Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #50314 > unrolled thread

Hard Forth

Back to article view | Back to comp.lang.forth

Contents

  Hard Forth lehs <skydda.barnen@gmail.com> - 2016-08-21 01:04 -0700
    Hard Forth Mark Wills <markwills1970@gmail.com> - 2016-08-21 01:12 -0700
      Re: Hard Forth lehs <skydda.barnen@gmail.com> - 2016-08-21 01:17 -0700
        Re: Hard Forth Mark Wills <markwills1970@gmail.com> - 2016-08-21 01:54 -0700
          Re: Hard Forth lehs <skydda.barnen@gmail.com> - 2016-08-21 07:26 -0700
            Re: Hard Forth Coos Haak <htrofhc@gmail.com> - 2016-08-21 17:07 +0200
          Re: Hard Forth rickman <gnuarm@gmail.com> - 2016-08-21 17:37 -0400
            Re: Hard Forth Mark Wills <markwills1970@gmail.com> - 2016-08-21 14:51 -0700
        Re: Hard Forth lawrencedo99@gmail.com - 2016-09-15 23:43 -0700
    Re: Hard Forth Julian Fondren <julian.fondren@gmail.com> - 2016-08-21 01:16 -0700
      Re: Hard Forth lehs <skydda.barnen@gmail.com> - 2016-08-21 01:20 -0700
        Re: Hard Forth julian.fondren@endurance.com - 2016-08-21 01:36 -0700
        Re: Hard Forth Julian Fondren <julian.fondren@gmail.com> - 2016-08-21 01:41 -0700
    Your definition of "Forth" is ...? - was [Re: Hard Forth] Richard Owlett <rowlett@cloud85.net> - 2016-08-21 14:16 -0500
      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lehs <skydda.barnen@gmail.com> - 2016-08-21 13:50 -0700
        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] hughaguilar96@gmail.com - 2016-08-21 19:46 -0700
          Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Cecil Bayona <cbayona@cbayona.com> - 2016-08-21 22:29 -0500
            Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Ron Aaron <rambamist@gmail.com> - 2016-08-22 07:03 +0300
              Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Cecil Bayona <cbayona@cbayona.com> - 2016-08-21 23:13 -0500
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Ron Aaron <rambamist@gmail.com> - 2016-08-22 07:18 +0300
            Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] hughaguilar96@gmail.com - 2016-08-21 22:08 -0700
              Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Cecil Bayona <cbayona@cbayona.com> - 2016-08-22 00:55 -0500
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] hughaguilar96@gmail.com - 2016-08-21 23:27 -0700
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-08-22 06:00 -0400
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lehs <skydda.barnen@gmail.com> - 2016-08-21 23:30 -0700
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Paul Rubin <no.email@nospam.invalid> - 2016-08-22 00:04 -0700
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] "HAA" <someone@microsoft.com> - 2016-08-23 16:11 +1000
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lehs <skydda.barnen@gmail.com> - 2016-08-22 23:45 -0700
            Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] "Elizabeth D. Rather" <erather@forth.com> - 2016-08-21 21:42 -1000
          Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-15 23:57 -0700
            Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-09-16 03:30 -0400
              Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-16 08:37 +0000
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Lars Brinkhoff <lars.spam@nocrew.org> - 2016-09-16 12:16 +0200
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-16 12:56 +0000
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-16 09:39 -0500
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-09-17 02:13 -0400
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-16 23:46 -0700
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-17 02:02 -0500
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-17 00:41 -0700
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-09-17 16:51 -0400
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-17 17:24 -0700
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-18 15:44 -0500
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Bernd Paysan <bernd@net2o.de> - 2016-10-13 00:49 +0000
                        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) - 2016-10-13 03:32 +0200
                        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-10-12 23:34 -0400
                        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-10-13 04:02 -0500
                          Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-10-13 15:04 -0700
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Mark Wills <markwills1970@gmail.com> - 2016-09-17 00:04 -0700
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] mhx@iae.nl - 2016-09-17 01:28 -0700
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-17 02:04 -0700
                      Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-17 15:38 +0000
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-17 10:20 -0500
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] rickman <gnuarm@gmail.com> - 2016-09-17 16:52 -0400
                  Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-17 14:18 +0000
                    Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-17 17:22 -0700
        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) - 2016-08-29 15:33 +0200
        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Bruce Axtens <bruce.axtens@gmail.com> - 2016-09-12 19:16 +0800
          Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] JUERGEN <epldfpga@aol.com> - 2016-09-16 10:29 -0700
        Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-15 23:54 -0700
          Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-16 03:03 -0500
            Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] lawrencedo99@gmail.com - 2016-09-16 01:19 -0700
              Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-09-16 08:32 +0000
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] albert@cherry.spenarnc.xs4all.nl (Albert van der Horst) - 2016-09-16 13:15 +0200
                Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Bernd Paysan <bernd@net2o.de> - 2016-10-13 01:05 +0000
              Re: Your definition of "Forth" is ...? - was [Re: Hard Forth] Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-09-16 09:45 -0500

Page 3 of 4 — ← Prev page 1 2 [3] 4  Next page →

#51087 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Sunday, September 18, 2016 at 8:51:05 AM UTC+12, rickman wrote:
>
> On 9/17/2016 2:46 AM, Lawrence D’Oliveiro wrote:
>>
>> On Saturday, September 17, 2016 at 6:13:10 PM UTC+12, rickman wrote:
>>>
>>> The 68000 has numerous issues such as higher
>>> transistor count and slower throughput compared to the 8086.
>>
>> Yet Microsoft Windows (on x86) was not able to match the performance
>> of the Macintosh (on 680x0). It was only when all the Windows code was
>> able to move to a flat 32-bit address space (late 1990s onwards) that
>> it was finally able to pull ahead. The segmented-address issue delayed
>> this transition by about a decade compared to the Macintosh.
> 
> Huh?  How can you compare two different machines with two different OS 
> and say the difference is purely the hardware?

Either that, or Microsoft writes crap code...

[toc] | [prev] | [next] | [standalone]

#51100 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

rickman <gnuarm@gmail.com> wrote:
> Only gear heads argue about which CPU *should* have won the race.

Well, yeah, but gearheads 'r us.  :-)

Andrew.

[toc] | [prev] | [next] | [standalone]

#51816 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

Am Sat, 17 Sep 2016 16:51:03 -0400 schrieb rickman:

> That's your opinion.  The x86 was the design the world wanted at the
> time.

You mean IBM.  IBM choose the 8088 over the 68000 due to time to market 
pressure.  At that time, the 68000 was just being debugged, and the 8088 
was ready.  It was Microsoft which convinced IBM to actually use a 16 bit 
processor; they had planned to use an 8 bit processor originally.  It was 
supposed to be a home computer, nothing serious, as they feared 
cannibalizing their workstation market (which it eventually did, but 
first, it cannibalized the typewriter market, something IBM could live 
with).

>  In other words, it best met the needs of users.  Only gear heads argue
> about which CPU *should* have won the race.

Of course we argue about what should have won from a technical point of 
merit.  Mundane marketing issues like availability on the market at a 
specific point in time doesn't count.  For several years to come after 
the initial PC, the 68k competitors like Atari ST, Mac, Amiga, where all 
better offerings than the original IBM PC, and those machines had found a 
lot of customers, too.  That changed when the clones came and pushed the 
price down.  And at that point in time, the 386 was already there; the 
clone makers were the first to introduce the 386 (Compaq, IIRC), just 
with ISA, when IBM wanted to transition people to Microchannel, a failed 
attempt of gaining back control.  Later, PCI replaced ISA, it came from 
Intel and was open.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
net2o ID: kQusJzA;7*?t=uy@X}1GWr!+0qqp_Cn176t4(dQ*
http://bernd-paysan.de/

[toc] | [prev] | [next] | [standalone]

#51820 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

In article <ntmlnn$qa9$3@dont-email.me>, Bernd Paysan  <bernd@net2o.de> wrote:
>Am Sat, 17 Sep 2016 16:51:03 -0400 schrieb rickman:
>
>> That's your opinion.  The x86 was the design the world wanted at the
>> time.
>
>You mean IBM.  IBM choose the 8088 over the 68000 due to time to market
>pressure.  At that time, the 68000 was just being debugged, and the 8088
>was ready.  It was Microsoft which convinced IBM to actually use a 16 bit
>processor; they had planned to use an 8 bit processor originally.  It was
>supposed to be a home computer, nothing serious, as they feared
>cannibalizing their workstation market (which it eventually did, but
>first, it cannibalized the typewriter market, something IBM could live
>with).
>
>>  In other words, it best met the needs of users.  Only gear heads argue
>> about which CPU *should* have won the race.
>
>Of course we argue about what should have won from a technical point of
>merit.  Mundane marketing issues like availability on the market at a
>specific point in time doesn't count.  For several years to come after
>the initial PC, the 68k competitors like Atari ST, Mac, Amiga, where all
>better offerings than the original IBM PC, and those machines had found a
>lot of customers, too.  That changed when the clones came and pushed the

I remember how all technically savvy people immediately bought an Atari
or an Amiga, as soon as they became available. Then Veltman had his
Schoonschip program running on the Mac, Atari and Amiga.
(Started in the 1980 this got him the Nobelprice around 2000).

>price down.  And at that point in time, the 386 was already there; the
>clone makers were the first to introduce the 386 (Compaq, IIRC), just
>with ISA, when IBM wanted to transition people to Microchannel, a failed
>attempt of gaining back control.  Later, PCI replaced ISA, it came from
>Intel and was open.

The intel 386 inspired nobody until Linux Torvalds came around.

Groetjes Albert

>
>--
>Bernd Paysan

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#51823 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On 10/12/2016 8:49 PM, Bernd Paysan wrote:
> Am Sat, 17 Sep 2016 16:51:03 -0400 schrieb rickman:
>
>> That's your opinion.  The x86 was the design the world wanted at the
>> time.
>
> You mean IBM.  IBM choose the 8088 over the 68000 due to time to market
> pressure.  At that time, the 68000 was just being debugged, and the 8088
> was ready.  It was Microsoft which convinced IBM to actually use a 16 bit
> processor; they had planned to use an 8 bit processor originally.  It was
> supposed to be a home computer, nothing serious, as they feared
> cannibalizing their workstation market (which it eventually did, but
> first, it cannibalized the typewriter market, something IBM could live
> with).
>
>>  In other words, it best met the needs of users.  Only gear heads argue
>> about which CPU *should* have won the race.
>
> Of course we argue about what should have won from a technical point of
> merit.  Mundane marketing issues like availability on the market at a
> specific point in time doesn't count.  For several years to come after
> the initial PC, the 68k competitors like Atari ST, Mac, Amiga, where all
> better offerings than the original IBM PC, and those machines had found a
> lot of customers, too.  That changed when the clones came and pushed the
> price down.  And at that point in time, the 386 was already there; the
> clone makers were the first to introduce the 386 (Compaq, IIRC), just
> with ISA, when IBM wanted to transition people to Microchannel, a failed
> attempt of gaining back control.  Later, PCI replaced ISA, it came from
> Intel and was open.

You can exclude such commercial aspects as availability, but then why 
not exclude price?  If you want to be a gearhead and look at what 
"should" have happened, then why consider any of the practicality issues?

I guess I'm just getting too old to care about such things.  I'd rather 
look ahead and see where things are going.

-- 

Rick C

[toc] | [prev] | [next] | [standalone]

#51826 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

Bernd Paysan <bernd@net2o.de> wrote:
> Am Sat, 17 Sep 2016 16:51:03 -0400 schrieb rickman:
> 
>>  In other words, it best met the needs of users.  Only gear heads argue
>> about which CPU *should* have won the race.
> 
> Of course we argue about what should have won from a technical point of 
> merit.

Well, yes.  Gearheads 'r us.  :-)

> Mundane marketing issues like availability on the market at a
> specific point in time doesn't count.  For several years to come
> after the initial PC, the 68k competitors like Atari ST, Mac, Amiga,
> where all better offerings than the original IBM PC,

Better than the original PC, sure, but the PC did have the open bus
architecture which was extremely well-documented.  This spawned a
thriving industry.

Andrew.

[toc] | [prev] | [next] | [standalone]

#51847 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Thursday, October 13, 2016 at 10:03:01 PM UTC+13, Andrew Haley wrote:
> Better than the original PC, sure, but the PC did have the open bus
> architecture which was extremely well-documented.  This spawned a
> thriving industry.

The “original” PC--the LINC <https://www.youtube.com/watch?v=ZgPfLWt5FWE>?

[toc] | [prev] | [next] | [standalone]

#51060 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

X86 dominance in the market had nothing to do it being better. Anyone who's ever written a line of assembler on both devices knows the 68K is the better device by a mile. The X86 success was purely down to IBM's decision to use it in their new PC line, and possibly the dominance of CPM prior to that.

[toc] | [prev] | [next] | [standalone]

#51065 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Saturday, September 17, 2016 at 9:05:00 AM UTC+2, Mark Wills wrote:
> X86 dominance in the market had nothing to do it being better.
> Anyone who's ever written a line of assembler on both devices
> knows the 68K is the better device by a mile.

I did write *a lot* of assembly language for both
processors, an embedded board with CP/M for the 68K
and a 386 PC under 32bit Windows NT. I liked (or
disliked) both architectures equally well. There was
no objective technical reason to call either of them
(much) 'better' than the other one. At some point
the x86 developed itself much faster (higher clock
speeds, more memory, FP co-processor), and 68K
simply stopped to be a serious alternative.

An x86 versus 68K dispute is only useful to identify 
the real engineers from the dilettantes.

-marcel

[toc] | [prev] | [next] | [standalone]

#51067 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Saturday, September 17, 2016 at 8:28:41 PM UTC+12, m...@iae.nl wrote:
> There was no objective technical reason to call either of them
> (much) 'better' than the other one.

68K had a linear 32-bit address space. More registers. Better addressing modes. A much less painful transition to full 32-bit operation.

Did I mention it had a 32-bit linear address space?

> At some point the x86 developed itself much faster (higher clock
> speeds, more memory, FP co-processor), and 68K simply stopped to
> be a serious alternative.

That only happened because Intel could afford to spend ten times as much as Motorola on developing its processors. It needed to, to keep up.

[toc] | [prev] | [next] | [standalone]

#51078 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

mhx@iae.nl writes:
>On Saturday, September 17, 2016 at 9:05:00 AM UTC+2, Mark Wills wrote:
>> X86 dominance in the market had nothing to do it being better.
>> Anyone who's ever written a line of assembler on both devices
>> knows the 68K is the better device by a mile.
>
>I did write *a lot* of assembly language for both
>processors, an embedded board with CP/M for the 68K
>and a 386 PC under 32bit Windows NT. I liked (or
>disliked) both architectures equally well. 

The 386 has a different architecture than the 8086, and is in fact
more orthogonal than the 68000: with the exception of a few relatively
rare instructions (and 8086 holdouts) and addressing modes, every
instruction and addressing mode could use every register; i.e., the
386 is a register machine.  By contrast, the 8086 was a mess of
special-purpose instructions and especially addressing modes.  There's
a reason the registers are not called R0-R7 like in a from-the-start
register architecture (such as the PDP-11).

>An x86 versus 68K dispute is only useful to identify 
>the real engineers from the dilettantes.

The use of "x86" is useful to identify dilettantes.  Real engineers
know that 8086, IA-32, and AMD64 are different architectures and won't
mix them by applying one moniker to refer to some or all of them,
because that just confuses.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#51076 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

Mark Wills <markwills1970@gmail.com> wrote:

> X86 dominance in the market had nothing to do it being
> better. Anyone who's ever written a line of assembler on both
> devices knows the 68K is the better device by a mile.

You can't judge the quality of a processor solely by its instruction
set architecture.  It's just as important to think about the issue
rate, code density, and so on.  These days what's really important
is just as likely to be memory bandwidth, cache behaviour, and so on.

> The X86 success was purely down to IBM's decision to use it in their
> new PC line, and possibly the dominance of CPM prior to that.

And Intel's execution.  They are very good at that,

Andrew.

[toc] | [prev] | [next] | [standalone]

#51082 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On 9/17/2016 3:04 AM, Mark Wills wrote:
> X86 dominance in the market had nothing to do it being better. Anyone who's ever written a line of assembler on both devices knows the 68K is the better device by a mile. The X86 success was purely down to IBM's decision to use it in their new PC line, and possibly the dominance of CPM prior to that.

Your analysis is one dimensional.

-- 

Rick C

[toc] | [prev] | [next] | [standalone]

#51077 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

rickman <gnuarm@gmail.com> writes:
>On 9/16/2016 4:37 AM, Anton Ertl wrote:
>> Anyway, Intel could have achieved assembly-source compatibility while
>> still providing a flat address space.
>
>Huh?  The 8080 has a flat address space.  How do you extend that to a 
>larger address space without sectors.  The basic instructions assume 16 
>bit addresses.  Are you suggesting a totally different instruction set 
>on *top* of the 8 bit register instructions?

There are various ways to do it.

The 65C816 extended the 6502 instruction set to support a 16MB address
space, and was not only assembly language compatible, but even binary
compatible.

The MIPS, SPARC, and Power architectures were extended from 32-bit to
64-bit and were not only assembly language compatible, but even binary
compatible, and similar things happened to the S/360 architecture.

The 386 architecture included the 8086 and 80286 architecture through
16-bit modes.

Similarly, the AMD64 architecture has modes for running 386
architecture programs and 8086 and 80286 programs in addition to its
64-bit mode for its 64-bit architecture.

ARM went in the same direction with ARMv8: It has a 64-bit
architecture and a 32-bit architecture and modes for executing
programs in these architectures.

I am sure there are other examples.

>>> In the end Intel ended up with a full 32 bit CPU with 32 bit addressing
>>> and a 64 bit CPU with 64 bit addresses.  The purpose of the segment
>>> registers now is not so much a way to utilize a large address space, but
>>> to separate the various memory sections and support security.
>>
>> The only purpose of segment registers (in 64-bit mode) is to provide
>> thread-local storage.  Instead of extending segment descriptors for
>> 64-bit use (which would have required more space), segments in 64-bit
>> mode were reduced in capability and can no longer be used for
>> security; they have not been used for security before that, anyway.
>> Segment fans always whine about security; paging goes home and fscks
>> the system:-).
>
>You seem to be totally ignoring the incremental nature of the 
>development of the technology.  Remember how the 8086 was available 
>earlier than the 68000?  That's because they didn't try to solve 
>tomorrow's problems today.  The 68000 has numerous issues such as higher 
>transistor count and slower throughput compared to the 8086.  It was 
>only if you looked many years ahead that the 68000 looked like a better 
>design.

I did not have to look ahead.  I just had to look at the assembly
language.  Anyway, I don't know what this has to do with my refutation
of your claim about segments being used for security these days.
Although, come to think of it, maybe with "now" you mean a time when
the 80286 was current and there were attempts to use protected mode
for security (maybe in OS/2 1.x, now 25 years gone).

>  In the end the x86 line dominated, so clearly it was the better 
>choice.

They had to invent a new architecture for 32 bits, while the 68000
line could continue with its existing architecture.  The 8086 does not
dominate, AMD64 (and ARMv8) dominates; AMD64 still has 8086 in some
dark corner, but it's hardly used these days, if at all.

As for better choice: If IBM had chosen the 68000, it would have
dominated, and it would clearly have been the better choice.

As for performance,
<http://performance.netlib.org/performance/html/dhrystone.data.col0.html>
lists an 8MHz 8086 (ATT PC6300) at 0.44 Dhrystone MIPS and a 7.16MHz
68000 (Amiga 1000) at 0.54 Dhrystone MIPS, and an IBM PC (4.77MHz
8088) at 0.22 Dhrystone MIPS.  So no, they did not chose the 8088 for
performance.

- anton
-- 
M. Anton Ertl  http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
     New standard: http://www.forth200x.org/forth200x.html
   EuroForth 2016: http://www.euroforth.org/ef16/

[toc] | [prev] | [next] | [standalone]

#51086 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Sunday, September 18, 2016 at 3:38:20 AM UTC+12, Anton Ertl wrote:
> The MIPS, SPARC, and Power architectures were extended from 32-bit to
> 64-bit and were not only assembly language compatible, but even binary
> compatible, and similar things happened to the S/360 architecture.

Can’t speak for the others, but I did do some code generation for the PowerPC instruction set. It was clear that was designed to be a 64-bit architecture from the outset, so the 32-bit chips were basically operating on cut-down subsets. This would have made it fairly straightforward to transition code to the full 64-bit processors--analogously to how the original Motorola 68000 16-bit chip was really running a cut-down 32-bit architecture.

[toc] | [prev] | [next] | [standalone]

#50550 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

In article <f14aebd4-a311-4d91-b84e-225054af6583@googlegroups.com>,
lehs  <skydda.barnen@gmail.com> wrote:
>Den s=C3=B6ndag 21 augusti 2016 kl. 21:17:00 UTC+2 skrev Richard Owlett:
>> On 8/21/2016 3:04 AM, lehs wrote:
>> > Shouldn't it be possible, when using Forth, to address sectors on an ex=
>ternal disk,
>> >  read and write on this sectors regardless of what the disk contain?
>>=20
>> I've been following this thread and began wondering if everyone=20
>> is using the same definition of the word {the character sequence}=20
>> "Forth".
>>=20
>> I see multiple possible definitions of "Forth/FORTH".
>>=20
>> Question to lehs in particular and list in general -
>> Does http://www.forthos.org/ answer &/or raise pertinent questions?
>First, I'm not negative to the Forth standards. A standard is a compromise =
>and if you want to use a standard or not is up to the forther. Also, I don'=
>t want to define Forth. My idea of the definition of Forth is the core word=
>s we all know about.
>
>But it seems to me that Forth systems maybe should be more oriented to hard=
>ware. Algol didn't even had standards for input or output, since it was a l=
>amguage for algorithms. Forth is excellent for algorithms, but has always h=
>ad rather general words for input amd output.=20

Algol68 had the most elaborate definition of transput (as they called
it) for its time and it was part of the language.
It was not based on existing practice and did get well known.

>
>These days, input and output means a lot more than keyboards, screens and p=
>rinters, but Forth seems to have stopped develop in that direction since Fo=
>rth community highjacked Forth from Charles Moore.

What do you want? Have drivers in Forth developed for all hardware since
1970? I'm glad that manufacturers sometimes supply c-files to access
devices, or even documentation.

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#50885 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On 22/08/2016 4:50 AM, lehs wrote:
> but Forth seems to have stopped develop in that direction since Forth community highjacked Forth from Charles Moore.
Maybe it's time to give up on Forth (by that name) and re-badge as "Moore".

[toc] | [prev] | [next] | [standalone]

#51041 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Monday, September 12, 2016 at 12:16:34 PM UTC+1, axtens wrote:
> On 22/08/2016 4:50 AM, lehs wrote:
> > but Forth seems to have stopped develop in that direction since Forth community highjacked Forth from Charles Moore.
> Maybe it's time to give up on Forth (by that name) and re-badge as "Moore".

I cannot understand the reasoning: 
I would agree with: Keep Forth, backward compatibility, ANS ... to keep and continue, actually this is what the commercial SW manufacturers do for backward compatibility.  
MPE has introduced the SockPuppet to bridge the gap to C - one of the most important evolutions of Forth and it seems not appreciated by the Forth Community. Forth++ and MOORE was open for the last 30 years, but not many branches came out of it, which indicates that commercially the interest was limited.  

Chuck always did what he wanted. Did not like ANS, instead did the Forth Processors, which took over the world - or might do in the future.

[toc] | [prev] | [next] | [standalone]

#51021 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

On Monday, August 22, 2016 at 8:50:42 AM UTC+12, lehs wrote:
>
> Algol didn't even had standards for input or output, since it was a 
> lamguage for algorithms.

Algol 60 didn’t, but Algol 68 did. The reason for having the concept of an “algorithmic, not a programming language” back in those days was precisely because I/O was such a complicated and machine-dependent mess. To deal with that, the Algol 68 I/O layer description ended up taking a quarter (over 70 pages) of the entire language spec.

POSIX got rid of all that. So nowadays we can assume the existence of a common, fairly simple “stdio” layer on top of whatever the OS or hardware might actually be doing. Any OS that cannot provide such a thing simply isn’t worth using any more.

[toc] | [prev] | [next] | [standalone]

#51024 — Re: Your definition of "Forth" is ...? - was [Re: Hard Forth]

lawrencedo99@gmail.com wrote:
> 
> POSIX got rid of all that. So nowadays we can assume the existence
> of a common, fairly simple 'stdio' layer on top of whatever the OS
> or hardware might actually be doing. Any OS that cannot provide such
> a thing simply isn't worth using any more.

Umm, what?  This is comp.lang.forth.  A brilliant language for small
embedded systems.  An order of magnitude or two smaller than POSIX.
It doesn't need to run on top of a heavyweight OS.

Andrew.

[toc] | [prev] | [next] | [standalone]

Page 3 of 4 — ← Prev page 1 2 [3] 4  Next page →

Back to top | Article view | comp.lang.forth

csiph-web