iBelieve NY: If You Don’t Like Change. . .
October 1, 2012 Timothy Prickett Morgan
If you are trying to get to the future, you can’t always be looking over your shoulder at the past. You have to keep your eyes on the road ahead. And if any message was clear at the iBelieve event hosted by looksoftware last Thursday in my hometown of New York City, it is that we can celebrate our long history but we have to keep moving, keep learning, keep changing, and keep pace with modern technologies like the IBM i platform itself has managed to do. And perhaps with a lot more grace than the humans who program and operate the machines.
“We have a very strong heritage, and sometimes people get our heritage confused with our future,” explained Trevor Perry, an independent consultant who does implementation work sometimes using looksoftware tools. Perry has become the standard bearer and chanticleer for the platform in recent years and he emceed the iBelieve NY event. With a foil showing the many generations of hardware leading from the System/3 to modern Power Systems behind him, with the venerable AS/400s from the late 1980s and early 1990s in the middle of the pile, Perry pointed the finger of blame squarely on all of us who have not kept pace.
“Part of the problem is that we can’t get people to start coding like it is a modern system,” he said. “When you create a physical file, you are living in that white box area,” Perry said, pointing at those AS/400 machines. (I realize that as a man with a shortage of cones in my retinas, I am chromatically challenged, but I thought the AS/400s were light beige?)
Like many of us, Perry gets fired up about the great-great grandchildren of the System/38, architected in large part by Frank Soltis, who gave the opening keynote of the event and who is a fabulous story teller and gave us his own story of how he came to the central role he has played in IBM midrange systems since the 1960s. I am not going to get into that now–I will share that with you separately–but rather want to focus on what the speakers at iBelieve had to say about the past and future of the IBM i platform that we all have a stake in. Personally, as I told Soltis later in the day when I thanked him for all he has done to create the machine that has fed me for so decades now, I plan on the IBM i platform footing the bill through these newsletters for my children to go to college, and he quipped that the AS/400 and its progeny had put his three sons through college and would very likely do so for a bunch of his grandchildren.
We all have some skin in this game, and to his credit, Perry simply would not allow the discussion to degrade into wallowing in a mudpit discussion about IBM’s lack of marketing for the IBM i platform.
“The platform has a very bright future and is in fact selling well in any place where they are feeling good about the economy,” Soltis said, and obviously even though he no longer works at IBM since retiring at the end of 2008, Soltis is very hooked into IBM Rochester and keeping people from talking to him would be precisely as difficult as it would be to keep Soltis from speaking his mind. Which he has always done, to his great credit. Even when it was not politically savvy.
“If I had a dollar for every time someone asked me if this product, call it IBM i or whatever, is going away, I would have a very fat bank account,” quipped Soltis, explaining that once the System/38 was designed and launched in 1978 and shipped in 1980, he had been expecting for it to be changed by something else within five years or so after the launch as was customary fashion for Big Blue at the time. “For me, this product has been going away since 1983.”
That’s not to say that there hasn’t been pressure on the AS/400 and its successors. IBM has recently said that it has more than 150,000 customers using OS/400, i5/OS, and IBM i platforms, a number that was considerably larger than what many of us had expected. (My own estimate was closer to 120,000.) And Soltis cited internal IBM statistics that he was privy to from the late 1990s that showed the base peaked at around 240,000 customers. I had Bill Zeitler, who used to run the AS/400 business and then IBM Systems Group, tell me in 1997 when the AS/400e machines were launched that there were 275,000 unique customers worldwide for the OS/400 platform, and what this shows is that it is very difficult to count your platform base when a large portion of it goes dormant. Microsoft is equally clueless about how many Windows Server shops there are because a customer doesn’t necessarily pay Microsoft money, and unless a platform has a call-home feature, the vendor can’t know it is still running.
The absolute number of current and past peak customers using OS/400, i5/OS, and IBM i platforms is not the point. The relative decline and what IBM learned from it was, as Soltis explained.
We have all railed against IBM’s lack of marketing for the platform against rivals both inside of Big Blue and out in the IT market at large, and we all know IBM’s strategy: If IBM can’t keep them on OS/400 or IBM i, then it can try for AIX on Power Systems, and if that doesn’t work, then there is Windows on System x with Linux on System x as a failsafe.
“About two years ago, someone inside of IBM decided to survey these customers who left, and guess what? 95 percent of all customers who left i left IBM. They were not good IBM customers, “Soltis explained. “They were good IBM i customers. Anybody who walks away from this platform walks away from IBM.”
Well, we could have told you that. And I think I might have a few hundred times. (Possible four hundred, even. . . .)
The good news, as Alison Butterill, the IBM i product offering manager, explained in her session, is “that the IBM i business is doing well” and IBM is seeing “good uptake on Power7 hardware and IBM i 7.1.” She did not give any precise statistics–IBM has just finished its third quarter and she cannot talk about it yet because the books are not done, but as we have pointed out in past issues of The Four Hundred, shipments and revenues have been up as 2011 cane to a close and 2012 got rolling. The interesting bit of data Butterill could share was that about two-thirds of the Power Systems machines being configured with IBM i coming out of IBM’s factories today are at the 7.1 release level. “You need to be at 7.1 to keep going down the road with us.”
And, by moving to IBM i 7.1, you will be in position to take advantage of the relatively new and very useful Technology Refresh method of updating IBM i that has been created to add features and functionality to the operating system without necessitating a new release and therefore a new qualification cycle.
This is exactly how the major Linux platforms are all doing their releases these days because no one wants to have to go through the hardware and application software qualification and compliance testing that a new release requires. It is just too expensive and too much of a pain. With the prior cadence, you would have been expecting for IBM to put a new IBM i release–we’ll call it 7.2–into the field about now (let’s say next week with the new Power7+ iron), but that is not the plan. IBM listened to customers who can’t take a new release and has instead decided to keep using Technology Refreshes to update 7.1, enabling it to run on new iron and add new functions.
“Everybody wants us to slow it down,” says Butterill. “At some point, we will have to do a new release. Something that is large enough and pervasive enough that we’d have to give you a PTF that ate the world to do.”
Just like there are accounting rules that govern how much of a machine can be changed and you can call it an upgrade rather than a new system, there are rules on compliance boards that govern whether you can call an operating system patch an update (or in this case, a Technology Refresh) or a new release or version
IBM has been updating IBM i in April and then October, so it is reasonable to surmise that Technology Refresh 5 for IBM i 7.1 is due this month. (Funny bit of trivia: This April-October cadence is precisely the same that Canonical‘s Ubuntu Server variant of Linux and the OpenStack cloud controller are on, and it would be funny if all operating systems fell into a similar lockstep. Red Hat has been doing March-November updates for a while and may synchronize with OpenStack.)
As for what that those updates to IBM i might be in TR5, Butterill was not about to say much. Prior Technology Refreshes have added a lot of database functionality, and we can expect more tweaks and tunings as IBM helps customers push more work through every Power thread and core. IBM is making enhancements to its PowerHA clustering and Backup Recovery Media Services (BRMS) hierarchical storage software, and will continue to make investments in the RPG programming language and its Open Access extensions.
The one thing that is still on the To-Do list is getting COBOL enabled for Open Access, and Butterill said that what we should do is send her emails about it so she can forward them on to the Rational tool team to get this bumped up to a higher priority. COBOL is still widely used in financial services and insurance industries, and in many cases, those are old IBM mainframe shops that jumped from System/390s and ES/9000s many years ago to the AS/400 platform.
As for a native .NET environment, don’t hold your breath. “A relationship like that would require participation from both sides, and I am just going to leave it at that,” Butterill said. As I have said before, IBM could just use the Mono open source C# runtime if it really wanted to do something interesting and perhaps get Microsoft to play along. But, Big Blue can’t do anything that might antagonize Microsoft and upset that System x business.
I will report on the other presentations at the event in next week’s issue, provided I have room and IBM doesn’t dump a lot of announcements on us on October 3. But the theme coming out of the iBelieve conference resonated with a famous quite from General Eric Shinseki, former chief of staff for the U.S. Army: “If you don’t like change, you’re going to like irrelevance a lot less.”
We all need to learn new things that are part of the IBM i platform and make sure they actually get used solving real business problems. It is difficult, and frustrating. But it needs to be done. And, this being an application system, that will be the topic of the next story that comes out of the iBelieve NY event. Stay tuned.