What’s the Deal? Battling the ‘L’ Word
October 7, 2020 Alex Woodie
It’s hard to read or write about the IBM i server or its big brother, the System z mainframe, without seeing the word “legacy” bandied about. These business systems have stood the test of time, providing value for millions of organizations, and the reward for that amazing longevity is to be called old and out of date. What’s the deal?
We have searched for other terms to replace legacy, as if the word itself is the problem. Calling these “heritage” systems is one way people have tried to avoid the stigma associated with the “L” word (we’ve have even tried out the “H” word here at IT Jungle, to mixed success). At the end of the day, heritage pretty much means the same thing as legacy: It’s something from the past.
Those labels beg the question: Why would somebody who’s looking to the future invest in a platform from the past? How can a system that was created before most users were born be capable of supporting those users’ technological needs, especially now, when IT is becoming a differentiator thanks to advances in machine learning and the Internet of Things?
As consumers, we are trained to desire the latest, greatest electronic gizmo, and a huge industry has been created to serve those desires. Social media conditions us to constantly seek new stimuli, usually from the comfort of a color, touch-screen smart phone. In 13 short years, this new class of product has reached nearly 100% saturation in the United States, Europe, and parts of Asia, which is amazing when you think about it.
We are all so comfortable living at the cutting edge of technological progress when it comes to consumer technology that it can be a disjointing experience when we encounter a business system that doesn’t seem to read our minds. “You mean the program can’t automatically anticipate my next desire? Whatevs…”
IBM’s System/38 was cutting edge in 1978, but it’s midrange line has come quite a ways since. (Image Source: IBM)
The Big Iron platforms from IBM suffer from a perception gap that will probably never close. The truth is that IBM i and System z are different animals compared to X86-based Windows and Linux servers that they compete with today. We’re not going to go into all the differences, but suffice it to say, there are substantial ways these systems differ.
The debate over the fate of these “L”-word systems mostly impacts the organizations that have already invested in them. Trying to sell these systems into new accounts in the current “digital first” age, when CIOs look to the cloud instinctively for business applications, is fraught with problems. For that reason, the battlelines are drawn primarily around the tens of thousands of companies, schools, non-profits, and governmental agencies that are running critical applications on IBM i and System z servers.
For these customers, the debate comes down to migration or modernization. Businesses need to advance their IT capabilities, there is no question about that. But what is holding them back? Is it outdated applications or outdated platforms that are preventing customers from achieving their goals?
Unfortunately, much of the blame seems to be falling on the platform these days. It has become too easy for CIOs to dismiss the IBM i and System z platforms with a wave of the hand. “They’re just legacy systems, after all. They’re too old and outdated to take us where we need to go.”
These “legacy” systems are causing problems. Take, for example, the State of New Jersey, which has struggled to keep up with unemployment claims as a result of the COVID-19 pandemic. Earlier this year, New Jersey Governor Phil Murphy put out a call for COBOL programmers to help maintain the mainframe system that processes the claims.
“Literally, we have systems that are 40 years-plus old, and there’ll be lots of postmortems,” he said. “And one of them on our list will be how did we get here where we literally needed COBOL programmers?”
Article after article reported Governor Murphy’s embarrassing situation of being stuck in a pandemic with a 40-year-old computer. News editors dove into the archives for images of those 40-year-old systems: CNBC showed with a man at a terminal in the 1960s, CNN Business had a black-and-white shot of a suited gentleman working a giant console with reels of tapes in the background, while PC Mag’s story had a picture of ancient 2966 Model 30, a type of mainframe from British computer manufacturer ICL. Slate ran a similar image in its story.
CNN Business suggested a mainframe like this was responsible for the slow-down in unemployment claim processing in New Jersey.
If the State of New Jersey was indeed using one of the ancient systems depicted in those images, there would be a big problem. However, the odds that the Garden State’s unemployment agency had not upgraded its hardware since the Johnson Administration remains quite low. (Wired arguably had the best take on the problem, and quoted a New Jersey official as saying that the unemployment system at issue was composed of various pieces, including Web servers, databases, application servers, and, yes, mainframes.)
Rich Berman, who handles PR for Rocket Software, a provider of modernization solutions for mainframes and IBM i servers, gets annoyed at tech reporters who paint too broadly with the legacy brush.
“When talking to reporters who had written negative stores about legacy systems, a lot of them were under the impression that literally the state was using 50-year-old machines,” Berman told IT Jungle. “We had to explain, no the computer itself is six-month’s old, but the underlying technology is the legacy part. And they just were envisioning some sort of steampunk 50-year-old room-size computer in a basement somewhere.”
In an attempt to cut through the legacy clutter and get some data on the matter, Rocket Software recently tapped IDC to conduct a study. The study involved investigating the costs associated with application modernization on IBM i and System z versus migrating an application to a new platform, as well as the resulting levels of technological and business agility of the companies that made those moves. We covered this report two weeks ago in our story “Modernization Trumps Migration for IBM i and Mainframe, IDC Says.”
Lo and behold, the companies that migrated off the IBM i and System z platforms not only spent more money than their counterparts who modernized their applications. But they also reported greater technological agility in implementing next-gen IT projects like machine learning and IoT. Those results ran counter to the prevailing wisdom that anything running on a “legacy” system was inherently inferior.
Another blast from the past: The IBM System/34 is truly a legacy machine at this point in time.
Chris Wey, who runs Rocket’s new Power Systems group, says the IDC study validates what his company and others treading the IBM i and System z waters have been saying for some time: That there’s nothing inherently bad about these machines, and they can be as modern as you care to make them.
“It’s one thing for us to say it,” Wey says. “Now with the help of IDC and the independent analysis of going to those users, it’s really just more ammunition to [back our view].”
If you fail to keep your IBM i and System z mainframe environments up to date, then they become old and stale. Nobody wants that. But if you put the time and effort into keeping those systems current and ensuring that they meet business requirements, then they remain agile and new.
There are things that IBM i and System z servers can do that are compelling and awesome, and there are aspects that are less-than-awesome and require some extra care. There’s no free lunch in IT, and the current hoopla around running all business apps in the cloud simply doesn’t jibe with the real world.
“The reality is that cloud environment and a mix of public and hybrid and distributed systems and SaaS services — that comes with its own problems around security and where the data is and compliance,” Wey says. “Those are problems we’ve been at for decade on legacy systems and have solved.”
At the end of the day, “legacy” exists primarily in people’s minds. It’s an easy way to refer to platforms that have been there and done that for decades. That history of technological longevity may not be high on your list of priorities at the moment. But for organizations that exist in the real world, such a track record of success is something to be proud of, not to hide from.
RELATED STORIES
Modernization Trumps Migration for IBM i and Mainframe, IDC Says
The 1980s Were Great, Just Not for Business Computers, Apparently
Battling the L Word is a great column. Thanks Alex.
I would just add that Unix and Windows predates OS/400, Linux started right after that, C and RPG in same time frame. All have been continually updated. No one else talks about legacy. There’s a reason. It doesn’t actually have to do with legacy. It has to do with perception that cheaper software and hardware is to be found than IBM products. It’s just a perception. In the end the Oracle and Microsoft and SAP solutions of the world have their own exorbitant expenses and are just as proprietary and just as updated legacy as IBM products. But C-level execs are consistently lured by the promises of cheap replacements. Legacy is the excuse, not the reason.
Instead of “Legacy”, I like to think of the IBM i as “Proven Technology”. Does that work?
Prime example of the Fake news… so if this is how they report regarding the IBM systems, what makes you think they are reporting everything else accurately??