It’s been about a year and a half since Neon Enterprise Software released its game-changing zPrime. ZPrime, for those of you who have been living in a cave, is software that allows any type of workload to run on mainframe specialty
A little history on the Neon zPrime, IBM saga
Neon released zPrime in July 2009 and caused quite a stir among mainframe customers. ZPrime saved money and had the potential to break IBM’s stranglehold on big iron. It also offered a possible escape from expensive independent software vendor contracts that tie revenues to processor capacity.
After the initial enthusiasm, most customers held their breath, waiting to see what IBM would do. A few weeks after Neon’s announcement, Big Blue kicked off a fear, uncertainty and doubt (FUD) campaign by issuing a letter that warns IBM customers that using zPrime and other specialty engines would violate compliance with IBM contracts.
The next couple of months were predictable. IBM quietly went to work on its customers. Catching wind of the campaign, Neon sued IBM for unfair trade practices. Neon had a point: IBM’s efforts seemed to be aimed directly at the company and at preserving a monopoly on the mainframe.
Recent developments in the Neon zPrime case
The alternative dispute resolution report was due at the end of August. Since we haven’t heard of a resolution, we must assume one wasn’t reached. According to court documents on Neon’s website, amended pleadings, along with written offers of settlement from both parties, were due by Oct. 22. Neon and IBM were supposed to respond to the proposed settlements by Nov. 5. At this point, I haven’t been able to find any word on a settlement outcome.
The good news is that the judge moved some of the court dates forward. In the likely event that the two parties disagree, fact discovery will continue until March 18, 2011, giving both parties almost half a year to interview witnesses, plow through source code and carefully parse contracts. Motions are due by the end of April 2011, and jury selection is scheduled for the following June.
Not to be caught flat-footed, Neon filed an antitrust complaint against IBM with the European Commission (EC) and started a campaign to sell zPrime for IMS at one dollar for the whole year. The only condition is the customer must have the newest version (2.1) in production before Dec. 31, 2010.
Neon zPrime will live on
I am not, of course, privy to Neon’s and IBM’s strategy sessions or legal team discussions. But ignorance has never stopped me before.
In my opinion, time is on IBM’s side. IBM has deep pockets and I would expect it to use delaying tactics to keep Neon tied up. The same strategy helped IBM beat the government’s antitrust suit in the ‘80s.
IBM would want to slow things down for several reasons. First, it can wait for customer contracts to expire and then try to introduce language prohibiting software like zPrime. Second, it will have the chance to reformulate its specialty engine policies or get rid of them altogether. Lastly, with zPrime’s source code in hand, IBM has half a year until jury selection to come up with an operating system or microcode change that stops Neon zPrime.
Neon has fewer options. Selling IMS zPrime for a buck is good because IMS, a typically CPU-heavy workload run in big mainframe shops, would give those shops immediate capacity relief. More customers and machines running zPrime will make Neon harder to stamp out, and means Neon can claim a groundswell of customer protest against IBM and persuade other shops on the sidelines to join in on the fun.
Neon is also more likely to get a sympathetic ear from the EC. You may remember that the EC took Microsoft down a couple of years ago when the U.S. justice system couldn’t.
Whatever the outcome, in a sense, the genie is already out of the bottle. In the last year, rumors have surfaced that vendors figured out the the trick to Neon zPrime. Even if IBM manages to kill zPrime, how many other products will spring up in its place? If the secret becomes common knowledge, what's to stop any shop with a moderately gifted assembler programmer from using the same exploit? This may only be the beginning.
What did you think of this feature? Write to SearchDataCenter.com's Matt Stansberry about your data center concerns at firstname.lastname@example.org.
This was first published in November 2010