On z9, mainframe blogging, serendipity, biography, killer marketing, and MVS geekdom

One of things that the blogosphere drives is serendipity, "the faculty of making fortunate discoveries by accident".

Links can make you lucky, and as Louis Pasteur said: "Chance favours the prepared mind".

What do i mean "links can make you lucky". For one thing most search engines work by counting links, so as links increase so does relevancy. For another, people follow and cluster and distribute links.Thus, when i recently posted a blog asking where are all the mainframe bloggers, I was expecting some kind of response. Someone somewhere was going to pull the link chain in.

What I didn't know was just how good my timing was. I had written the post months before, but never got round to posting it. As soon as I did though, July 15th, IBM's mainframe communications people jumped. They wanted to introduce blogging as an element of System z9 marketing and I had just made the kind of call to action they were already planning to.

So naturally they asked if I wanted to contribute to a group blog. I believe in making a contribution. But I am also zealous in guarding my independence. I wanted to know about the blog ownership, copyright, who else was involved. It may seem odd to say the one question I didn't ask was: "how much will I get paid?" But that is blogs for you.

I was reassured that IBM would include representatives from both the user community, but also, more importantly for credibility's sake, other mainframe vendors. I would not have got involved if it was just an IBM marketing vehicle. We'll see how this plays out. Finally, you will notice the blog is not even hosted by IBM.com.

That is because, increasingly, IBM gets it. People like Catherine Helzermann, James Snell, and executives with cojones, are leading IBM into the world of conversational marketing. One of the tenets of blogging, as far as I am concerned, is Be More Competitive By Being Less Competitive. If you are in attack mode all the time you tend to lose credibility, Ed Brill's solid FUD/anti-FUD work notwithstanding. I regularly link to other smart analysts with good ideas. I would rather do that than steal the idea and claim it as my own. And I know other firms can do things we at RedMonk can't, don't or won't.

So that's a few words on mainframe blogging, but what about biography? One key thing in starting a blog is to say a little bit about who you are and why you're doing it. Who is this James Governor cat, and why should I pay a jot of attention to what he has to say?

I am 35 and I have been covering the IBM mainframe ecosystem for ten years now. I used to be a reporter and editor. I think i can hear the knives on the grind stone. "This guy has been a journalist and analyst and expects to have any credibility? "

I will take that on the chin, but with a caveat.

When I started watching the mainframe in 1995 it really was grey hairs and geeks. I had to learn a new language and felt I had to fight for credibility in every meeting. I wasn't just 25; I was a young looking 25. Sometimes I thought about wearing glasses just to try and look more mature. The fear of being found out meant that I immersed myself in gorpy stuff like SNA, partitioning and virtualization, job scheduling, message-queing, workload management, the meaning of mixed workloads and I/O intensive operations. But you had to extend this stuff too - terminal emulation and middleware were bread and butter. I covered Amdahl and HDS, EMC and StorageTek, BMC and CA.

My colleagues called me "legacy boy". So much for legacy - many of the dotcom and networking firms they covered don't exist any more... What they called legacy I called production environments. And the age thing? I couldn't help but notice that boas, one of the other posters on this group blog, was born two years after me                        .

These days of course our skills are in demand. Every systems vendor out there has mainframe envy, and needs to get better at delivering on, and communicating, hardcore technical concepts for today's computing challenges. When I talk to Michael Emanuel about Microsoft's systems management strategy we both know that mainframe quality is the high bar-that is what Microsoft is shooting for.

According to this August 2004 interview in news.com Bill Gates is still measuring his company's efforts against the IBM System/360 project. He said:

 

Our scheduling and predictability on this project has   been better than it was on OS 360 (the mainframe operating system created by   IBM). So software has not gotten more complex. Software with this kind of   scope of features and compatibility has always been complex. That's the   business we're in.

I think the admission  is somewhat revealing of Gates's psychology, but that's a subject for a different blog...

Its worth taking a step back, at this point, to compare a couple of systems launches this week.

Microsoft began the long beta road to delivery of Vista.

IBM announced System z9 and some other assorted goodies.

What is kind of amazing is that, as far as I can see, IBM blew away Microsoft from a PR perspective.

News stories about Vista tend to raise more questions than they answered. What's the point, what's the value, when will it be here? Words like gripe and delay peppered the coverage.

IBM's z9 press coverage meanwhile was more like the US governments': it was all about certainty.

It was like John Kerry against George W Bush, with IBM as the Republicans.

I should make an admission at this point. I was pre-briefed on the IBM systems news last Thursday and I didn't think it quite made it. I fired off an email to that effect. You sure you tied the concepts together? I wasn't at the event in NY earlier this week but evidently IBM did just that.

IBM wanted to frame the debate using terms like "collaborative processing" and collaborative design. What the hell is collaborative processing? Who cares, as long as it means brand z9 and brand IBM are associated with one of today's hot memes, collaboration. And the association worked bloody well, as these links show.

So IBM got it right. z9 associations were aspirational and all about clarity.

What is one of businesses biggest concerns at the moment? Identity leakage and theft. While Microsoft is doing some good work with respect to security and ID, it was IBM earning the approval of column inches. Erich Clementi's story about tape and encryption evidently caught the imagination.

And for pranksters what about the names of these systems, and the cars they evoke?

Would you prefer a BMW concept car or a 1970s Oldsmobile station wagon?

What I also can't understand is why nobody at Microsoft used their search engine to see about other Vista Software companies. Its called due diligence. And it only takes a second. I guess when legal and out of court settlements become part of the business model you can get a little complacent. But if I was Steve Ballmer I would be firing someone this week.

No such problems for IBM z9, unless BMW wants to claim it is a software company now.

Not everyone was impressed with IBM's moves.

But all in all the proof is in the press. IBM has the third estate on side, and on message. At least for now. We'll know soon enough if that translates into sales.

One thing that I wanted to make sure I did when posting to the mainframe blog was to give hardcore geeks something to check out. Other bloggers or techies with similar interests. So how about MVS Turnkey emulation, open source style? SAP R/2 hands keep popping out of the woodwork. Here is some 3278 pr0n.

One final note before I go. If you want to read blogs you can make life very easy for yourself by using a simple online service called bloglines. Here is a how to guide. You don't need to visit a bunch of sites every day. Blogs come to you, which one of the reasons they are so useful.

Thus ends my first Mainframe Blog post.

by James Governor July 29, 2005 in People
Permalink | Comments (7) | TrackBack (1)

The Return of the 9

I am not your typical mainframer which might have to do with the fact that I was born too late. After the Virtual Machine Facility/370 was announces on August 2, 1972, it took me another 4 months to see the world with my own eyes. The announcement of that August are still preserved whereas my own general availability went mostly unnoticed. To make up for this lack of first hand experience of the early computer years, my grey haired mentor suggests history lessons.

One of these took me to Konrad Zuse. A civil engineer, he build his first computer in 1938. Later in 1941, he build the Z3, which today is widely acknowledged as being the first fully functional automatic digital computer. Although build with relays, no CMOS nor Bipolar, you will find a lot of the concepts still deployed today. The Z3 had a memory of 64 words, each 22 bit wide. Additional machines followed. The Z4 had a clock with 30 Hertz and could perform 11 multiplications per second. The Z5 was six times faster. He also build a machine that was called Z9. To get a sense of progress and innovation it is good to look back at those numbers and compare it with the 1 billion transaction of the new z9 mainframe of 2005. But to understand Konrad Zuse's genius we have to compare today's typical teams of several hundred or thousand engineers to this single pioneer. He entered a completely new field of technology. He could not really build on existing work and communication with other Researchers was difficult during the war time. He cut across different fields of expertise: mechanics, electrics and logic to name a few. Through the 1940s he also invented a sophisticated programming language called Plankalkül.

Zuse's accomplishments are proof that imagination is more important than knowledge.

by Boas Betzler July 27, 2005
Permalink | Comments (0) | TrackBack (0)

The Mainframe Geek

Recently on Slashdot I found myself reading many postings on the topic "What is Mainframe Culture?". It turns out that Mainframers have a real sense of identity. In fact, they seem to me like the prototype geeks. A geek is defined as a "person who is interested in technology". I went to Geekfinder and entered 'mainframe' as search expression. Sure enough, the result list had over 2000 matches. What would qualify you to apply for one of those?

You might be a real mainframe geek if ...

  • YOU WRITE ALL YOU EMAIL UPPERCASE.
  • when called to court, you look for a Principle Of Operation to take your oath on.
  • you still think they should have never started casual Friday.
  • you have a large supply of punchcards that you use to write shopping list on the back
  • you have either grey hair or no hair.
  • all your tattoos are in EBCDIC.

Seriously. There is one personality trait that mainframe people learned early on: to share. Sharing a bag of candies means that you don't eat them all yourself. It also means not to give them away to the first guy around, but to save some for your friends. Mainframers grew up with the concept that they don't own the machine and are not the only one online. That they cannot grab all the resources all the time. They had to apply for machine time to run their programs through the reader. When they started to use screen interface, they logged on to the "Time Sharing Option". Writing batch jobs, the first thing to declare is which resources are needed.

The concept of virtualization, sharing a system through the use of abstracted and architected interfaces, is inherent in the mainframe culture. Slowly the Unix geeks find out that more than one application can run in an Operating System. Recently more and more Windows geeks run multiple Operating Systems on the same machine. On Virtual Machines. Pretty soon, they all will cherish the mainframe culture. They will think twice before re-booting an OS just because there could be many other applications running. They will no longer turn the power off just because they see the blue screen of death. Clint Boulton has written a piece about a 24 year old programmer now hooked on mainframes.

In a virtualized IT infrastructure the server, network and storage resources are shared among all the business applications. Running a successful data center means to assign these resources to the applications and transactions according to business priorities along efficiency, reliability, availability and performance. If this what you do, treat yourself to a state-of-the-art caffeine delivery vehicle. You are real mainframe geek.

by Boas Betzler July 25, 2005 in Future, History, People
Permalink | Comments (11) | TrackBack (1)



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.