Re: Ill-considered complaints [was: RE: early networking (was Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...))]

[Liam Proven]

On 15 September 2016 at 01:28, Rich Alderson
 wrote:
> Any decent newsreader or threading mail
> reader knows how to deal with that, and threading is unbroken.


Would that this were true.

Of course, many would say that Gmail is not a decent MUA; however,
it's the best for my needs these days. Every subject edit creates a
new thread in it, and it's very confusing.

-- 
Liam Proven • Profile: http://lproven.livejournal.com/profile
Email: lpro...@cix.co.uk • GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lpro...@hotmail.com • Skype/AIM/Yahoo/LinkedIn: liamproven
Cell/Mobiles: +44 7939-087884 (UK) • +420 702 829 053 (ČR)

Re: Ill-considered complaints [was: RE: early networking (was Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...))]

[geneb]

On Wed, 14 Sep 2016, Rich Alderson wrote:


If you want to change a subject please start a new thread, and if you
wish you can give the new thread a subject line such as "New Subject (was
Old Subject)" to reflect its origin.


Actually, Mr. Cook, the standard for the last 35 years or so has been to


*grabs popcorn*

g.

--
Proud owner of F-15C 80-0007
http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.
Some people collect things for a hobby.  Geeks collect hobbies.

ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.
http://scarlet.deltasoft.com - Get it _today_!

Re: early networking (was Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...))

[Liam Proven]

On 14 September 2016 at 19:12, j...@cimmeri.com  wrote:
> Cool.  I was a big fan of running Netware over Token Ring.   But remember
> eventually
> just getting crushed by cheap and easier to install ethernet.   One of my
> main clients at
> the time was on 4mb Token, and we were asked for a proposal to speed it up.
> 16mb Token Ring had just come out, and the per-card cost was very high.
> Another
> vendor proposed with 10mbit ethernet and stole the client... despite them
> having to
> ditch the expensive, genuine IBM 4mb Token setup (whose wiring could have
> still
> been used for 16mb) and rewire the place.   Still bugs me to this day, which
> is
> probably why I'm writing about it now.  :-)


I saw a few "BrokenString" deployments, but it always was very
expensive, relatively speaking, and except on very heavily-loaded
networks, it was slower, so even at the turn of the '90s, in my world,
it was being replaced by Ethernet unless IBM big iron connectivity was
required.

I've also, just the once, removed an ATM network adaptor from a
desktop PC. Never saw it in real life, but that client had just
returned from Singapore where their home broadband was delivered over
ATM.

-- 
Liam Proven • Profile: http://lproven.livejournal.com/profile
Email: lpro...@cix.co.uk • GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lpro...@hotmail.com • Skype/AIM/Yahoo/LinkedIn: liamproven
Cell/Mobiles: +44 7939-087884 (UK) • +420 702 829 053 (ČR)

Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...)

[Cameron Kaiser]

> So unless there was only a single *nix machine on campus there would
> be NFS.

When I was a University of California student, the news spool and
(depending on the system) the mail spool were both NFS mounts on the
system I used -- sdcc12 and sdcc13/17, respectively, for any other
UCSD students who want to reminisce.

Some of the freenets used, and still use, NFS for a similar purpose,
and the university I later worked for had a couple Sun systems that
did this for specific departments although it was primarily a NetWare
shop initially.

That said, on my own systems, the only NFS server is set up so it can
be a swap device for the Dreamcast. ;)

-- 
 personal: http://www.cameronkaiser.com/ --
  Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckai...@floodgap.com
-- One can live in Paris ... on just grief and anguish. -- Henry Miller ---

Re: early networking (was Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...))

[Liam Proven]

On 14 September 2016 at 18:15, tony duell  wrote:
>> > * LittleBigLAN(never heard of or saw)
>> > * The $25 Network (never heard of or saw)
>>
>> Odd... They were sold in the UK as being American imports...
>
> Dare I suggest that perhaps they flopped in the states so they
> tried to flog them to us :-)

I share your cynicism in general on that point, but this was sold for
_years_ by, IIRC, a little company called EQ Consultants, IIRC.

http://eqc.co.uk/

It was definitely a thing when Ethernet was still too expensive.
Here's a mini-review:

http://www.atarimagazines.com/compute/issue142/70_Getting_wired.php

> I've never seen it in operation, but the Gemini
> Galaxy (somewhat based on the Nascom, with
> the same bus) had a network option. It was a little
> board that hung off the parallel connector on the
> CPU board. The one I have had had all the numbers
> scratched off the ICs, it took me about 2 minutes
> to realise that the main 40 pin IC was a dumb
> UART. The rest of the board was a bit of logic
> to interface it to the parallel port, a clock
> generator and RS485 buffers.

Interesting. I heard of it, certainly, but I never knew of anyone who
actually used it.

> Of course the common network in UK schools
> in the early 80s was Econet (Acorn's network
> for the BBC micro, Atom, etc).

I saw quite a few decommissioned machines with Econet adaptors, and
I've seen a demo network set up at a show in the last decade I think,
but I don't think I ever saw one in action.

I did see IEE-488 in use, both on CBM PETs and BBC Micros, in
education -- both for storage and for connecting to lab equipment.


-- 
Liam Proven • Profile: http://lproven.livejournal.com/profile
Email: lpro...@cix.co.uk • GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lpro...@hotmail.com • Skype/AIM/Yahoo/LinkedIn: liamproven
Cell/Mobiles: +44 7939-087884 (UK) • +420 702 829 053 (ČR)

RE: More Weird Stuff

[tony duell]

> The VT240 is interesting, but a bit large.  Does anyone have any good

Having brought one home (some years ago) on public transport, I found
the VT240 to be easier to move than many other terminals.

Given that there's a T11 processor in there, I do think it's a pity there's
no way to run your own programs on the terminal

-tony

Re: More Weird Stuff

[Ethan Dicks]

On Thu, Sep 15, 2016 at 12:39 AM, Chris Hanson
 wrote:
> ... at Weird Stuff a couple days ago...
>
> Also a bunch of DEC stuff including a VT240 base and several keyboards.

The VT240 is interesting, but a bit large.  Does anyone have any good
programs that exercise the graphics/color?  For my own uses, RT-11 or
VMS preferred.

-ethan

Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...)

[Tor Arntsen]

On 15 September 2016 at 11:43, Liam Proven  wrote:
> On 15 September 2016 at 09:30, Tor Arntsen  wrote:

>> A bit like not noticing
>> that the USB stick runs Linux.. which happens.
>
> Er. Explain? How can a dumb storage device run any OS?
>
> I have various bootable USB sticks around the place, all with Linux
> on, but they don't _run_ it. I won't touch the "smart Wifi enabled"
> ones.

Sorry, yes, not the dumbest mass-storage ones, but the slightly
smarter ones, e.g. wi-fi enabled sticks and SD cards. I have a couple
of those. On the other hand the electronics of even "simple" USB
gadgets is getting pretty sophisticated, and I wouldn't be at all
surprised if a mass-storage USB stick turns out to actually run an OS.

Re: Tips for getting to CHM from SJC airport without a car?

[Al Kossow]

On 9/14/16 10:32 PM, Sam O'nella wrote:
Any tips, tricks or
> warnings?
> 

Take a cab. It's $25 and you'll be there in 15 mins off
traffic peak.

That's what I do when coming back

Public transit to SJC is a sick joke. Billions going
into BART and there are no plans to connect it to the airport.

Re: Linux at 25

[Chuck Guzis]

On 09/15/2016 07:54 AM, tony duell wrote:

> My thermostat contains about 2 dozen parts, even if you count every
> nut , bolt, and washer. It does the job and is not hard to understand
> or repair if/when it needs it.
> 
> Quite why I would want a thermostat with presumably several million
> components, running a multi-user operating system is, to be honest,
> beyond me.

(I don't mean to divert the thread stream; this is an explanation of why
an IOT-connected unit wound up on my wall)

It was a necessary (according to the installer) move as part of having
my home's 24-year-old heat pump replaced.  In that period, the US EPA
has been in the picture very actively.   The old R-22 refrigerant units
are becoming a liability with ever-stricter restrictions on replacement
refrigerant and technical certifications for handling.  By 2020, dealing
with them will be nigh impossible.  The nature of failure of my own
system was such that repairing it was impractical.

The thermostat cable for the old one was the standard 7 wire hookup; the
new system uses (IIRC) 9.  Given the  expense of pushing a new cable
through  finished walls and ceilings, a simple cable upgrade would have
been prohibitively expensive.  So a new 2-wire thermostat was employed
instead (at the installer's expense) and it has WiFi, Web and Bluetooth
connectivity as part of the package.  Fortunately, all of the
aforementioned can be disabled via appropriate selection on the (color)
LCD graphic touchscreen.

The previous heat pump started out with the usual mercury-switch bimetal
thermostat--at some point I upgraded to a programmable electronic one
and hoped to keep it.

As an aside, I just about fainted when I saw the controller electronics
for the heat pump.  Boards full of SMT; lots of  it.  On the other hand,
the unit does boast a lot of efficiency improvements and is very quiet.
What with EPA regulations now, the old "builder's model"
no-frills-barely-does-the-job units seem to be a thing of the past.

--Chuck

Re: Linux at 25

[Cameron Kaiser]

> >  I was a bit surprised to
> > find that my home thermostat was running BusyBox.
> 
> My thermostat contains about 2 dozen parts, even if
> you count every nut , bolt, and washer. It does the job
> and is not hard to understand or repair if/when it needs
> it.

I intentionally kept the old T87-type thermostat when I bought this
house because I understood its properties well and it was completely
predictable.

My sister, against my request, bought me a Nest. It's still in the
box. I think it will make a dandy paperweight.

-- 
 personal: http://www.cameronkaiser.com/ --
  Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckai...@floodgap.com
-- /etc/motd: /earth is 98% full. please delete anyone you can. ---

RE: Linux at 25

[tony duell]

>  I was a bit surprised to
> find that my home thermostat was running BusyBox.

My thermostat contains about 2 dozen parts, even if
you count every nut , bolt, and washer. It does the job
and is not hard to understand or repair if/when it needs
it.

Quite why I would want a thermostat with presumably
several million components, running a multi-user operating
system is, to be honest, beyond me.

-tony

Re: 68K Macs with MacOS 7.5 still in production use...

[Peter Corlett]

On Mon, Sep 12, 2016 at 01:40:56PM -0700, Chris Hanson wrote:
> On Sep 12, 2016, at 2:21 AM, Peter Corlett  wrote:
[...]
>> The 68020 onwards made the CPU fully 32 bit, although various bits of legacy
>> 16 bit cruft remained for compatibility.
> No, the 68000 was a 32-bit CPU, as defined by the register width and
> programming model. The fact that it was implemented with a 16-bit ALU and had
> a 16-bit data path to memory is immaterial.

By that logic, the Z80 is a 16 bit processor because ADC HL, BC and the like
exist. It even has a 4 bit ALU and passes data through twice, but this is an
invisible implementation detail and it's generally considered to be an 8 bit
CPU. Likewise, the existence of zmm0-zmm31 registers don't mean that
contemporary x86 is 512 bit.

The essential 16 bitness of the 68000 pokes through in a lot of places despite
the register file giving the illusion of 32 bitness. Instructions are 16 bit
aligned. The status register is 16 bits. There are no true 32 bit multiply or
divide instructions. Shift/rotate instructions to memory can only be used on 16
bit quantities. MOVEP exists.

TMB11 drawings?

[Noel Chiappa]

Does anyone know of the whereabouts of a set of engineering drawings for the
TMB11 (also a Technical Manual, although that's more of a luxury)? All I could
find on it, online, was the Operator's Manual.

 Noel


PS: In an older thread on TU10's/TM11's here:

  http://www.classiccmp.org/pipermail/cctalk/2015-September/011810.html

I saw some queries about whether a TU10 could be connected to a TMB11. The
answer is apparently 'yes', for two reasons:

First, I found docs on a thing called a TMA11 (apparently intermediate between
the TM11 and TMB11), and one version of the docmentation about it talks about
the TMA11 and the TU10, but another version talks about the TMA11 and the
TS03. So, by transitivity, if the TU10 works with a TMA11, and a TMA11 works
with a TS03, and the TS03 works with a TMB11, the TMB11 must work with a
TU10...

Second, I have a report of a TU10 found plugged into a TMB11 in a retired
computer.

Re: 68K Macs with MacOS 7.5 still in production use...

[Chuck Guzis]

On 09/15/2016 11:03 AM, Peter Corlett wrote:
> On Mon, Sep 12, 2016 at 01:40:56PM -0700, Chris Hanson wrote:

>> No, the 68000 was a 32-bit CPU, as defined by the register width
>> and programming model. The fact that it was implemented with a
>> 16-bit ALU and had a 16-bit data path to memory is immaterial.
> 
> By that logic, the Z80 is a 16 bit processor because ADC HL, BC and
> the like exist. It even has a 4 bit ALU and passes data through
> twice, but this is an invisible implementation detail and it's
> generally considered to be an 8 bit CPU. Likewise, the existence of
> zmm0-zmm31 registers don't mean that contemporary x86 is 512 bit.

My 68K manual, as handed to me by the Moto sales guy at Wescon back in
the day says "MC68000 16-BIT MICROPROCESSOR User's Manual".   Page 1-1
reiterates essentially the same sentiment:

"The MC68000...combines state-of-the-art technology and advanced circuit
design techniques with computer sciences to achieve an architecturally
advanced 16-bit microprocessor."

Call it anything you want, but we know what Motorola called it.

--Chuck

Re: 68K Macs with MacOS 7.5 still in production use...

[Norman Jaffe]

And, to 'put a nail in it', the bitsavers file for the MC68000 is 
'68000_16-Bit_Microprocessor_Apr83.pdf'. 
- Original Message -

From: "Chuck Guzis"  
To: "General Discussion: On-Topic and Off-Topic Posts"  
Sent: Thursday, September 15, 2016 11:24:11 AM 
Subject: Re: 68K Macs with MacOS 7.5 still in production use... 

On 09/15/2016 11:03 AM, Peter Corlett wrote: 
> On Mon, Sep 12, 2016 at 01:40:56PM -0700, Chris Hanson wrote: 

>> No, the 68000 was a 32-bit CPU, as defined by the register width 
>> and programming model. The fact that it was implemented with a 
>> 16-bit ALU and had a 16-bit data path to memory is immaterial. 
> 
> By that logic, the Z80 is a 16 bit processor because ADC HL, BC and 
> the like exist. It even has a 4 bit ALU and passes data through 
> twice, but this is an invisible implementation detail and it's 
> generally considered to be an 8 bit CPU. Likewise, the existence of 
> zmm0-zmm31 registers don't mean that contemporary x86 is 512 bit. 

My 68K manual, as handed to me by the Moto sales guy at Wescon back in 
the day says "MC68000 16-BIT MICROPROCESSOR User's Manual". Page 1-1 
reiterates essentially the same sentiment: 

"The MC68000...combines state-of-the-art technology and advanced circuit 
design techniques with computer sciences to achieve an architecturally 
advanced 16-bit microprocessor." 

Call it anything you want, but we know what Motorola called it. 

--Chuck 

Re: 68K Macs with MacOS 7.5 still in production use...

[Noel Chiappa]

> From: Chuck Guzis

> Call it anything you want, but we know what Motorola called it.

The _first implementation_ may have been 16-bit, but I am in no doubt
whatsover (having written a lot of assembler code for the 68K family)
that the _architecture_ was 32-bit:

- 32-bit registers
- many operations (arithmetical, logical, etc) defined for that length
- 32-bit addresses

Etc, etc, etc, etc.

Noel

Re: TMB11 drawings?

[Al Kossow]

On 9/15/16 10:29 AM, Noel Chiappa wrote:
> Does anyone know of the whereabouts of a set of engineering drawings for the
> TMB11


http://www.computerhistory.org/collections/catalog/102753063

Re: 68K Macs with MacOS 7.5 still in production use...

[Chuck Guzis]

On 09/15/2016 11:38 AM, Noel Chiappa wrote:
>> From: Chuck Guzis
> 
>> Call it anything you want, but we know what Motorola called it.
> 
> The _first implementation_ may have been 16-bit, but I am in no
> doubt whatsover (having written a lot of assembler code for the 68K
> family) that the _architecture_ was 32-bit:
> 
> - 32-bit registers - many operations (arithmetical, logical, etc)
> defined for that length - 32-bit addresses
> 
> Etc, etc, etc, etc.

Hence my comment.  It's a matter of what to believe--Motorola or your
lyin' eyes.  :)

Clearly, external bus size doesn't mean much in this discussion.  Take
the NS32K series--from the -008 to the  -032, all basically the same
internally.  For what it's worth, WikiP refers to them as the "first
general-purpose 32-bit microcomputer".

There have been a great many bit-serial computers in history,but I've
never heard them called "one bit" architectures.  Register length is
similarly no indication as internal registers can be any length and
"visible" registers may not exist at all.

'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it
means just what I choose it to mean — neither more nor less.'

--Chuck

Re: TMB11 drawings?

[Al Kossow]

just took a look at this
the TMB11 is a special widget for the small Kennedy 7" 800bpi tape drive

On 9/15/16 11:41 AM, Al Kossow wrote:
> 
> 
> On 9/15/16 10:29 AM, Noel Chiappa wrote:
>> Does anyone know of the whereabouts of a set of engineering drawings for the
>> TMB11
> 
> 
> http://www.computerhistory.org/collections/catalog/102753063
> 

Re: TMB11 drawings?

[Al Kossow]

but we do have the TMA11 drwngs
http://www.computerhistory.org/collections/catalog/102753071

On 9/15/16 12:38 PM, Al Kossow wrote:
> just took a look at this
> the TMB11 is a special widget for the small Kennedy 7" 800bpi tape drive
> 
> On 9/15/16 11:41 AM, Al Kossow wrote:
>>
>>
>> On 9/15/16 10:29 AM, Noel Chiappa wrote:
>>> Does anyone know of the whereabouts of a set of engineering drawings for the
>>> TMB11
>>
>>
>> http://www.computerhistory.org/collections/catalog/102753063
>>
> 

Re: 68K Macs with MacOS 7.5 still in production use...

[Pete Turnbull]

On 15/09/2016 20:34, Chuck Guzis wrote:

On 09/15/2016 11:38 AM, Noel Chiappa wrote:

From: Chuck Guzis



Call it anything you want, but we know what Motorola called it.


The _first implementation_ may have been 16-bit, but I am in no
doubt whatsover (having written a lot of assembler code for the 68K
family) that the _architecture_ was 32-bit:

- 32-bit registers - many operations (arithmetical, logical, etc)
defined for that length - 32-bit addresses



Hence my comment.  It's a matter of what to believe--Motorola or your
lyin' eyes.  :)


FWIW, when it came out, and for some years after, we were all talking 
about "Motorola's (new) 16-bit micro", and when I attended a later 
ComputerSci course on 68000 architecture and programming in the 1990s 
the family was still treated as 16-bit until the 68020.  The first THREE 
implementations are 16-bit as far as I'm concerned.


--
Pete
Pete Turnbull

Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[Toby Thain]

On 2016-09-15 2:38 PM, Noel Chiappa wrote:

> From: Chuck Guzis

> Call it anything you want, but we know what Motorola called it.

The _first implementation_ may have been 16-bit, but I am in no doubt
whatsover (having written a lot of assembler code for the 68K family)
that the _architecture_ was 32-bit:

- 32-bit registers
- many operations (arithmetical, logical, etc) defined for that length
- 32-bit addresses


GPR width, being the visible programmer model, is the most common and 
convenient definition of "architecture" I've come across. But there's no 
reason we can't just say the *visible* architecture is 32 bit (which it 
is), but the "internal" architecture is sort of 16.


Why not have two concepts? They're free. Gordon Bell probably laid this 
out somewhere. Or Blaauw and Brooks.


--Toby



Etc, etc, etc, etc.

Noel

Re: Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[Cameron Kaiser]

> Also, Apple Computer referred to [the 68000] as a 32 bit microprocessor in
> their early Macintosh ads.

And Apple *never* oversells *anything.* ;)

-- 
 personal: http://www.cameronkaiser.com/ --
  Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckai...@floodgap.com
-- He whose face gives no light, shall never become a star. -- William Blake --

Re: 68K Macs with MacOS 7.5 still in production use...

[Cameron Kaiser]

> Few, if any processors could be unambiguously classified.

Some can. The TMS 9900 is indisputably 16-bit; it even has 16-bit memory
addressing, in addition to 16-bit registers, a 16-bit ALU and 16-bit
internal datapaths. For that matter, the CRU is 16-bit, too.

The PDP-8 and Intersil 6100 had 12-bit registers and 12-bit addressing,
at least natively.

If you ignore address bus size, lots of the 8-bits are clearly 8-bit, like
the 6502 and the Signetics 2650, which have 8-bit registers and 8-bit ALUs
and no doubling or splitting. I'm not aware of an 8-bit with an 8-bit address
bus [that isn't multiplexed], though I'm sure there's some wacky
microcontroller variants around that qualify.

And then there's the MC14500B ...

I think this only starts getting blurry when you get above the 8-bit
threshold.

> Which part of the elephant are we looking at?

"No, honey, that's *not* his trunk."

-- 
 personal: http://www.cameronkaiser.com/ --
  Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckai...@floodgap.com
-- Jesus loves you, and I'm trying to. -- Jack Thompson ---

Re: Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[Chuck Guzis]

On 09/15/2016 05:17 PM, j...@cimmeri.com wrote:

> Also, Apple Computer referred to it as a 32 bit microprocessor in
> their early Macintosh ads.   I always just considered it a 32 bit CPU
> with a 16 bit external bus.

So what's the width of an IBM 1620?

--Chuck

Re: TMB11 drawings?

[Al Kossow]

On 9/15/16 3:46 PM, Al Kossow wrote:
> 
> 
> On 9/15/16 1:42 PM, Noel Chiappa wrote:
>> > From: Al Kossow 
>>
>> >  http://www.computerhistory.org/collections/catalog/102753063
>>
>> Ah, excellent! Any chance those can be scanned at some point?
>>
> 
> I'll try to take care of both this evening
> 
> 

done. the mirrors should pick them up by tomorrow morning

Re: Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[Kyle Owen]

On Thu, Sep 15, 2016 at 8:07 PM, Cameron Kaiser 
wrote:

> > Also, Apple Computer referred to [the 68000] as a 32 bit microprocessor
> in
> > their early Macintosh ads.
>
> And Apple *never* oversells *anything.* ;)
>

How about 65k of RAM?!

http://tr3.cbsistatic.com/hub/i/2011/04/06/5174da48-c3aa-11e2-bc00-02911874f8c8/2e9edc467174c362dd543d51b886196a/02-IMSAI_1977.jpg

That's like...a whole 1k more than the competition!

Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[j...@cimmeri.com]

On 2016-09-15 2:38 PM, Noel Chiappa wrote:

> From: Chuck Guzis

> Call it anything you want, but we know what Motorola called it.

The _first implementation_ may have been 16-bit, but I am in no doubt
whatsover (having written a lot of assembler code for the 68K family)
that the _architecture_ was 32-bit:

- 32-bit registers
- many operations (arithmetical, logical, etc) defined for that length
- 32-bit addresses



Also, Apple Computer referred to it as a 32 bit microprocessor in their 
early Macintosh ads.   I always just considered it a 32 bit CPU with a 
16 bit external bus.


- J.

Re: Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[Paul Koning]

> On Sep 15, 2016, at 5:57 PM, Toby Thain  wrote:
> 
> On 2016-09-15 2:38 PM, Noel Chiappa wrote:
>>> From: Chuck Guzis
>> 
>>> Call it anything you want, but we know what Motorola called it.
>> 
>> The _first implementation_ may have been 16-bit, but I am in no doubt
>> whatsover (having written a lot of assembler code for the 68K family)
>> that the _architecture_ was 32-bit:
>> 
>> - 32-bit registers
>> - many operations (arithmetical, logical, etc) defined for that length
>> - 32-bit addresses
> 
> GPR width, being the visible programmer model, is the most common and 
> convenient definition of "architecture" I've come across. But there's no 
> reason we can't just say the *visible* architecture is 32 bit (which it is), 
> but the "internal" architecture is sort of 16.

So would you call a PDP-8/S a one bit machine?  I suppose you could, but that 
seems rather odd.

paul

Re: TMB11 drawings?

[Al Kossow]

On 9/15/16 1:42 PM, Noel Chiappa wrote:
> > From: Al Kossow 
> 
> >  http://www.computerhistory.org/collections/catalog/102753063
> 
> Ah, excellent! Any chance those can be scanned at some point?
> 

I'll try to take care of both this evening

Need help with DSD-440 QBus Controller

[Julian Wolfe]

Hi folks,



I recently acquired a DSD-440 drive and purchased its accompanying controller 
on ebay.  The controller is configured at defaults according to the manual on 
Bitsavers.  However, unless it is on the bus by itself after the RAM cards, it 
halts the CPU at location 270.



Here are the 4 card configurations I tried,  or  denotes direction of 
serpentine QBus:



Doesn't work:


1 CPU - CPU - CPU - CPU 

2 RAM - RAM - RAM - RAM 

3 RAM - RAM - RAM - RAM 

4 RAM - RAM - RAM - RAM 

5 RAM - RAM - RAM - RAM 

6 DSD - DSD - RAM - RAM 

7 UC07-UC07- --DEQNA-- 

8 --- -  --- - --- - --- - --- 

9 --- -  --- - --- - --- - --- 



OR


1 CPU - CPU - CPU - CPU 

2 RAM - RAM - RAM - RAM 

3 RAM - RAM - RAM - RAM 

4 RAM - RAM - RAM - RAM 

5 RAM - RAM - RAM - RAM 

6 RAM - RAM - --- - --- 

7 DSD - DSD - --- - --- 
8 UC07-UC07- --DEQNA-- 

9 --- -  --- - --- - --- - --- 





Works (or at least leaves the CPU in the "RUN" state):



1 CPU - CPU - CPU - CPU 

2 RAM - RAM - RAM - RAM 

3 RAM - RAM - RAM - RAM 

4 RAM - RAM - RAM - RAM 

5 RAM - RAM - RAM - RAM 

6 DSD - DSD - RAM - RAM 

7 --- -  --- - --- - --- - --- 

8 --- -  --- - --- - --- - --- 

9 --- -  --- - --- - --- - --- 



OR



1 CPU - CPU - CPU - CPU 

2 RAM - RAM - RAM - RAM 

3 RAM - RAM - RAM - RAM 

4 RAM - RAM - RAM - RAM 

5 RAM - RAM - RAM - RAM 

6 RAM - RAM - --- - --- 

7 DSD - DSD - --- - --- 

8 --- -  --- - --- - --- - --- 

9 --- -  --- - --- - --- - --- 



The card says (C)1978 Data Systems Design on it, and the latest manual is 
(C)1980 - which makes me concerned this is an 18-bit only QBus card.



I have not yet tried attaching the drive box to the card, thinking it would at 
least see the controller without hanging.  The box isn't yet clean and ready to 
go.
 

Any help with this card would be appreciated.  Am I doing something wrong, or 
is this just a bad card? 



Thanks



Julian

RE: Meaning of "architecture width" - Re: 68K Macs with MacOS 7.5 still in production use...

[tony duell]

> Why not have two concepts? They're free. Gordon Bell probably laid this
> out somewhere. Or Blaauw and Brooks.

In a talk on the HP9810 desktop calculator I said, somewhat tongue-in-cheek
that it could claim to be 1, 3, 4, 6, 8 or 16 bits.

The justifications for those are as follows : The processor has 16 bit 
registers.
It's bit serial, so the normal internal ALU and data path width is 1 bit. But 
it 
does BCD additions between the A (accumulator) and T (memory data) registers
a nybble (4 bits) at a time.

The memory system is odd, with hardware translating a read or write into 2 
cycles on successive locations. One location stores the odd bits, the other the
even bits. So the physical width of memory for things like the system stack
is 8 bits.

Even odder, the machine stores user programs as sequences of 6 bit keycodes.
User program memory is logically 6 bits wide. But because of the memory 
control system, such memory is physically 3 bits wide, user program memory
expansion consists of Intel 1103 (1K*1bit) DRAMs fitted in sets of 3. 

So : 
1 bit -- physical width of binary ALU and data paths
3 bits -- physical width of user program memory
4 bits -- physical width of BCD ALU and data paths
6 bits -- logical width of user program memory
8 bits -- physical width of data memory
16 bits -- logical width of data memory, physical width of CPU registers

Call it what you will :-)

-tony

Re: Need help with DSD-440 QBus Controller

[Jerry Weiss]

> On Sep 15, 2016, at 7:51 PM, Julian Wolfe  wrote:
> 
> I recently acquired a DSD-440 drive and purchased its accompanying controller 
> on ebay.  The controller is configured at defaults according to the manual on 
> Bitsavers.  However, unless it is on the bus by itself after the RAM cards, 
> it halts the CPU at location 270.
> 
See below.  Assuming you have a Q-Q backplane here.  If its an Q-CD, full stop.

> Here are the 4 card configurations I tried,  or  denotes direction of 
> serpentine QBus:
> 
> …….
> 
> The card says (C)1978 Data Systems Design on it, and the latest manual is 
> (C)1980 - which makes me concerned this is an 18-bit only QBus card.
> 

Yes this is an 18 bit card, but it will work in a 22 bit Qbus in some OS’es 
(TSX+) with the correct configuration.I.e. it will only move data to the 
between the lower 18bits of memory and media.  

> 
> I have not yet tried attaching the drive box to the card, thinking it would 
> at least see the controller without hanging.  The box isn't yet clean and 
> ready to go.
> 

That is a good approach.   Using ODT make sure the CSR is at 1170 (for an 
11/73) or 777170 for an 11/23 and the status bits appear to be valid.

> 
> Any help with this card would be appreciated.  Am I doing something wrong, or 
> is this just a bad card? 
> 
> 
A halt at 270 is a bit odd, but its only just above the interrupt vector. 
Disable the DSD controller on-board bootstrap.  Without drives and an OS disk, 
its going to halt at some point,

It will conflict with the UC07, if that is strapped for bootstrap as well.   
Then try to boot your normal OS.
If RT11SJ, FB, or SB and the DY handlers are present, the machine should boot 
and the handler loadable.  Without the drives however, accessing them will not 
be useful.

Jerry

Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...)

[Tor Arntsen]

On 14 September 2016 at 16:51, Liam Proven  wrote:
> On 14 September 2016 at 15:59, Tor Arntsen  wrote:
>> On 14 September 2016 at 15:50, Liam Proven  wrote:
>>
>>> To this day, I have never once used any form of NFS or ever seen it in use.
>>
>> A typo, I presume? NFS, as in Network File System?
>>
>> Used, for example, everywhere where Sun boxes were installed, for our
>> (European) company that would be from around 1989 and onwards (and we
>> were a bit late with Sun), and it's going strong, although (almost)
>> without Sun boxes. And common with NAS, of course.
>
>
> No typo. As far as I know, working in tech since 1988, I have never
> ever seen it used.
>
> All Unix types assume that it is very common. In mixed environments --
> which means more or less every one I've ever worked in -- it isn't.

I still find it very strange. Back in 1988 was probably when I first
got to use a Sun machine, and from 1989 we started using more of them,
but it was certainly a mixed environment - it had to be, we had a lot
of infrastructure from earlier. It took quite a few years until it was
all *nix (and *nux, after a while). The thing is that a Sun machine at
the time more often than not didn't have a disk big enough to keep all
the software (and other files), and/or the sysadm found it difficult
to maintain even a small number of workstations or servers
individually. So at that time it was extremely common to have a lot of
the software accessible via NFS.  In other words, an NFS path was
usually part of the computer user's $PATH.

So unless there was only a single *nix machine on campus there would
be NFS. I saw the same thing in other local companies and in other
European countries, e.g. Italy, which I visited extensively back then.
Companies would have different environments, but most of them would
have a small number of Unix servers and maybe a single workstation,
the rest would be terminals. And later Windows PCs. The small number
of Unix boxes would in any case always use NFS in one way or another.

Back to 2016. In an environment with mixed Unix, Linux and Windows
computers, the shared documents are simply made available via NFS for
*nix and Samba for Windows. From the same server(s). Been that way
forever by now. If there's a single *nix box which needs anything from
somewhere else then NFS is the natural choice, mixed environment or
not.

In any case I'm absolutely certain that as far as NFS is concerned
there couldn't and can't possibly be any difference in usage between
Europe and the US.  I spent so many years travelling between various
sites on both sides of the Atlantic pre- and post-2k, and I never
noticed anything different there. I actually suspect that you *have*
seen NFS in use, you just didn't notice it. A bit like not noticing
that the USB stick runs Linux.. which happens. And various other more
or less surprising places.

Re: TMB11 drawings?

[Noel Chiappa]

> From: Al Kossow 

>  http://www.computerhistory.org/collections/catalog/102753063

Ah, excellent! Any chance those can be scanned at some point?

(No rush, I'm not about to start working with one instantly - too much else
backed up in the queue! :-)

>> I saw some queries about whether a TU10 could be connected to a TMB11.
>> The answer is apparently 'yes', for two reasons:
>> First, I found docs on a thing called a TMA11 (apparently intermediate
>> between the TM11 and TMB11), and one version of the docmentation about
>> it talks about the TMA11 and the TU10, but another version talks about
>> the TMA11 and the TS03. So, by transitivity, if the TU10 works with a
>> TMA11, and a TMA11 works with a TS03, and the TS03 works with a TMB11,
>> the TMB11 must work with a TU10...
>> Second, I have a report of a TU10 found plugged into a TMB11 in a
>> retired computer.

> the TMB11 is a special widget for the small Kennedy 7" 800bpi tape drive

That's the TS03, right?

Anyway, there are indications (above) that it will also work with a TU10. If
the TMB11 prints get scanned in, I can take a look them, and see what gives.

Further (third) clue: the TMB11 Ops manual says "The TS03 tape transport
operates at only one density (800 bpi) and iin only one mode of operation (9
track). The TMB11 is capable of other densities and can operate in the
7-track .. modes." I expect this is to support these modes in the TU10...

I originally thought the TB11 was TS03-specific, but after reviewing all the
above, I have changed my mind. Having the drawings would be great; I could
check them out to confirm that it really can drive a TU10.


> we do have the TMA11 drwngs

Also excellent! The TM11 ones are online, but not (AFAIK) the TMA11. So if
that could get done at some point, too... :-)

The two are very similar (a 19" rack backplane full of smaller FLIP CHIPs),
so it's not like the RK11-C -> RK11-D, where they re-implemented it to make
it cheaper. I'm _guessing_ the latter one can handle 1600 bpi, or some such,
but with the prints, the difference could be confirmed.


> From: Henk Gooijen

> IIRC, I have the printset of the TMB11 and ISTR it is one inch thick!

Hmm. No idea why - it's only a quad card and a hex card and 4 smaller
standard UNIBUS FLIP CHIPs (M105, M7821, etc). Hard to see that generating 1"
of paper (even with the wire list for the custom backplane - a hex-high
system unit).

Maybe that set includes the TS03 drawings too?

Noel

Re: 68K Macs with MacOS 7.5 still in production use...

[Fred Cisin]

family) that the _architecture_ was 32-bit:

On Thu, 15 Sep 2016, Chuck Guzis wrote:

Hence my comment.  It's a matter of what to believe--Motorola or your
lyin' eyes.  :)


Ah, but can the manufacturer be trusted?
What would motivate them to take a 32 bit processor and CALL it 16 bit?
Was that a marketing decision?  "We can call it a good competitive 32 bit 
processor, or we can make an indisputable claim that it is the BEST 16 bit 
processor!"



'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it
means just what I choose it to mean ? neither more nor less.'


There are more than a hundred different parameters involved.
There is no definitive agreement as to which parameter is to be used for 
the classification, nor even which parameterS.
Whenever somebody has the efrontery to build a product that has one 
classification by one parameter, and a different classification by another 
parameter, there will be disputed classifications, since different 
parameters are more or less important to each viewer.


Few, if any processors could be unambiguously classified.
What size is a
Z80?
8088?
8086?
80286?
6809?
Everybody here can agree that that is a stupid question, 
because the list of answers is "OBVIOUS".  And yet a comparison of 
answers WILL have discrepancies.


Register size would be an obvious one.  But, as soon as provision is made 
for accessing half registers and double ones, it is open to variant 
interpretations even of what the register size is.


Should we classify them, instead by their heat output?


Which part of the elephant are we looking at?


--
Grumpy Ol' Fred ci...@xenosoft.com

RE: TMB11 drawings?

[Henk Gooijen]

I have the TMB11 and the TS03. IIRC, I have the printset of the TMB11

and ISTR it is one inch thick!

Will check this weekend.



-  Henk





Van: Al Kossow
Verzonden: donderdag 15 september 2016 21:37
Aan: cctalk@classiccmp.org
Onderwerp: Re: TMB11 drawings?



just took a look at this
the TMB11 is a special widget for the small Kennedy 7" 800bpi tape drive

On 9/15/16 11:41 AM, Al Kossow wrote:
>
>
> On 9/15/16 10:29 AM, Noel Chiappa wrote:
>> Does anyone know of the whereabouts of a set of engineering drawings for the
>> TMB11
>
>
> http://www.computerhistory.org/collections/catalog/102753063
>

Re: G4 cube (was Re: 68K Macs with MacOS 7.5 still in production use...)

[Liam Proven]

On 15 September 2016 at 09:30, Tor Arntsen  wrote:
> I still find it very strange.

Yes, that was the point of my post!

> So unless there was only a single *nix machine

Virtually every *nix deployment I have ever worked upon, yes, the was
one, single *nix machine in the site/building/company.
Between 2000 and 2005, in most instances, it was an old PC running Smoothwall.

It has always been, in my working life, very very rare.

One AIX box, ever.
One AS/400, ever.
One S/36, ever.
One PDP-11, ever.
I don't think I've ever had a client who had a connect, working Sun
workstation. A handful of servers in 1 job, but I never touched them.

> Back to 2016. In an environment with mixed Unix, Linux and Windows
> computers, the shared documents are simply made available via NFS for
> *nix and Samba for Windows. From the same server(s).

This might surprise you, but in my own home network, I used Samba for
connections from Linux to Linux. It was easier.

> In any case I'm absolutely certain that as far as NFS is concerned
> there couldn't and can't possibly be any difference in usage between
> Europe and the US.

I suspect that there can. That in what I am now told are not SMEs
after all, but "microbusinesses", the bulk of my career, Sun kit was
always just too expensive to be worth it. Windows is utterly, totally
dominant, with these days some Macs and a tiny bit of Linux. Linux is
usually in niche roles, such as firewalls, web proxies, DNS servers or
caches, occasionally as NAS hosts. NFS does not enter into the
equation, at all, ever.

Obviously -- well, it's obvious to me, but I'm spelling it out because
I find that people make assumptions... Obviously, I am not claiming
that my experience is universal or general or applies to anyone else.
Others will have had different experiences.

However, I have had a fairly varied career in IT. I've worked for
VARs, for shops, in repair, in systems design, in sysadmin, in 1st,
2nd & 3rd line support, I've done a smattering of development work
early on, I've worked in training, I've worked for end users, I've
been a consultant, I've been a journalist and an editor and a
technical writer. I've worked for software houses, charities,
government divisions, multinationals, banks, stockbrokers, investment
houses.

I've done a lot of stuff over 28 years in IT.

And I have never seen NFS in use, anywhere, ever.

I guess it figures that the thing I consider massively marginal and
niche is the one that provoked the reaction of surprise and disbelief
that I was actually looking for and trying to cause.

> A bit like not noticing
> that the USB stick runs Linux.. which happens.

Er. Explain? How can a dumb storage device run any OS?

I have various bootable USB sticks around the place, all with Linux
on, but they don't _run_ it. I won't touch the "smart Wifi enabled"
ones.

-- 
Liam Proven • Profile: http://lproven.livejournal.com/profile
Email: lpro...@cix.co.uk • GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lpro...@hotmail.com • Skype/AIM/Yahoo/LinkedIn: liamproven
Cell/Mobiles: +44 7939-087884 (UK) • +420 702 829 053 (ČR)

Re: Tips for getting to CHM from SJC airport without a car?

[Corey Cohen]

corey cohen
uǝɥoɔ ʎǝɹoɔ
> On Sep 15, 2016, at 1:32 AM, Sam O'nella  wrote:
> 
> A bit off topic other than tips for anyone else trying to travel cheaply. I
> have a trip to California in a month although the final destination isn't
> San Jose.
> 
> I can fly in to SJC and Google is sounding like I might be able to get a
> few bus hops from the airport and find my way there? Any tips, tricks or
> warnings?
> 
> *(more off-topic below, feel free to stop above for helping anyone venture
> around to CHM).
> No idea if I'm testing my luck neighborhood wise and debating taking a bus
> or train from Mountain view to Pasadena after my free day.
> 
> Feel free to email me offline if this is too far from list interest. Mostly
> if anyone was to benefit it would be what's the cheapest route to spend a
> day at the CHM and then return to the airport again.
> 
> - John

Your better off using Uber or lift.  Taking a bus might eat up all your time at 
CHM. An uber or lift if it's not rush hour will only take 20 minutes max to 
CHM.  If you have a little more time and want to save money use Uber pool.  

Cheers,
Corey

Re: Subjects, Topics and Threading

[Dale H. Cook]

At 06:40 PM 9/14/2016, KnoppixLiveKiller wrote:

>Is it really that hard to bottom post?

With that things have gotten really ridiculous, so goodbye for good.

Dale H. Cook, Contract IT Administrator, Roanoke/Lynchburg, VA
http://plymouthcolony.net/starcityeng/index.html