Why don't toasters work properly?

I have just read an intriguing article in the Daily Telegraph. It would appear that despite the fact that the Romans (what did the Romans ever give us? – see below)  mastered toast, manufacturers are still struggling to create an effective electric toaster.

Now, I do wonder how they determined that the Romans made good toast – did they find a piece lying around somewhere? However, that’s not the point.

In the article it says “dazzling displays and design can’t hide the fact that most toasters aren’t very good at their core function – making toast.”

I would like to offer a slight change to that sentence: “dazzling displays and design can’t hide the fact that most distributed systems aren’t very good at their core function – supplying a production level operating system.”

Yes, I am a lot older than James and the others here, and I was brought up in a mainframe world. I was raised on an operating system that was designed to do multi-tasking, recover itself from errors, was secure, and did not have to be rebooted every ten minutes. For more boring details about my life, click here or here.

Do I want everyone to throw away their wondrous new pretty Unix, Linux and Wintel boxes? No. Do I want people to throw away 30 years of hard-earned experience in how to run production systems? Also no.

What I want people to do is to take the best pieces from each world and build truly useful systems, which are designed to make businesses run better.

Ah yes, the Romans


They knew which side of the road you should drive on – Napoleon ruined it all later

by pwarmstrong August 4, 2005 in History, Innovation, People, Systems Technology


TrackBack URL for this entry:

Listed below are links to weblogs that reference Why don't toasters work properly?:



I found your view on distributed systems and mvs systems finding their individual strenghts to build an overall better OS to be enlightening.

I’m new to the mainframe, only being exposed to it in the past 2 months as an intern at Kohls, but an avid user of distributed systems all my life, just like every young college student today. I myself see these systems striving to have the stability, availability, security, and virtualization found in mvs, most notably the availibility "The accessibility of a system resource in a timely manner; for example, the measurement of a system's uptime." and the powerful virtualization. These two aforementioned characteristics which i have noticed cannot be compared to any supplimentary utility product(VMware) in the distributed systmes world in their effectiveness.

Here they have just one z990, but use it to keep up with many innovative concepts i have read about in my quest for the future of the mainframe and i take it as a privilege being given the opportunity to learn this very useful information, and valuable skills.

I am currently getting my associates degree in CNS(Computer Network Systems) and am very seriously thinking about going the remaining two years for a bachelors in Information Security(now i'm not so sure). My only trouble is the saturation I foresee this field having in the future, since it is the next big thing, and the cost of school(IS). But i see how my education and interests could go hand in hand with the innovative course the mainframe is taking knowing how security central it is.

Like the mainframe geek quiz says all of my team is either balding or has grey hair(well i see it as white), except for my manager who i presume is in his mid to late 30's. This generation gap is huge, since i just turned 20, but they still can relate to me in some aspects and do keep me entertained.

Comming from someone of your background what innovations do you foresee in the mainframe’s future? And do you really see the Mainframe as being still the best at virtualization and parallel computing with the advances in VMware and the release of the p series server(and i series for that matter)? Do you also foresee the possible outsourcing of the Mainframe programmer/support since IBM seems to offer all the resources and funding to the Chinese?

By the way, i use a toasteroven :)

Posted by: Kyle Horton | Aug 4, 2005 12:59:49 PM

holy crap this blog thing is actually working. hello kyle, great to here a voice direct from the trenches of today's IT education.

I would not worry about saturation. infosec is a huge field and frankly there is an awful lot of dross out there.

a guy with mainframe experience, with all the latest educational bells and whistles, is going to be very well positioned for the next decades.

do those guys think in hex? ;-)

on the VMware side it will be interesting to see what happens when IBM and HP start putting more resource into Xen.

regarding p-series style parallel processing - that is obviously not the mainframe's core competence. probably where blue gene comes in. and of course a data center consisting of z at the center and a bunch of bladecenters would make an interesting topology.

a 20 year old coming on line. ok i feel old.

Posted by: james governor | Aug 4, 2005 1:24:55 PM

Security - agree with James, loads of room to grow / expand there for years yet especially with things like Sarbox, Basel II etc pushing people to get it right.

Virtualisation - see the innovation coming in DS world from necessity (the operating systems are so awful compared with MVS that they have to reinvent a virtual environment via the hardware).

Mainframes need to become easier to manage and more flexible in their definition, because there are fewer skilled people out there to look after them.

This in turn sets up opportunities for the emerging geographies to grasp the problem, teach loads of people MVS and take over the world (of mainframes).

Posted by: Peter | Aug 5, 2005 9:32:14 AM

Kyle wrote: "And do you really see the Mainframe as being still the best at virtualization and parallel computing with the advances in VMware and the release of the p series server(and i series for that matter)?"

I hear questions like this all the time. Those who ask usually approach it with the mistaken assumption that the "alternate platforms" are improving but the mainframe is just standing still, waiting for everyone else to catch up. The virtualization technologies on pSeries are "mainframe-inspired", not "just like the mainframe" (as my pSeries sales friends occasionally mistakenly claim). There are a number of differences in how they work, such as how the mainframe actually uses a hardware assist (SIE) to dispatch LPARs, and the time slicing algorithms in the microcode, etc. And as far as VMWare is concerned, it is a long, long way from the maturity level of PR/SM or even the pSeries hypervisor.

All of the IBM server platforms are maturing their virtualization functions with time, including the mainframe. One must include z/VM in the equation alongside PR/SM in that maturity/capability equation, a factor that none of the alternative platforms have.

So, two things to remember: 1) The mainframe technologies (virtualization and beyond) continue to grow and improve, at or above the pace of distributed systems, and 2) The distributed technologies that appear to be "just like the mainframe" are merely "mainframe-inspired" - they are not the same


Posted by: Bill Seubert | Feb 2, 2006 9:59:06 AM

The comments to this entry are closed.

The postings on this site are our own and don’t necessarily represent the positions, strategies or opinions of our employers.
© Copyright 2005 the respective authors of the Mainframe Weblog.