Friday, July 11, 2008

The Real Thing!

I was reminded recently of my times living in the ‘70s. I was thumbing back through old pictures, and came across this picture taken of me at the Sydney Auto Show back in the early ‘80s.

You can see that I maintained my ‘70s hair style, even then, with not a single grey hair in sight!

The car I am pictured with is the race car, a Mazda RX7, of Canadian/Australian touring car driver Allan Moffat, with which he successfully campaigned against the local V8 cars of Ford and General Motors. My brother, Greg, and I were anxious to take a good look at the real thing, as we had only ever seen it on TV.

I had returned to Sydney in late 1977 as the head of technical support for the local subsidiary of the US-based software company, The Computer Software Company (TCSC). The company had built an alternate, but fully compatible, operating system for the IBM mainframe that it called the Extended Disk Operating System (EDOS).

Working closely with the then-dominant third party mainframe leasing companies, EDOS gave users features that could only be found in the more expensive OS/MFT and OS/MVT variants and supported the attachment of the latest disk drives. This was a time when earlier generations of mainframes still retained residual value and, as new models were introduced into the marketplace, often provided the newer technology with stiff competition.

After only a few months at EDOS (Australia), I became its Managing Director. These were heady days for IT in Australia as by 1979 the IBM Plug Compatible Manufacturers (PCM) really began to flourish. Amdahl, with its 470 line was the first to arrive and found quick success at QANTAS. It was followed by the short-lived offshoot of the leasing company Itel, with its line of Advanced Systems that found a niche among the service bureaus of the day.

Fujitsu, Hitachi, and even Mitsubishi then began building IBM PCM mainframes and the competition between the vendors intensified. A number of European vendors began to re-badge these systems as they too saw an opportunity to enter this market. Business opportunities for a company like EDOS (Australia) looked promising, and we began to validate EDOS on a number of different PCM offerings.

At the height of these PCM wars, Nixdorf Computer bought TCSC and I was given the opportunity to fold the local subsidiary into Nixdorf Computer (Australia). A new business unit was created, and with the full support of Nixdorf’s Managing Director, Dieter Monch, we established out own PCM presence in the Australian marketplace in 1982. Only three years later however, I could see the end of the PCM business as the hardware vendors began to view Unix as the way to go!

I have included a picture here of my own Mazda RX7 that I had built as a copy of Moffat’s racing car. While Moffat was not allowed to modify his engine, my imitation racing car was fitted with a turbocharger and become one of the earliest examples of a Mazda RX7 turbo! I had added an aero kit, and added the same wheels. It wasn’t the real thing, and was never a true race car, but I enjoyed driving it all the same.

It was definitely as fast as Moffat’s race car! And I will always remember the ticket I was given early one Sunday morning as I drove down the freeway from the Blue Mountains, just west of Sydney, with the Managing Director of Nixdorf Computer, New Zealand. As I talked with the highway patrol officer, my cause wasn’t helped at all when the New Zealand boss asked for a copy of the ticket so that he could prove to his friends just how fast we were going!

While IBM remains just as strong as it was in the ‘70s, the rest of the landscape has changed dramatically. Today IBM is no longer facing the challenges of many vendors imitating their product offerings and flooding the marketplace with copies, but rather, is coming to terms with the world of industry standard and open systems. Furthermore, IBM now has to operate from the position of second place as HP has passed it in revenues and market share for the first time in IBM’s history.

While IBM has shed its PC business, spun off its line of printers, and exited the disk and tape market, HP continues to invest in all of these technologies and continues to gain enormous leverage from them. And for the first time in a long while, I considered the possibility that, once again, we could revisit the glory days of the ‘70s PCM marketplace. Could we see HP supporting the mainframe zOS operating system and would there be HP product offerings that were compatible with IBM?

But the IT landscape is so different these days. In the late ‘70s, it wasn’t just our hair that were long, it was our portfolio’s of custom code. The need for a PCM marketplace reflected the dire straights CIOs were in – they couldn’t walk away from the applications and would only entertain new product offerings if they could simply pick-up their applications and drop them, unchanged, onto the new platform.

Today, we rely on packaged solutions with the software vendors providing support of frameworks that allow portability, and this gives us the option to select a platform from any vendor supporting the framework. If an application is written in Java, for example, and the run-time environment is supported on a variety of platforms, the opportunities are extensive. At this year’s HPTF&E I came across the company, Tmaxsoft, who could re-host a mainframe application, CICS and all, onto HP platforms including the HP Superdome. Samsung Life Insurance talked about how they used Tmaxsoft’s OpenFrame and TJES products to replace not only CICS but the Job Control Language (JCL), IBM’s mainframe scripting language, as well. And isn’t leveraging the investment in the business logic that is important?

Pretty impressive accomplishment and something I have a sense we will see a lot more often. Even at my own company, GoldenGate, where the focus has evolved to embrace real time data integration, I am seeing a lot of activity around data migrations and the desire of CIO’s to move off older systems and onto newer, industry-standard, platforms. As GoldenGate continues to build on its Transactional Data Management (TDM) offerings, I was particularly impressed when I came across a recent analyst report, Bloor Research’s June 2008 “Data Migration – Market Update”, that noted “data migration is a subset of the data integration market … a very large subset!" And again, isn’t protecting the investment we have made in the data the other important consideration?

The need to maintain compatibility with the hardware and operating system would appear to be no longer a real and important requirement. We have the ability to mover the applications and to preserve the data no matter the specifics of the situation. Servers that simply imitate mainframes and that are plug-compatible with mainframe operating systems, aren’t likely to make a come-back given this.

For a while I did think that this may eventuate as I watched the fortunes of one company considering such an opportunity. Readers of the December 19, ’07 blog posting: “Virtualization? Unreal, mate!” will recall the reference I made to the Sunnyvale, CA, company Platform Solutions, Inc (PSI). In that posting I referenced their web site where PSI described their new Open Mainframe Servers as being “Intel Itanium2 based Mainframe servers for z/OS.”

I have included a picture of me alongside just such a product offering where I saw z/OS running on Intel’s Itanium processors. But at the time I took this picture, PSI and IBM were locking horns and had gone to the courts for a resolution of their differences. However, earlier this month, IBM purchased PSI and PSI’s President promises “to collaborate on future offerings and maximize our combined knowledge skills for the benefit of IBM clients globally.”

Not sure what this all means and where it will be headed, but at the end of the announcement, there’s a reference to both IBM and PSI dropping their respective claims against each other. I have been in dialogue with a number of folks from PSI over the years, and believe that they are genuinely happy with the outcome and are looking forward to working with their peers at IBM but I am pretty sure that this puts an end to any speculation that HP may support z/OS on its Intel Itanium-based servers. Perhaps, following the announcement that both the System i and System p would share a blades package, IBM will follow HP’s lead and use the PSI firmware to add support for System z on blades. But I somehow doubt it.

As the ‘70s came to a close I was reminded of a song that appeared at the close of the ‘60s. Australian songwriter Johnny Young wrote the song “The Real Thing”, and local singer Russell Morris made it a hit across the country. The lines from that song that have stayed with me through the years included the catchy phrase:
“There's a meaning there, but the meaning there doesn't really mean a thing.Come and see the real thing! Come and see the real thing!”

The ‘70s saw me leaving the user community and moving over to the vendor side, where I have remained ever since. Having spent the first years in IT writing mainframe applications, it was pretty easy to gravitate to vendors in that marketplace. With the experience I gained at EDOS and at Nixdorf Computers, I was convinced that the mainframe was the real thing. And even today, the mainframe remains a viable option for many users and its supporters are as enthusiastic as I was all those years ago.

But chasing the mainframe and emulating it “doesn’t mean a thing” anymore. The industry has moved on, and it’s the infrastructure and solutions providers who ensure we retain options. After all, it’s our investments in business logic and the data that is important, not the platform. And that, by way of contrast, really does mean a lot!

No comments: