HP, sad

On Wed, 16 Sep 2015 04:30:31 -0700, John Larkin Gave us:

What is sad is your ignorance.

(Is it even worth the bother to correct your mistakes?) George H.

You are wrong.

Always.

Totaly from the Motorola and then Kodak playbook. What other companies did this?

It's called progress.

the x86 stuff works great, is cheaper, faster and more plentiful than anything else out there unless you cross off cheaper. There's also no lead time on PC based servers. I can have any intel based HP server next day. That's not the case with old timers and their stories of VAX 11-780s and other large computers that were really more of a "machine" than a computer.

Reliability is excellent with properly made PC servers too. The shit will run nonstop for years, I see this all the time. Never seen uptimes of years on a VAX or Alpha, although somebody may have pulled that off.

The software and OS selection for intel based servers is out of this world too. No more vendor lock-in with proprietary everything.

We're far ahead of the "good old days" by any measure. Yeah, big old machines were really fun and amazing, but they've long been eclipsed in any measure.

No joke. Their server grade machines are still top of the line. Are they perfect? no, but nothing is. They have corrected some of their stupid mistakes too, like the unusable sun/fujitsu style "drop-in" server rail kits that appears with the G8 series machines. That stuff was garbage.

I used to run a PDP-11 timeshare system that would stay up for months between power failures. I don't know if my new Win7 is that good, but XP benefitted from regular reboots.

The problem with x86 is not that the architecture is slow - massive pipelining and silicon speed fix that - but that it is so insecure.

I'm hoping that massively multicore ARM processors will take over.

Xenix was even earlier. Microsoft licensed the Unix V7 source from AT&T in about 1978. The plan was to sell Xenix as the OS for running MS BASIC. The plan was put on hold when IBM arrived and wanted BASIC for their new IBM PC. When IBM mentioned that they could also use an operating system, MS suggested Xenix, but IBM wanted no part of anything attached to AT&T. So, MS offered QDOS (Quick and Dirty OS) which eventually became MSDOS and PCDOS. Meanwhile, Microsoft sub-licensed Xenix to various OEM's. Eventually, Xenix became sufficiently popular to create a demand. MS did not have the resources to do both Xenix and MS/PCDOS, so they hired SCO to do a joint development with Xenix in about 1982(?). Fairly soon, SCO was doing all the Xenix development. The rest you can read in the Wikipedia article:

I was selling and servicing Xenix in SCO's back yard starting in about

1985 with Xenix 2.0 on a 286. I also had customers running Xenix on 8086, Lisa 68000, and one ancient PDP-11 something. Speed was acceptable but what was impressive was the uptime. Xenix 286/386 would go for months without any reason to reboot. Compared to the other systems of the day, this was a minor miracle.

Bottom line: If IBM had not decided to build a PC in 1981, we would all be running Xenix today instead of Windoze.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

You too can invite Jeff Bezos into your living room to listen to every conversation.

I believe it was a II. A full rebuild took fourteen hours for a system that fit in < 1 meg RAM. Same codebase on a 286 was minutes.

Your mileage may vary :) I got a lot of very good work done with nothing but DOS and a compiler, maybe the odd little board here and there for level shifting.

--
Les Cargill

It was klunky, but it's shocking what you could get outta one if you knew how. The systems software was all very clean.

Huh - I knew about the 1000 series, but not the 2xxx series. Looks funky - and sounds just like a PDP-8, including the "store the return address at the top of ..." hack.

I spent a year on a PDP-8 one week.

I have several contacts who were flushed out. The senior management literally knew nothing about the company. After all, look at what Fiorina did to Lucent before joining HP

Same basic thing.

I've always been reasonably happy with it. But I had several jobs where we'd roll in PeeCees to do what took a much more expensive computer before, so the money was good.

Best book to own from 1985 to 1995: "Advanced MS-DOS" by Ray Duncan.

--
Les Cargill

On Wed, 16 Sep 2015 19:37:24 -0500, Les Cargill Gave us:

She led one of the most successful IPOs in US history at the time.

You might want to re-think that. Unless you are giving her credit for finding and flushing the idiots.

During her time at Lucent, Fiorina added 22,000 jobs, grew revenues from US$19 billion to US$38 billion, net income went from a small loss to US$4.8 billion profit and the company's market share increased in every region for every product.

Was the uVAX starved of ram? VM thrashing would explain it. The two should have been broadly comparable if it was a uVAX II.

A Sun would have eclipsed either of them.

--
Regards, 
Martin Brown

I hope he enjoys my conversations. My only family is my two dogs.

I remember I worked with XENIX about 1984/1985 when we started using it to deploy "our application" that had been developed on SysV Unix on lower-end hardware. I think something had changed then that made it useful for us, but I do not remember exactly what it was. (the version for the XT, the version for the 286, or maybe a reduction of the price)

Anyway, a usable operating sytem for the IBM PC was available back then and maybe a bit earlier. Most people claiming the OS of the PC was unusable only have knowledge of MS/PC DOS.

(there even was a UCSD P-system for the PC, but of course it was similar in DOS w.r.t its limitations)

With software building the situation was even worse than with other CPU intensive operations, because the compilers on VMS were so dog slow even on a fast system.

On MSDOS you had "Turbo C" (later renamed "Borland C") that compiled much and much faster even on slow hardware.

On 17 Sep 2015 12:03:50 GMT, Rob Gave us:

Folks also forget that the 80186 did exist, and was one of the most widely used CPUs of the age in embedded and industrial applications. Certainly the most widely used of the x86 family until makers like Tektronix, etc. started placing entire SBCs into their designs. And that wasn't until well into the Pentium realm.

As far as OSes go, one of the forgotten ones was DesqViewX, which would run a process on a remote PC long before anything like that was in vogue, as long as the network was under TCP/IP. It was an X server and would run C script from the CLI, just like Linux did when it appeared.

I used to really like the HP300-series MC680x0 boxes. My fave was the

9836. As a grad student, I had a 9816 with a math coprocessor board and a compiler compatible with HP's Rocky Mountain Basic, which I loved. To this day AFAIK there's nothing as nice for instrument control.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net