f



Auction: Vintage IBM 5362 System 36 Computer System S/36 (LA local pickup only)

BIN http://search.ebay.com/200344827100
Santa Cruz, California
0
Louis
11/30/2009 1:52:34 AM
comp.sys.ibm.ps2.hardware 10363 articles. 0 followers. Post Follow

9 Replies
584 Views

Similar Articles

[PageSpeed] 14

On Nov 29, 8:52=A0pm, Louis Ohland <ohl...@charter.net> wrote:
> BIN http://search.ebay.com/200344827100

"Here's your chance to have a small IBM system that's constructed like
their mainframes of years gone by."

That's either literary license, or the seller has no concept of
classic mainframe construction.  The 5362 is built in a manner similar
to the RT/PC or the RS/6000 7013s, even the high-end 5360 was not up
to mainframe construction standards.

Rick Ekblaw
0
RickE
11/30/2009 12:44:18 PM
RickE wrote:
> On Nov 29, 8:52 pm, Louis Ohland <ohl...@charter.net> wrote:
>> BIN http://search.ebay.com/200344827100
> 
> "Here's your chance to have a small IBM system that's constructed like
> their mainframes of years gone by."
> 
> That's either literary license, or the seller has no concept of
> classic mainframe construction.  The 5362 is built in a manner similar
> to the RT/PC or the RS/6000 7013s, even the high-end 5360 was not up
> to mainframe construction standards.
> 
> Rick Ekblaw

It's successor: the white AS/400 had expansion cards that did look a bit 
like the mainframe stuff. My advise would be to look for a more recent 
AS/400 if you like the concept of an all in one sturdy box.

One thing to say about the AS/400 is that it's I/O was generally better 
designed than the contemporary PCs and servers (no idea on how the 
iServers compare in that respect now). One of M$'s NT versions was 
developed with AS/400s as file servers if I'm correct (surprising, given 
that Sun also had a good file serving reputation too).

Now, I'm diverting too much.

It's a nice machine if you care to collect. Finding software might be hard.

Regards,
Alvin.
0
Alvin
11/30/2009 6:10:26 PM
Alvin Andries wrote:
> RickE wrote:
>> On Nov 29, 8:52 pm, Louis Ohland <ohl...@charter.net> wrote:
>>> BIN http://search.ebay.com/200344827100
>>
>> "Here's your chance to have a small IBM system that's constructed
>> like their mainframes of years gone by."
>>
>> That's either literary license, or the seller has no concept of
>> classic mainframe construction.  The 5362 is built in a manner
>> similar to the RT/PC or the RS/6000 7013s, even the high-end 5360
>> was not up to mainframe construction standards.
>>
>> Rick Ekblaw
>
> It's successor: the white AS/400 had expansion cards that did look a
> bit like the mainframe stuff. My advise would be to look for a more
> recent AS/400 if you like the concept of an all in one sturdy box.
>
> One thing to say about the AS/400 is that it's I/O was generally
> better designed than the contemporary PCs and servers (no idea on how
> the iServers compare in that respect now). One of M$'s NT versions was
> developed with AS/400s as file servers if I'm correct (surprising,
> given that Sun also had a good file serving reputation too).
>
> Now, I'm diverting too much.
>
> It's a nice machine if you care to collect. Finding software might be
> hard.

If anyone wants photos of the internals of an AS/400 and the mood strikes me 
I might be persuaded to open mine up.



>
> Regards,
> Alvin. 

0
J
11/30/2009 8:45:25 PM
J. Clarke wrote:
> Alvin Andries wrote:
>> RickE wrote:
>>> On Nov 29, 8:52 pm, Louis Ohland <ohl...@charter.net> wrote:
>>>> BIN http://search.ebay.com/200344827100
>>> "Here's your chance to have a small IBM system that's constructed
>>> like their mainframes of years gone by."
>>>
>>> That's either literary license, or the seller has no concept of
>>> classic mainframe construction.  The 5362 is built in a manner
>>> similar to the RT/PC or the RS/6000 7013s, even the high-end 5360
>>> was not up to mainframe construction standards.
>>>
>>> Rick Ekblaw
>> It's successor: the white AS/400 had expansion cards that did look a
>> bit like the mainframe stuff. My advise would be to look for a more
>> recent AS/400 if you like the concept of an all in one sturdy box.
>>
>> One thing to say about the AS/400 is that it's I/O was generally
>> better designed than the contemporary PCs and servers (no idea on how
>> the iServers compare in that respect now). One of M$'s NT versions was
>> developed with AS/400s as file servers if I'm correct (surprising,
>> given that Sun also had a good file serving reputation too).
>>
>> Now, I'm diverting too much.
>>
>> It's a nice machine if you care to collect. Finding software might be
>> hard.
> 
> If anyone wants photos of the internals of an AS/400 and the mood strikes me 
> I might be persuaded to open mine up.
> 
> 
> 
>> Regards,
>> Alvin. 
> 

What type(s) and model(s) do you have?
- white CISC
- early RISC with still the large cards (I think they used V3R4 or so 
for the transition)
- later RISC with PCI
- iSeries (probably still look the same)

I have 2, but not enough time to do much with them. I did reset the 
password (and already forgot it).
I also tossed 2 whites 5 years ago (a B + expansion and a smaller D with 
built-in adapters). It did hurt (mentaly en physically), but I couldn't 
do anything with them, because:
-> I mixed the adapters and disks because I couldn't lift them 
otherwise, but I forgot to note where wat was located
-> when I got them (for free), I didn't know that finding software would 
be such a disaster
-> when I learned about the battery held license key, I knew I was in 
deep sh*t, especially because V2R1 was non-transferable, already dated 
and "hobby" use is something IBM hadn't heard of


Their total lack on eBay seems to confirm that keeping them up and alive 
is (nearly) impossible.

I must say that it's a shame how IBM is treating these machines: 
software for the S/36 might still run unmodified on the current 
generation (it certainly did till V4R1): imagine running CP/M programs 
under Windows XP :-)
Also, the concept is quite interesting. But then again, IBM did sell 
their PC division and let Apple drop the PowerPC.

Regards,
Alvin.

0
Alvin
11/30/2009 11:10:07 PM
Alvin Andries wrote:
> J. Clarke wrote:
>> Alvin Andries wrote:
>>> RickE wrote:
>>>> On Nov 29, 8:52 pm, Louis Ohland <ohl...@charter.net> wrote:
>>>>> BIN http://search.ebay.com/200344827100
>>>> "Here's your chance to have a small IBM system that's constructed
>>>> like their mainframes of years gone by."
>>>>
>>>> That's either literary license, or the seller has no concept of
>>>> classic mainframe construction.  The 5362 is built in a manner
>>>> similar to the RT/PC or the RS/6000 7013s, even the high-end 5360
>>>> was not up to mainframe construction standards.
>>>>
>>>> Rick Ekblaw
>>> It's successor: the white AS/400 had expansion cards that did look a
>>> bit like the mainframe stuff. My advise would be to look for a more
>>> recent AS/400 if you like the concept of an all in one sturdy box.
>>>
>>> One thing to say about the AS/400 is that it's I/O was generally
>>> better designed than the contemporary PCs and servers (no idea on
>>> how the iServers compare in that respect now). One of M$'s NT
>>> versions was developed with AS/400s as file servers if I'm correct
>>> (surprising, given that Sun also had a good file serving reputation
>>> too).
>>>
>>> Now, I'm diverting too much.
>>>
>>> It's a nice machine if you care to collect. Finding software might
>>> be hard.
>>
>> If anyone wants photos of the internals of an AS/400 and the mood
>> strikes me I might be persuaded to open mine up.
>>
>>
>>
>>> Regards,
>>> Alvin.
>>
>
> What type(s) and model(s) do you have?
> - white CISC

White CISC--too lazy to go downstairs and refresh my memory on the details.

> - early RISC with still the large cards (I think they used V3R4 or so
> for the transition)
> - later RISC with PCI
> - iSeries (probably still look the same)
>
> I have 2, but not enough time to do much with them. I did reset the
> password (and already forgot it).
> I also tossed 2 whites 5 years ago (a B + expansion and a smaller D
> with built-in adapters). It did hurt (mentaly en physically), but I
> couldn't do anything with them, because:
> -> I mixed the adapters and disks because I couldn't lift them
> otherwise, but I forgot to note where wat was located
> -> when I got them (for free), I didn't know that finding software
> would be such a disaster
> -> when I learned about the battery held license key, I knew I was in
> deep sh*t, especially because V2R1 was non-transferable, already dated
> and "hobby" use is something IBM hadn't heard of

I ran into the same thing--it's mainly a conversation piece.  Haven't even 
powered it up in years.  No compiler with it and not much in the way of 
utilities and using it for a word processor is just, well, let's just say 
that the state of the art in that area has moved on quite a lot.

> Their total lack on eBay seems to confirm that keeping them up and
> alive is (nearly) impossible.
>
> I must say that it's a shame how IBM is treating these machines:
> software for the S/36 might still run unmodified on the current
> generation (it certainly did till V4R1): imagine running CP/M programs
> under Windows XP :-)

Actually most CP/M programs run fine under 32-bit XP.  You need to load an 
emulation layer called "22nice" (or any of several competitors--there used 
to be a lot of them) that maps the system calls.  Won't run directly under 
64-bit, but will if you put a DOS session in a virtual machine.  It should 
run under Windows XP mode on 64-bit Windows 7--don't have Windows 7 to try 
it, but the XP mode restores the 16-bit DOS box that was removed in the 
64-bit versions.

> Also, the concept is quite interesting. But then again, IBM did sell
> their PC division and let Apple drop the PowerPC.

Not sure how IBM could have prevented Apple from dropping the PowerPC--it 
was just behind the performance curve at that point.  Difference between a 
broadly based company and a specialist--at Intel Microprocessors R Us, at 
IBM they're just one product line out of many.  One can debate the 
theoretical merits of the design, but the fact remains that Intel and AMD 
are making processors that run like the hammers of Hell and the end isn't in 
sight.  I mean in 1970 who'd a thunk a 4 core machine running several 
gigahertz that was within the reach of someone working for minimum wage?

0
J
12/1/2009 1:13:29 AM
Alvin Andries wrote:
> It's successor: the white AS/400 had expansion cards that did look a bit
> like the mainframe stuff. My advise would be to look for a more recent
> AS/400 if you like the concept of an all in one sturdy box.

Alvin, you have no way of knowing that you're talking heresy.  I worked 
on the 9370 in Endicott, and we were betrayed by the AS/400 folks, so 
I've got nothing but vitriol for the AS/400 and the iSeries.

To top it off, I had to support a couple of the original AS/400 white 
boxes.  It didn't take long for me to hate OS/400 more than I disliked MVS.

And yes, those original AS/400s had cards that looked just like the ones 
in the 9370 (because they *were* the same I/O cards, that was part of 
the "unified architecture" plan that the AS/400 team threw out the 
window when it didn't suit them).  I spit on the SPD I/O bus, too -- 
which is only fair, I spent way too many hours helping to debug the I/O 
cards designed by idiots who didn't think they needed to abide with the 
functional specifications (this on the S/370 side, of course, with no 
help from the AS/400 team -- who were only too happy to pick up the card 
set after the grueling debug work was done).  As you can tell, these are 
not happy memories!

Lastly, while the 9309 rack *looks* sturdy, it's unibody design suffers 
from the same problem as uniframe vehicles -- if it ever takes a hit 
that bends the frame, you can stick a fork in it -- it's done!

Rick Ekblaw
0
Rick
12/1/2009 1:37:39 AM
"Rick Ekblaw" <ekblaw@vnet.ibm.com> wrote in message
news:MK6dnYXSnMaQ6YnWnZ2dnUVZ_q-dnZ2d@westnet.com.au...
> Alvin Andries wrote:
> > It's successor: the white AS/400 had expansion cards that did look a bit
> > like the mainframe stuff. My advise would be to look for a more recent
> > AS/400 if you like the concept of an all in one sturdy box.
>
> Alvin, you have no way of knowing that you're talking heresy.  I worked
> on the 9370 in Endicott, and we were betrayed by the AS/400 folks, so
> I've got nothing but vitriol for the AS/400 and the iSeries.

Over the years, I often wondered how much co-operation or competition there
was between the various IBM hardware divisions, and I couldn't entirely see
the need for quite so many different system lines. It's always interesting
to get an insider's views and experiences.

> To top it off, I had to support a couple of the original AS/400 white
> boxes.  It didn't take long for me to hate OS/400 more than I disliked
MVS.
>
> And yes, those original AS/400s had cards that looked just like the ones
> in the 9370 (because they *were* the same I/O cards, that was part of
> the "unified architecture" plan that the AS/400 team threw out the
> window when it didn't suit them).  I spit on the SPD I/O bus, too -- 
> which is only fair, I spent way too many hours helping to debug the I/O
> cards designed by idiots who didn't think they needed to abide with the
> functional specifications (this on the S/370 side, of course, with no
> help from the AS/400 team -- who were only too happy to pick up the card
> set after the grueling debug work was done).  As you can tell, these are
> not happy memories!

Hmm, I recall a similar story from working at NatWest's data centre, post
the RBoS takeover of NatWest. There was a lot of work going on to integrate
the various networks and software, but the NatWest technical guys considered
the managerial decision to adopt the RBoS model "as is" very short-sighted
and technically inferior. Having also worked at RBoS in a relatively minor,
backwater technical post, but also having been up to the Edinburgh centre
for meetings etc. I could feel sympathy for their position. But it's often
hard to know what factors were involved in the decision-making process.
-- 
Regards,
    Tim Clarke (a.k.a. WBST)
    Guildford, U.K.


0
Tim
12/1/2009 2:10:27 AM
On Nov 30, 8:13=A0pm, "J. Clarke" <jclarke.use...@cox.net> wrote:

> > I have 2, but not enough time to do much with them. I did reset the
> > password (and already forgot it).
> > I also tossed 2 whites 5 years ago (a B + expansion and a smaller D
> > with built-in adapters). It did hurt (mentaly en physically), but I
> > couldn't do anything with them, because:
> > -> I mixed the adapters and disks because I couldn't lift them
> > otherwise, but I forgot to note where wat was located
> > -> when I got them (for free), I didn't know that finding software
> > would be such a disaster
> > -> when I learned about the battery held license key, I knew I was in
> > deep sh*t, especially because V2R1 was non-transferable, already dated
> > and "hobby" use is something IBM hadn't heard of
>
> I ran into the same thing--it's mainly a conversation piece. =A0Haven't e=
ven
> powered it up in years. =A0No compiler with it and not much in the way of
> utilities and using it for a word processor is just, well, let's just say
> that the state of the art in that area has moved on quite a lot.

I did a ton of research into getting a AS/400 for hobby use.  In the
end I decided against it due to all the peculiarities with the
software and hardware.  Unless I manage to pick up a newer model
system locally WITH software, I've abandon the idea.

I am primarily interested in learning about the database-class
filesystem and unified storage architecture.  I think IBM did a lot of
innovative work here that seems to be coming up again in contemporary
UNIX derivatives (Linux, Mac OS X, Solaris, BSD) and Windows (with
WinFS).

The RS/6000 line and newer linage are the opposite and pretty easy to
"hobby" since AIX media is readily available and machine licensed.

>
> Not sure how IBM could have prevented Apple from dropping the PowerPC--it
> was just behind the performance curve at that point. =A0Difference betwee=
n a
> broadly based company and a specialist--at Intel Microprocessors R Us, at
> IBM they're just one product line out of many. =A0One can debate the
> theoretical merits of the design, but the fact remains that Intel and AMD
> are making processors that run like the hammers of Hell and the end isn't=
 in
> sight. =A0I mean in 1970 who'd a thunk a 4 core machine running several
> gigahertz that was within the reach of someone working for minimum wage?

I don't think anyone lost sleep when Apple dropped the PPC from their
personal computer line.  The margin off those was pretty
insignificant.  IBM instead got a hat trick with all video gaming
consoles coming out at that time.

IBM Semi's real dropped ball is in embedded CPUs, something they were
leading for some time.  For example, the Apple iPhone (and most other
smartphones) use an ARM chip.  The volume on these is staggering, and
only predicted to increase.

FWIW IBM still leads the pack at silicon performance.  POWER6+ is the
fastest CPU out there, and its been out for a while.  POWER7 is
nearing release with even more punch.  Intel's Itanium offering at
this level is pathetic in comparison.
0
Kevin
12/1/2009 2:51:36 AM
> I am primarily interested in learning about the database-class
> filesystem and unified storage architecture.  I think IBM did a lot of
> innovative work here that seems to be coming up again in contemporary
> UNIX derivatives (Linux, Mac OS X, Solaris, BSD) and Windows (with
> WinFS).

Kevin, in such a case you would find this book helpful:

Owen Hanson, "Design of Computer Data Files", 2nd Ed., Computer Science
Press, 1988. ISBN 0-7167-8197-2.

Most of the design principles outlined in this book apply to the design of
database files, for this book discusses structured file design and fast
access and retrieval methods. There are also specs and results from tests on
IBM and ICL tape drives and disk units, including even the early floppy
drives.

I extended and applied the "inverted files" technique from this book with
great success, it was pretty amazing to see how fast that DOS database app
(around 1990-92) could search. The trick was to collect the "inverted" data,
coming from a number of files, in a master file early at the time of data
insertion and update. At query time, you just read this master file and
displayed the results.

"Inverted" file means that you collect data based on secondary attributes
and not on the primary key. For a simple example, in a file system, that
could be a master list of all executable files so that the OS can instantly
locate the executable to load instead of (sequentially) scanning one or more
directories.




0
UZnal
12/4/2009 10:45:33 AM
Reply: