Don't worry, HP's project Moonshot will save us

[Neil Rieck]

Don't worry, HP's project Moonshot will save us

http://www.zdnet.com/hp-launches-project-moonshot-powered-with-intels-atom-at-first-7000013686/

NSR

[MG]

On 9-apr-2013 1:19, Neil Rieck wrote:
> Don't worry, HP's project Moonshot will save us
>

> [...]

HP will sell you a box with a ton of 'smartphone parts' smashed
together and calls it a server? I guess that could work in
today's exciting world of computing...

Perhaps it's any consolation that it will be RISC (so in that
specific regard like AXP... for what it's worth, /hooray/?) and
it's at least not another 'blade' thing... right?

Then my eye spied this little interesting /thing/:


<http://cdn-static.zdnet.com/i/r/story/70/00/013686/moonshot2-614x343.png>

- "77% less costly": I wonder, compared to what?

- "89% less energy": Could be true, but (again), compared
to what?

- "80% less space": 'Compressing' 720 cores in a 4U
RM enclosure /only/ results in 80%? Compared to a fully
loaded blade thing, perhaps? Although I'd think a blade
enclosure is a 'bit' bigger than a 4U RM system, but that
could be me.

- "97% less complex": ...? I'd think that 180 processors
inside a single system would anything but less complex?
(End user perspective, 'not having to deal with sockets'
kind of sentiment, perhaps?)


- MG

[Stephen Hoffman]

On 2013-04-08 23:19:21 +0000, Neil Rieck said:

> Don't worry, HP's project Moonshot will save us
>
> http://www.zdnet.com/hp-launches-project-moonshot-powered-with-intels-atom-at-first-7000013686/
>

Moonshot is clearly not aimed at the same market and the same style of
computing that is performed with the typical OpenVMS server.

HP's <http://www.hp.com/go/moonshot> has some more details.

For some tasks, those boxes should work very nicely. Many web
services tasks should do nicely with that. Also for dedicated hosting
providers and for folks offering VPS computing, for instance.

List price of US$1260 per board/blade, plus the cost of the chassis box
and I/O and storage giblets. Not cheap, but certainly dense. The
question is whether it's cheaper (in terms of power or density or ease
of management or... whatever) than available and competitive ODM
hardware configurations that various organizations are now using.

Facebook has made available designs <http://www.opencompute.org>
including Open Compute and Open Rack, based on what they're using in
their data centers, for instance.

Strictly on core count, the Moonshot configuration would have once
involved a whole rack of AlphaServer DS10L boxes. It'd be interesting
to see some comparisons of a 466 MHz or 600 MHz EV6 against a dual-core
2 GHz Atom, too. At least in terms of simplicity of installation and
density, Moonshot easily wins.

The issue with software-defined computing — or utility computing or
whatever it's called these days — tends to be with the system and
management and application software, and the associated documentation.
Great in theory, certainly. If that whole stack all works and meets
your needs, then you're good to go. If not, then you just bought a
gazillion idle cores, and with the requisite integration and deployment
and management headaches. (Not that you woudn't have had these
headaches at this scale already. One common bit of "fun" for any of
these projects includes integrating third-party hardware and software
into whatever you're using.)

Odd that the HP site shows "Sorry, there are no documents available for
this product" in the referenced Moonshot library. Ah, well.

Interesting. But I'll reserve judgement, for now...



--
Pure Personal Opinion | HoffmanLabs LLC

[Neil Rieck]

Yep, and on top of that, there's a little voice in my head saying "Really, HP is going to build a new server line on another specialty chip from Intel?".

NSR

[MG]

On 9-apr-2013 17:31, Neil Rieck wrote:
> Yep, and on top of that, there's a little voice in my head
> saying "Really, HP is going to build a new server line on
> another specialty chip from Intel?".

I saw also others, mockingly, remark that in the comments
there.

In Dutch they have the saying: "A donkey won't bump itself
twice against the same stone."

- MG

[Stephen Hoffman]

On 2013-04-09 15:31:02 +0000, Neil Rieck said:

> Yep, and on top of that, there's a little voice in my head saying
> "Really, HP is going to build a new server line on another specialty
> chip from Intel?".

Atom is a low-power x86-64 processor.
<http://www.intel.com/content/www/us/en/processors/atom/atom-processor.html>
Intel has been pushing Atom for microservers, among other uses.
Various examples of Windows systems running on Atom clients are
available too, including the "HP Spectre XT TouchSmart 15t-4000
Ultrabook".

[Bill Gunshannon]

In article <kk1d07$cr8$1...@dont-email.me>,

Gee, I wonder if that's what is in the HP mini-laptop I have that
prints the message "processor not supported" when I try to install
drivers for any HP printer......

bill


--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>

[Stephen Hoffman]

On 2013-04-09 15:54:29 +0000, Bill Gunshannon said:

> Gee, I wonder if that's what is in the HP mini-laptop I have that
> prints the message "processor not supported" when I try to install
> drivers for any HP printer......

Finally time to upgrade that Pentium 90 notebook, eh?

[Paul Sture]

In article <d77caace-9362-4c89...@googlegroups.com>,

Neil Rieck <n.r...@sympatico.ca> wrote:

> Yep, and on top of that, there's a little voice in my head saying "Really, HP
> is going to build a new server line on another specialty chip from Intel?".

Have you seen this one?

<http://www.theregister.co.uk/2013/04/09/hp_moonshot_server_analysis/>

--
Paul Sture

[MG]

On 9-apr-2013 18:30, Stephen Hoffman wrote:
> On 2013-04-09 15:54:29 +0000, Bill Gunshannon said:
>
>> Gee, I wonder if that's what is in the HP mini-laptop I have that
>> prints the message "processor not supported" when I try to install
>> drivers for any HP printer......
>
> Finally time to upgrade that Pentium 90 notebook, eh?

LOL.

Those were far from "mini" though, from what I remember. Well,
then it depends what people consider a "mini".

I had a little debate with IBM people the last few days, so I
can tell from first-hand experience that opinions vary quite
a bit...

- MG

[Keith Parris]

On 4/8/2013 5:45 PM, MG wrote:
> - "77% less costly": I wonder, compared to what?
>
> - "89% less energy": Could be true, but (again), compared
> to what?

"Compared to traditional servers", says http://hp.com/go/moonshot

> - "97% less complex": ...?

Compared with cabling together than many servers, with power, Ethernet /
Fibre Channel, management connections, etc.

[Keith Parris]

On 4/9/2013 9:31 AM, Neil Rieck wrote:
> there's a little voice in my head saying "Really, HP
> is going to build a new server line on another specialty
> chip from Intel?".

From the white paper at
http://h20195.www2.hp.com/V2/GetDocument.aspx?docname=4AA4-6179ENW&cc=us&lc=en

"HP intends to assemble a catalog of HP ProLiant Servers, powered with
chips from different manufacturers, with widely varying performance and
features, and with overlapping new introductions and update cycles. That
will be new for our industry. These server cartridges will implement
accelerators using a variety of technologies: x86 and ARM CPU cores,
compute offload via GPU, DSP, FPGA, and fixed function logic, and
eventually APUs (accelerated processing units)."

So both x86 and ARM processors as well as other options are envisioned.

[Bill Gunshannon]

In article <kk1fig$2mo$1...@dont-email.me>,

Stephen Hoffman <seao...@hoffmanlabs.invalid> writes:
> On 2013-04-09 15:54:29 +0000, Bill Gunshannon said:
>
>> Gee, I wonder if that's what is in the HP mini-laptop I have that
>> prints the message "processor not supported" when I try to install
>> drivers for any HP printer......
>
> Finally time to upgrade that Pentium 90 notebook, eh?

Wishful thinking. It's an HP 2133 and has been like this since
it came out of the box. Reminds me of the Celeron days. Glad
I wasn't the one who paid for it. I ought to just put a Linux
Desktop on it, but wait, everyone keeps telling me all these
laptops I have with Linux on them can't possibly work cause
Linux can't do a Desktop. :-)

[MG]

On 9-apr-2013 22:11, Bill Gunshannon wrote:
> Wishful thinking. It's an HP 2133 and has been like this since
> it came out of the box. Reminds me of the Celeron days. Glad
> I wasn't the one who paid for it. I ought to just put a Linux
> Desktop on it, but wait, everyone keeps telling me all these
> laptops I have with Linux on them can't possibly work cause
> Linux can't do a Desktop. :-)

Well, go ahead and find out. It's about time that a big advocate
would try it for himself...

- MG

[Stephen Hoffman]

The processor in that HP 2133 MiniNote is a VIA C7-M processor
<http://en.wikipedia.org/wiki/VIA_C7> x86-compatible processor.

For Linux on that box...
<http://www.linlap.com/hp_2133>
<https://wiki.ubuntu.com/LaptopTestingTeam/HP2133>

HP reportedly had a Linux distro for some models of that laptop, too.

There's an old printer driver kit for Windows on the support page for
that laptop, but whether that supports your printer....?

On Linux, CUPS will work with most printers.

Given the dates I'm encountering on files associated with that laptop,
it looks to be, well, not as old as Pentium 90, but still pretty old.

[MG]

On 9-apr-2013 23:15, Stephen Hoffman wrote:
> The processor in that HP 2133 MiniNote is a VIA C7-M processor
> <http://en.wikipedia.org/wiki/VIA_C7> x86-compatible processor.
>
> For Linux on that box...
> <http://www.linlap.com/hp_2133>
> <https://wiki.ubuntu.com/LaptopTestingTeam/HP2133>
>
> HP reportedly had a Linux distro for some models of that laptop, too.
>
> There's an old printer driver kit for Windows on the support page for
> that laptop, but whether that supports your printer....?
>
> On Linux, CUPS will work with most printers.
>
> Given the dates I'm encountering on files associated with that laptop,
> it looks to be, well, not as old as Pentium 90, but still pretty old.

The overall Linux power management capabilities are poor, this is
hardly a secret either. That largely kills the experience on the
laptop/notebook, well, for me anyway.

As a matter of fact, even some big name Linux developers/contributors
have been critical about this, even fairly recently. (Remember, for
instance, that article that was shared here, a while back, about the
co-founder/creator of GNOME that spoke out?)

I run Linux Mint x86-64 on an older Acer 5536. It barely works, it
feels more like a 'poorly cooled computer with a low-grade UPS'.
I'm actually going to try FreeBSD on it soon, since I can then at
least also use interesting things like ZFS and the "jails".

- MG

[MG]

On 10-apr-2013 0:46, I wrote:
> I run Linux Mint x86-64 on an older Acer 5536.

Acer Aspire 5536, for completeness.

- MG

[Brad Hamilton]

On 2013-04-09, MG <marc...@SPAMxs4all.nl> wrote:
> The overall Linux power management capabilities are poor, this is
> hardly a secret either. That largely kills the experience on the
> laptop/notebook, well, for me anyway.
>

Can you quantify, "poor"? I ask, because I use Linux Mint 13 on an older
Lenovo with 1st gen Intel Dual-Core. XFCE, instead of Mate or Cinnamon,
allows me to use hibernate and sleep with few problems, as opposed to
Mate, Cinnamon, or Ubuntu Unity.

> As a matter of fact, even some big name Linux developers/contributors
> have been critical about this, even fairly recently. (Remember, for
> instance, that article that was shared here, a while back, about the
> co-founder/creator of GNOME that spoke out?)
>
> I run Linux Mint x86-64 on an older Acer 5536. It barely works, it
> feels more like a 'poorly cooled computer with a low-grade UPS'.
> I'm actually going to try FreeBSD on it soon, since I can then at
> least also use interesting things like ZFS and the "jails".
>

I avoided using FreeBSD years ago, because the wireless device drivers were
poor/nonexistent. That experience may have changed by now, but I won't risk
an otherwise "stable" platform. I may experiment on a "throwaway" laptop.
[...]

[Phillip Helbig---undress to reply]

In article <51645a0d$0$6367$e4fe...@dreader35.news.xs4all.nl>, MG

<marc...@SPAMxs4all.nl> writes:

> Those were far from "mini" though, from what I remember. Well,
> then it depends what people consider a "mini".

Minicomputers, anyone?

[David Froble]

Stephen Hoffman wrote:

> Given the dates I'm encountering on files associated with that laptop,
> it looks to be, well, not as old as Pentium 90, but still pretty old.
>
>

Well, so am I, pretty old that is. What are you trying to say ??

:-)

[Neil Rieck]

I guess I spoke too soon. I just assumed HP was only focusing on ATOM but it appears they also have plans to support ARM. There are two links of many:

Calxeda: ARM-based HP Moonshot 'later this year'
http://www.zdnet.com/calxeda-arm-based-hp-moonshot-later-this-year-7000013693/

HP Project Moonshot hurls ARM servers into the heavens (this one from 2011)
http://www.theregister.co.uk/2011/11/01/hp_redstone_calxeda_servers/

NSR

[Paul Sture]

In article <51649a55$0$6345$e4fe...@dreader35.news.xs4all.nl>,

MG <marc...@SPAMxs4all.nl> wrote:

> The overall Linux power management capabilities are poor, this is
> hardly a secret either. That largely kills the experience on the
> laptop/notebook, well, for me anyway.
>
> As a matter of fact, even some big name Linux developers/contributors
> have been critical about this, even fairly recently. (Remember, for
> instance, that article that was shared here, a while back, about the
> co-founder/creator of GNOME that spoke out?)
>
> I run Linux Mint x86-64 on an older Acer 5536. It barely works, it
> feels more like a 'poorly cooled computer with a low-grade UPS'.
> I'm actually going to try FreeBSD on it soon, since I can then at
> least also use interesting things like ZFS and the "jails".

FWIW the recommendations I have come across for ZFS say 4GB and
preferably 8GB minimum RAM. If you have less RAM don't be surprised
if ZFS is slow.

--
Paul Sture

[Paul Sture]

In article <slrnkm9de...@coyote.earwicker.home>,

Brad Hamilton <br...@coyote.earwicker.home> wrote:

> On 2013-04-09, MG <marc...@SPAMxs4all.nl> wrote:
> > The overall Linux power management capabilities are poor, this is
> > hardly a secret either. That largely kills the experience on the
> > laptop/notebook, well, for me anyway.
> >
>
> Can you quantify, "poor"? I ask, because I use Linux Mint 13 on an older
> Lenovo with 1st gen Intel Dual-Core. XFCE, instead of Mate or Cinnamon,
> allows me to use hibernate and sleep with few problems, as opposed to
> Mate, Cinnamon, or Ubuntu Unity.

The later versions of Mate, Cinnamon or Gnome required better graphics
than my 3 year old budget tower system has. XFCE is fine on that
however, and for consistency I use it elsewhere.

> > As a matter of fact, even some big name Linux developers/contributors
> > have been critical about this, even fairly recently. (Remember, for
> > instance, that article that was shared here, a while back, about the
> > co-founder/creator of GNOME that spoke out?)
> >
> > I run Linux Mint x86-64 on an older Acer 5536. It barely works, it
> > feels more like a 'poorly cooled computer with a low-grade UPS'.
> > I'm actually going to try FreeBSD on it soon, since I can then at
> > least also use interesting things like ZFS and the "jails".
> >
>
> I avoided using FreeBSD years ago, because the wireless device drivers were
> poor/nonexistent. That experience may have changed by now, but I won't risk
> an otherwise "stable" platform. I may experiment on a "throwaway" laptop.
> [...]

For those who don't have spare hardware, FreeBSD sits nicely inside
VirtualBox running on either Linux or Windows.

However, unlike Windows or any of the other Linux clients I have run
under VirtualBOx, FreeBSD doesn't get the latest time from the host when
resuming after a suspend. This might make it less suitable for use on a
laptop. There is of course nothing to stop you closing a FreeBSD client
rather than suspending it (or investigate other ways to get the correct
time after a suspend).

--
Paul Sture

[Paul Sture]

In article <asjaut...@mid.individual.net>,

bi...@server1.cs.uofs.edu (Bill Gunshannon) wrote:

> In article <kk1fig$2mo$1...@dont-email.me>,
> Stephen Hoffman <seao...@hoffmanlabs.invalid> writes:
> > On 2013-04-09 15:54:29 +0000, Bill Gunshannon said:
> >
> >> Gee, I wonder if that's what is in the HP mini-laptop I have that
> >> prints the message "processor not supported" when I try to install
> >> drivers for any HP printer......
> >
> > Finally time to upgrade that Pentium 90 notebook, eh?
>
> Wishful thinking. It's an HP 2133 and has been like this since
> it came out of the box. Reminds me of the Celeron days. Glad
> I wasn't the one who paid for it. I ought to just put a Linux
> Desktop on it, but wait, everyone keeps telling me all these
> laptops I have with Linux on them can't possibly work cause
> Linux can't do a Desktop. :-)

That's up there with someone a couple of years ago telling me that Linux
Mint was tuned for the desktop so if I wanted to run a LAMP system I
should install Ubuntu or Debian server on a separate box.

It was perfectly fine for development and testing :-)

--
Paul Sture

[Stephen Hoffman]

That you need to splurge, and get yourself something a little faster...
Think of it as your mid-life crisis laptop.
:-)

[Bob Koehler]

In article <kk1mfm$fc$1...@usenet01.boi.hp.com>, Keith Parris <keithparris...@yahoo.com> writes:
>
> "HP intends to assemble a catalog of HP ProLiant Servers, powered with
> chips from different manufacturers, with widely varying performance and
> features, and with overlapping new introductions and update cycles. That
> will be new for our industry. These server cartridges will implement
> accelerators using a variety of technologies: x86 and ARM CPU cores,
> compute offload via GPU, DSP, FPGA, and fixed function logic, and
> eventually APUs (accelerated processing units)."

The more they invent new buzzwords, the higher the stock prices go?

[Bob Koehler]

In article <51649a55$0$6345$e4fe...@dreader35.news.xs4all.nl>, MG <marc...@SPAMxs4all.nl> writes:
>
> I run Linux Mint x86-64 on an older Acer 5536. It barely works, it
> feels more like a 'poorly cooled computer with a low-grade UPS'.
> I'm actually going to try FreeBSD on it soon, since I can then at
> least also use interesting things like ZFS and the "jails".

I once loaded Linux 0.98 on a 386. Bringing up X-11 and doing
nothing was 100% CPU load. Not impressive. Even MS could do
a GUI on a 386 and leave some CPU power to do work.

Without X-11 running, the box met our needs as a low security
network gateway (lower than we'd want in a firewall). DOS would
not have.

[Bob Koehler]

In article <kk2qp6$8im$2...@online.de>, hel...@astro.multiCLOTHESvax.de (Phillip Helbig---undress to reply) writes:

>
> Minicomputers, anyone?
>

Super minis!

(Those who recognize the term probably all have grey hair.)

[Stephen Hoffman]

On 2013-04-10 15:01:42 +0000, Bob Koehler said:

> In article <kk1mfm$fc$1...@usenet01.boi.hp.com>, Keith Parris
> <keithparris...@yahoo.com> writes:
>>
>> "HP intends to assemble a catalog of HP ProLiant Servers, powered with
>> chips from different manufacturers, with widely varying performance and
>> features, and with overlapping new introductions and update cycles. That

>> will be new for our industry. These server cartridges ...

>
> The more they invent new buzzwords, the higher the stock prices go?

Yeah. What if HP just called it their new third modular-format
chassis? Adding the new Moonshot form-factor cartridges and modular
chassis to their existing c-series BladeSystem modular chassis
products, and their the SD2-class high-end Blade chassis? Nah. I'd
never survive in marketing. :-)

[Paul Sture]

In article <+d78Rk...@eisner.encompasserve.org>,

Yep. I wonder if Phillip has seen an 11/780?

A pair of them side by side took a lot of space when compared to the
racks available nowadays.

--
Paul Sture

[Paul Sture]

In article <kk3qvr$5sn$1...@dont-email.me>,

A lot cheaper than that sports car too :-)

Though maybe not so much fun.

--
Paul Sture

[Keith Parris]

On 4/8/2013 5:19 PM, Neil Rieck wrote:
> Don't worry, HP's project Moonshot will save us

There's an Executive Roundtable webcast today entitled "HP Moonshot -
for a new style of IT". See
http://h30507.www3.hp.com/t5/Converged-Infrastructure/Executive-Roundtable-on-April-10-HP-Moonshot-for-a-new-Style-of/ba-p/135601

[Keith Parris]

On 4/10/2013 8:46 AM, Keith Parris wrote:> There's an Executive

Roundtable webcast today entitled "HP Moonshot -
> for a new style of IT".

If the URL in your announcement doesn't work, use
http://hpbroadband.com/program.aspx?key=HP041013

[David Froble]

What I'd really like to do is splurge and get myself a little faster,
and younger. Haven't found the proper vendor yet.

:-)

[Bill Gunshannon]

In article <nospam-33907D....@news.chingola.ch>,

When I reached that point I already had two sports cars so I got a
Bradley. :-)

[Phillip Helbig---undress to reply]

In article <nospam-2A551E....@news.chingola.ch>, Paul Sture

<nos...@sture.ch> writes:

> > (Those who recognize the term probably all have grey hair.)

Well, the beard is turning grey.

> Yep. I wonder if Phillip has seen an 11/780?

I had one, or something similar, in my car once. Someone offered to
give away some hardware (I was interested in terminals) and said he
would rather have one person take it all. I figured I could pass on
anything I didn't need, but was surprised by this beast. It was a
refrigerator-size rack, either an early VAX or a PDP. (Come to think of
it, probably the latter; perhaps one can find my old posts on this.) I
had 12 offers to take it within 24 hours. I had a 1994 Polo "estate",
i.e. the one with two doors but the long windows in the back,
essentially a 2-door estate. With the front seats pushed all the way to
the front and the rear hatch open, it just fit. It also included
software on 10-inch floppies.

As to the chap who took it---now THAT'S a collector!

[glen herrmannsfeldt]

Phillip Helbig---undress to reply <hel...@astro.multiclothesvax.de> wrote:
> In article <nospam-2A551E....@news.chingola.ch>, Paul Sture
> <nos...@sture.ch> writes:

>> > (Those who recognize the term probably all have grey hair.)

> Well, the beard is turning grey.

>> Yep. I wonder if Phillip has seen an 11/780?

I saw one two days ago. (Not running, though.)

> I had one, or something similar, in my car once. Someone offered to
> give away some hardware (I was interested in terminals) and said he
> would rather have one person take it all. I figured I could pass on
> anything I didn't need, but was surprised by this beast. It was a
> refrigerator-size rack, either an early VAX or a PDP. (Come to think of
> it, probably the latter; perhaps one can find my old posts on this.)

The 11/780 is a little shorter, but a lot wider than the usual
refrigerator. I don't think it would fit in the usual large
car or minivan, but I suppose in a pickup truck.

> I had 12 offers to take it within 24 hours. I had a 1994 Polo "estate",
> i.e. the one with two doors but the long windows in the back,
> essentially a 2-door estate. With the front seats pushed all the way to
> the front and the rear hatch open, it just fit. It also included
> software on 10-inch floppies.

> As to the chap who took it---now THAT'S a collector!

-- glen

[Bill Gunshannon]

In article <kk4ghs$dgk$1...@online.de>,

hel...@astro.multiCLOTHESvax.de (Phillip Helbig---undress to reply) writes:

I amy be mistaken. but I seriously dowbt that an 11/780 would fit in
a Polo. The last one I saw would have been a challenge in the back
of my Silverado. :-)

And what is a 10" floppy? The biggest I ever saw were 8". (Still got
some. Even some still in the box and shrink-wrapped. :-)

>
> As to the chap who took it---now THAT'S a collector!

People here will probably remember that when I was at the University I
used to have not only PDP-11/44's but also some rather large VAXen.
Bigger than your average fridge. Two of them went to collectors and,
sadly, one went to the landfill. Oh yeah, all my 11/44's went to
other collectors, as is most of the rest of my stuff before too long.
It is rapidly reaching the point where I will be spending my time on
the golf course and no longer in a computer room.

[Paul Sture]

In article <kk4ghs$dgk$1...@online.de>,

hel...@astro.multiCLOTHESvax.de (Phillip Helbig---undress to reply)

wrote:

While Polos were surprisingly capacious I don't think a full size 11/780
would fit in one. A PDP seems more likely.

According to

http://www.computerhistory.org/revolution/mainframe-computers/7/182/736

"Dimensions
overall: 60 in x 48 in x 30 in"

The ones I used didn't have the side cabinet seen in the photo there;
the disks were free standing RM05 drives. A brief search brought up this
photo:

http://home.ntelos.net/~donbryan/Don/Pics/my_disk.jpg

--
Paul Sture

[Paul Sture]

In article <2DFFib...@eisner.encompasserve.org>,

My first experience with Linux was with a 300 MHz Pentium in 1999. Even
with 192MB RAM (which was plenty and then some for NT4) firing up X11
was like watching paint dry.

--
Paul Sture

[Brad Hamilton]

Or none...

[Phillip Helbig---undress to reply]

In article <nospam-95CC96....@news.chingola.ch>, Paul Sture

<nos...@sture.ch> writes:

> While Polos were surprisingly capacious I don't think a full size 11/780
> would fit in one. A PDP seems more likely.
>
> According to
>
> http://www.computerhistory.org/revolution/mainframe-computers/7/182/736
>
> "Dimensions
> overall: 60 in x 48 in x 30 in"

Sounds about right. Was there some model of PDP which was very similar
to a VAX? 11/750? 11/730?

[Johnny Billquist]

Since I think we have already established than VAXen came in all kind of
sizes, and so did PDP-11s, the answer is obviously "yes". :-)

Now, are you interested in similarities in size to some specific VAX model?

Johnny

[chris]

On 04/10/13 15:07, Bob Koehler wrote:

> Super minis!
>
> (Those who recognize the term probably all have grey hair.)
>

I still have a copy of the digital "Super Micro Systems" handbook, from
around
1985 ? or so, if I remember the title right. And yes, there is much grey
hair,
though whether due to years of toil with computer systems or something
else,
has yet to be determined :-)...

Regards,

Chris

[MG]

On 12-apr-2013 1:18, chris wrote:
> I still have a copy of the digital "Super Micro Systems" handbook,
> from around 1985 ? or so, if I remember the title right. And yes,
> there is much grey hair, though whether due to years of toil with
> computer systems or something else, has yet to be determined :-)...

No grey hair here. For some inexplicable reason, somehow, I always
found it interesting to read into these (especially) subjects and ---
by having done so --- have learned a great about things that took
place long before I was even conceived.

It's sad, yet also very understandable, why not more of (roughly)
my age (and especially below) show no interest for VMS or any
other of these 'legacy platforms'. But, it has nothing to do
with VMS itself. That can't be stressed enough in my opinion.

- MG

[MG]

On 12-apr-2013 1:18, chris wrote:

> I still have a copy of the digital "Super Micro Systems" handbook,
> from around 1985 ? or so, if I remember the title right. And yes,
> there is much grey hair, though whether due to years of toil with
> computer systems or something else, has yet to be determined :-)...

No grey hair here. For some inexplicable reason, somehow, I always

found it interesting to read into (especially) these subjects and ---

[MG]

On 10-apr-2013 14:08, Paul Sture wrote:
> FWIW the recommendations I have come across for ZFS say 4GB and
> preferably 8GB minimum RAM.

That's something I indeed came across, which is a /bit/ ridiculous.
LSM is a lot more efficient, but sadly unlike AdvFS, it's not freely
available as open source.

> If you have less RAM don't be surprised if ZFS is slow.

I also scrapped the idea to run that in a VM, as I was prompted
about it recently.

There is an override, but I don't feel much for it.

- MG

[chris]

Development pace was ragged edge and exciting when I first started
in computing and there were new developments and applications every
week. Now, everything has coalesced down to commodity and there's been
no real development in decades, just continuous refinement of existing
technology. You have a pc on your desk, you no longer need to get the
soldering iron or wire wrap gun out to build a custom interface. Just
buy it off the shelf at a fraction of the cost and effort. As the hardware
becomes commodity, the interfaces become more abstract and high level.
Fewer people know or need to know the low level stuff and end up seeing
only a fraction of the big picture.

Might also be something to do with lack of scientific curiosity in the
young. Computing, electronics and the sciences are no longer "fashionable"
enough, but it's just that prowess that has give the west it's competitive
advantage for decades...

Chris

[chris]

On 04/11/13 06:53, Paul Sture wrote:

>
> My first experience with Linux was with a 300 MHz Pentium in 1999. Even
> with 192MB RAM (which was plenty and then some for NT4) firing up X11
> was like watching paint dry.
>

So were sun and dec workstations of the time, though quite bit faster than
a Vaxstation 2 GPX running decwindows, which was almost unusable. VWS was
quite a bit faster, fwir.

Regards,

Chris

[Bob Koehler]

In article <asydnYU5vd_qGPHM...@bt.com>, chris <me...@devnull.com> writes:
>
> Development pace was ragged edge and exciting when I first started
> in computing and there were new developments and applications every
> week. Now, everything has coalesced down to commodity and there's been
> no real development in decades, just continuous refinement of existing
> technology.

That depends on what part of the industry you work in. When you
work close to custom hardware, typically one-off, there is no off
the shelf application.

In scientific computing, folks have to come up with new algorithms
to test new theories and analyze new data every day.

In accounting, nothing new has happened since the origin of the
row and column paper "spreadsheet", except that it has become
electronic.

If I'm running a business, I probaly want an off the shelf
application to track inventory, accounts recievable, ...

When the fellow down the hall was working on his Nobel winning
research, he might not have had a COTS product that could do the
analysis.

[John Wallace]

On Apr 16, 2:21 pm, koeh...@eisner.nospam.encompasserve.org (Bob
Koehler) wrote:

Unbeliever! You can do everything in the lab with NI Labview these
days without writing a single line of code, just pointy clicky draggy
things. DECrti (where rti = real time integrator) did a bit of that
pointy clicky data acquisition and dataflow programming too at one
time, and unlike generic Labview not just on Window boxes either, but
is presumably long gone from this earth (iirc, it transitioned from
DEC to Kinetic Systems in the mid 1990s).

NI still even get away with charging premium prices for premium
product and premium support and the customers are allegedly still
mostly happy, just like DEC used to at one time. It must all be true,
I've seen it in their monthly magazine, the IEE News. [OK it's called
IET Magazine now and it's much less techy than it used to be, which
includes not being as NI-centric as it used to be, but not too long
ago...]

[Neil Rieck]

Just saw this piece over on Ziff-Davis.

IBM's potential x86 server sale to Lenovo highlights oncoming train

http://www.zdnet.com/ibms-potential-x86-server-sale-to-lenovo-highlights-oncoming-train-7000014273/

Now I know that this thread deals with HP Moonshot (initially based upon Intel Atom but supposedly next-gen Moonshot will offer ARM) but this move by IBM "seems" to indicate they will will only manufacture systems based upon their own CPU rather than compete in the x86 marketplace. Perhaps lying low in a volatile market is their survival strategy.

NSR

[David Froble]

Well, I'm still selling my own version of kool-aid ....

People did WP on MicroVAX II systems because there were not any PCs ...

People did their browsing on desktops because there were not any
notebooks ...

People are now using notbooks, tablets, and smart phones ...

In each case, people were using what was available because what was
later developed and more suited to their usage wasn't at first
available. So, this erosion of the "old type" of devices was quite
natural. Inevitable. It's going to happen. Resistance is futile ...

But what this explosion of new users does is not replace past usage of
computers, but adds to the total of users. I think it's safe to say
that everyone has use for phone. But not everyone has use for a
"mainframe", "mini", or "micro".

There are still people using, needing, and wanting VMS systems. The
ones that have gone away are those who really needed a PC, Notebook,
tablet, or phone. Yeah, there is the defection to Unix and stuff, but
in the larger picture, all the "big iron" is in the same situation.

If a decent "system" still cost a quarter million or more, you'd see
more attention to that market by the vendors. The lowering cost of
hardware has confused the vendors, which as most all businesses end up
in the hands of beancounters who know nothing about computers,
automobiles, aircraft, and just about anything else. All they know how
to do is count beans, and there are a lot more phones to count than
mainframes.

The original market for "real" computing systems is still there, you
just cant find it among all the "beans" ....

The real problem is that the vendors have lost it in their huge pile of
beans ...

[MG]

On 23-apr-2013 18:27, David Froble wrote:
> There are still people using, needing, and wanting VMS systems.

I wonder. But, it's certainly a nice thought.

> The ones that have gone away are those who really needed a PC,
> Notebook, tablet, or phone.

An operating system with a reliable TCP/IP stack, recent and up-to-date
(network) security facilities (e.g. IPsec), improved software/package
management, at least one recent and common up-to-date web browser,
native integrated software development tools/suites, graphics software,
etc. ... besides (indeed, like you mentioned earlier) the usual 'word-
processing'/'office'-oriented software. Just to name a few things.

> Yeah, there is the defection to Unix and stuff, but in the larger
> picture, all the "big iron" is in the same situation.

IBM is never in trouble, you /evil infidel/!? They'd probably lynch
you, like in places like over at ClassicCmp, when you'd even think of
suggesting something like the above ... even /by accident/ and without
ill intentions. Certainly never get the 'funny idea' to cite, say, a
[The] Register article stating anything contrary to their mythical
preconceived conceptions about IBM's true popularity and success 'out
there'.

Luckily IBM's 'self-confidence' will, so they believe, make them
survive the coming tide of x86/ARM bipolarity and luckily the faith
in Big Blue is very strong amongst the faithful disciples of IBM.

In the meanwhile, fewer and fewer of the recent and upcoming
generations --- especially those born since the last 15~20 year ---
will even know what "IBM" itself is. They never even heard of the
"IBM PC", only if they somehow get the idea to 'look it up' at (e.g.)
Wikipedia.

> The original market for "real" computing systems is still there, you
> just cant find it among all the "beans" ....

I actually 'wonder' about that.

To me, the worrying thing in this all is not that the volume of "PC",
"Mac", "smartphone"/"tablet" far more exceeds things like, say, VMS
server-type systems. No, not at all. Even in the days of VAX, VMS
wasn't exactly visible inside every office and neither were other
platforms, like it. VMS was pretty much always a low(er) volume type
of platform. Nothing wrong with that, to everything its place.

What does seem quite worrying, to say the least, is that many of
those aforementioned high-volume platforms are getting the most
exposure and preferential treatment. Those platforms are also
making significant inroads into the living space of, well, VMS
and many other (to certain degrees, similar) server-type systems.
Especially if you count Linux as a typical PC-type system (since
it runs mostly on x86).

At least back in the VAX days, when computers in general weren't
nearly as widespread and rarely in people's homes, many universities
had VMS systems or still taught students about VMS; in the least
students were more likely lectured or somehow told that VMS existed
at all! (Unlike today, where most people would probably thing that
"VMS" refers to "virtual machines" and "OpenVMS" unsecured ones...)

So, in the end, most people --- including many future system and
software 'architects' --- will 'grow up', be 'educated' and use
these platforms. Platforms like VMS, if they will even exist
much longer in the foreseeable future, will be relegated to (even
more) niche roles and only require certain maintenance skills,
likely taught by smaller crowds of embedded or detached specialists
in corporations.

At least VMS has the (if I may say so, excellent) "Hobbyist Program".
But, even that doesn't seem to be enough to revive or let alone spike
its popularity.

Ironically, the last bit of fairly widespread public exposure, were
stories revolving around an ARM-powered SoC type system ("Raspberry
Pi") running VMS VAX through SimH. So, essentially VMS on 'smart
phone' equivalent hardware... I find that very telling, to be
honest.

Far more than desiring to 'reimplement' a "GNU VMS", people around
these places (like here on comp.os.vms) should actually be thinking
of guerrilla advertisement plans. Especially when you consider that
VMS was never properly marketed to begin with.

Maybe it's not too late to still attempt it now? That would also
nicely show how much people would truly care to see VMS survive
and how much of their free time it's worth.

- MG