, 05 February 2021 17:38 |
|
|
Originally posted by: J. Clarke
On Fri, 05 Feb 2021 10:35:46 -1000, Anne & Lynn Wheeler
<lynn@garlic.com> wrote:
> Jim Jackson <jj@franjam.org.uk> writes:
>> Ethernet v. IBM Token Ring - did you know that ethernet could never work
>> as it would collapse under moderately heavy loads?
>
> IBM E&S center in Dallas turned out report comparing 16mbit token-ring
> to ethernet ... my only explanation was that they used prototype 3mbit
> ethernet before standard that included listen-before-transmit.
>
> New IBM Almaden research was provisioned with extensive cat5 assuming
> 16mbit token-ring ... but found that ethernet had higher aggregate
> network throughput, lower latency, and ethernet cards had higher
> per card throughput
>
> ACM SIGCOMM 8/16-19, 1988, V18N4 had extensive analysis of 10mbit
> ethernet ... they demonstration 30station lan with low-level device
> driver code in constant loop transmitting minimum size packets with
> effective LAN throughput dropping off to 8mbit/sec. These were $69
> 10mbit ethernet cards.
>
> $899 16mbit T/R microchannel cards were significantly kneecapped ... IBM
> PC/RT had done their own at-bus 4mbit token-ring card and found it had
> higher per card throughput than microchannel 16mbit cards (and were
> forbidden from doing their own 16mbit T/R microchannel cards). The
> communication group was fiercely fighting off client/server and
> distributed computing and 16mbit T/R microchannel cards had design point
> of 300+ stations sharing same LAN doing dumb terminal emulation.
>
> Claimed motivation for development of T/R and CAT5 ... was that
> the 3270 dumb terminals was 3270 coax was starting to exceed
> bldg load weights ... i.e. each 3270 dumb terminal had coax
> that ran from machine room for each 3270 (cable trays were
> becoming massive) ... communication group was trying
> to address the problem with CAT5 T/R lans going to IBM/PCs
> emulating 3270 dumb terminals.
And the sad part is that other than it being wifi instead of T/R
that's _still_ the way the mainframe is accessed most of the time. I'm
considered a weirdo for "slowing myself down" by using Eclipse with
the CompuWare plugin. I admit, one of the old guys _can_ go through
three screens and type a keyword about as fast as I can right click
and do the same thing but he's been doing that particular sequence for
40 years.
|
|
|
|
| Re: Ethernet vs token ring, was Holy wars of the past - how did they turn out? [message #405210 is a reply to message #405207] |
Fri, 05 February 2021 17:39 |
John Levine
Messages: 1405
Registered: December 2011
Karma: 0
|
Senior Member |
|
|
In article <slrns1rerl.2ptb.grahn+nntp@frailea.sa.invalid>,
Jorgen Grahn <grahn+nntp@snipabacken.se> wrote:
> On Thu, 2021-02-04, Jim Jackson wrote:
> ...
>> Ethernet v. IBM Token Ring - did you know that ethernet could never work
>> as it would collapse under moderately heavy loads?
>
> Solved by redefining what "Ethernet" means, right? The studies claiming
it would collide to death were wrong. Someone else suggested they didn't
take into account listen-before-transmit.
No, the old DIX spec coax Ethernet worked fine under heavy load.
I agree that what we call Ethernet now has little in common with coax
Ethernet other than the framing, but the changes weren't because it
didn't work. It rapidly became clear that a star configuration with hubs
was easier to install and manage than a shared coax bus, and twisted pair
is a lot cheaper and easier to handle than either kind of coax, not to
mention now a lot faster.
--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405220 is a reply to message #405183] |
Sat, 06 February 2021 03:52 |
Anssi Saari
Messages: 327
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
gareth evans <headstone255@yahoo.com> writes:
> Many moons ago, I did a 3D Space Invaders game for the Oric. This was a
> 6502 machine code exercise saving back to a cassette recorder.
I remember that name, it seems Oric sold here in Finland back then. I
don't think I ever used one though.
In fact, with a quick look at a magazine archive, when the first(ish?)
home computer magazine "Mikrobitti" started here in 1984, the first
issue had a little basic game listing for the Oric.
|
|
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405222 is a reply to message #405208] |
Sat, 06 February 2021 04:25 |
Anssi Saari
Messages: 327
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
Jorgen Grahn <grahn+nntp@snipabacken.se> writes:
> Around here, very few use vim, and I've not met a fellow Emacs user
> for many years.
I guess I haven't either. Or I don't really know what my colleagues use
since I've started in a new job during the pandemic... I remember I had
an interesting talk years ago, probably in the naughties, with a
stranger because I was wearing my Gnus T-Shirt. As I recall he
recognized the elisp code on the shirt as lisp but wasn't an Emacs user.
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405234 is a reply to message #405155] |
Sat, 06 February 2021 05:47 |
Niklas Karlsson
Messages: 265
Registered: January 2012
Karma: 0
|
Senior Member |
|
|
On 2021-02-04, Anssi Saari <as@sci.fi> wrote:
> Thomas Koenig <tkoenig@netcologne.de> writes:
>
>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>> are using this even on Windows.
>
> Has this been decided then? What about Emacs in Windows? Actually I
> probably would've abandoned Emacs if it weren't for the VHDL and ORG
> modes.
I run Emacs for a sole purpose that's probably rather obscure outside
certain circles in this country: LysKOM. KOM is a kind of bulletin board
system that first turned up in the late 1970s. Like many other older
bulletin board solutions, it is in some ways a great improvement on its
successors. I don't believe it ever had much traction outside Sweden,
though I think someone managed to push it (in an English-language
version known as COM) as an EU standard (that very few used).
It received some Swedish media attention in the 1980s due to being
involved in a libel case.
KOM initially worked by logging in timesharing style, so the client was
local only, but LysKOM speaks TCP/IP. It is so named because it was
developed by Lysator, the computing society at the University of
Linköping.
There are various LysKOM clients including a JavaScript one, but the
Emacs LISP one is the most complete and feature-rich.
Wow, I'm actually on topic!
Niklas
--
In college, I wrote a TECO-like progamming language as a joke - one-letter
statements, totally unreadable. Then I discovered sendmail, and stopped,
because the joke had been done so much better than I ever could.
-- Mark 'Kamikaze' Hughes
|
|
|
|
| Re: Holy wars of the past - how did they turn out? [message #405235 is a reply to message #405208] |
Sat, 06 February 2021 06:03 |
|
|
Originally posted by: Thomas Koenig
Jorgen Grahn <grahn+nntp@snipabacken.se> schrieb:
> On Thu, 2021-02-04, Scott Lurndal wrote:
>> Anssi Saari <as@sci.fi> writes:
>>> Thomas Koenig <tkoenig@netcologne.de> writes:
>>
>>>
>>>> Emacs vs. vi: vim has led to a resurgence of vi, and many people
>>>> are using this even on Windows.
>>>
>>> Has this been decided then? What about Emacs in Windows? Actually I
>>> probably would've abandoned Emacs if it weren't for the VHDL and ORG
>>> modes.
>>
>> It's still around 50-50 here. The Verilog
>> types tend to use emacs (and csh), and the software folks vim (and bash/ksh).
> Around here, very few use vim, and I've not met a fellow Emacs user
> for many years.
I use both (vi for typing Usenet articles such as this and simple
texts, emacs for its nice syntax features and indentation).
What emacs is lacking is a block-oriented folding mode, where you can
just fold in that other side of the if statement when you want to
look at the condition.
|
|
|
|
Members
Search
Help
Register
Login
Home
