Monday, March 28, 2011

Control, choices, and options!

It was a cold and stormy night! This is not a parody of some dreadful work of fiction – it was true and I was no longer in control and I had few choices. For the NonStop community, this may hit too close to home …

It’s taken some time, but I have finally settled back into my Boulder home and the daily routines are being re-established. However, I made a quick dash back to Northern California and to the Bay area to participate in track event, but was caught out badly on the return trip when an unexpected weather change brought with it a fierce snow storm.

As the full fury was unleashed on the Sierras I became trapped for many hours on the border with Nevada, along with every other driver on the interstate at that time. I did get the opportunity to drive my all-wheel-drive Skyline, but it proved only partially effective as the summer tires struggled for control – a most uncomfortable situation. So much so, that the only real option I had was to return to Truckee, California, and ride out the storm. The picture above is of what the car looked like the following morning as I filled it with gas, outside Reno.

Sitting in the hotel lobby gave me the opportunity to scan newspapers and it was hard to miss the headlines surrounding HP. Ever since HP’s CEO, Leo Apotheker, unveiled his strategy for the company there’s barely been a day when some aspect of it did not attract the attention of one journalist or another. Pursuing the cloud, embracing mobility, beefing up its software business and stepping up to the plate to deliver services around everything, so as to ensure a much better customer experience.

These announcements haven’t come without a little comic relief. Pursuing growth the way HP’s CEO intends will come at a cost for others already in the marketplace and the angst we have seen coming from some of them has led to some executives making unfathomable decisions. The good thing about these arbitrary decisions and the posturing that accompanied them, customers remain firmly in control, despite the prevailing conditions, and there really are plenty of options out there – IBM must be chuckling in the wings!

Earlier in the week I had received an email from Jim Johnson, Chairman of The Standish Group, who had posed the question “does supporting older technology (imply you) are right-headed or wrong-headed? Is it investment protection or long-term disaster?” Even more strikingly, Jim then suggested “wouldn’t money spent on enhancing old technology be better spent on modernization or in reducing customer cost?”

There’s not too much IT managers feel they control these days, and as they sit through one meeting after another it must get a bit depressing after a while. For many of them who have been in the industry as long as I have, there’s a mindset of just sticking with what they have and riding it out till they retire. Tough perspective, perhaps, but nonetheless, somewhat scary as sticking with old and expensive technology may only hasten their push into retirement.

A few days later I had a call with Baldwin Hackett & Meeks, Inc. (BHMI) founders, Jack Baldwin and Mike Meeks. Readers may recall that in the post of October 8, 2010 “Remove the warnings!” (one of the most widely read posts in all of 2010) I had written of BHMI developing an alternative solution for customers running ACI’s XPNET. What I didn’t address in that earlier blog piece was how BHMI has a reputation of doing really hard things, often under trying conditions and where situations called for drastic actions far outside the box! BHMI have definitely become a vendor toiling away at the coal-face of modernization and their observations carry a lot of weight with me even with all that I have been reading!

While BHMI still “takes on big, complex custom projects,” Jack Baldwin explained, “the corporation’s primary focus is on its Concourse Financial Software suite, a collection of high performance near-real time back office applications.” The product BHMI used to replace XPNET was Concourse - Transaction Messaging System (TMS) and Jack was quick to add how “it’s the latest addition to the Concourse family, which transcends the back office environment.”

“As a lot of people have noted, after ACI announced its new relationship with IBM,” Jack continued, “BASE24 clients have started considering their options. Concourse - TMS does give current BASE24 customers increased flexibility for messaging choices, going beyond what XPNET users might be considering; Concourse - TMS goes beyond NonStop and the financial services marketplace!”

In talking with Jack and Mike and about the success they were having with their Concourse product suite, the conversation returned a couple of times to their observations about the NonStop marketplace. Recognizing that there where customers giving serious thought to migrating off the NonStop platform BHMI responded to this perceived requirement by engineering Concourse, including Concourse – TMS, to run on several platforms including Windows, Unix, Linux in addition to NonStop.

While BHMI is in discussion with some NonStop customers to add new functionality on platforms complementing the HP NonStop server, both Jack and Mike added “contemplating moving completely off NonStop is no longer a priority for many of these customers. We have found a cadre of customers who absolutely will not leave NonStop and, quite the opposite to what we expected, aggressively pursue other software solutions should there be any attempt to drop support for NonStop!”

With the NonStop product roadmaps becoming better understood and the migration of NonStop systems to more modern Blades packages, there are less concerns being raised by customers of NonStop – they can clearly see for themselves where the technology is headed and are electing to make the appropriate NonStop platform for their needs. Yes, they do have real choices today.

“I think one of the biggest reasons customers are leaving NonStop is that they view it as old and expensive technology; in letting them stay on the old technology increases both their company’s costs and risk,” Jim Johnson observed before adding “it also reinforces their company’s perception of NonStop being old and expensive, because for these companies (unaware of the options and choices they have), it is!”

Spending time talking with the founders of BHMI and listening to the experiences they have accumulated by being in business for more than 25 years, then yes, they will make sure their products run on multiple platforms. They will continue to support NonStop as they have customers who will not quit!

The success of NonStop continues to be in our hands. However, just how well informed are we keeping our management? How strong an advocate have we become within our companies for the NonStop platform? Considering there are many enjoying a successful career by doing this, shouldn’t we be making modernization a priority!

Rather than sticking to the “TAL and TACL Till we’re Terminated” theme, we can embrace the modern languages and programming frameworks and gain the skills that will let us remain productive for years to come! Vendors are demonstrating strong commitments to NonStop, shouldn’t we, as users, be doing the same, in lock-step?

Modern applications developed with no knowledge that the target platform is NonStop, and yet inheriting the traditional NonStop properties of availability and scalability, are becoming more widely known within our community. The ease with which this can be carried out is “firing-up” many within the NonStop community, all to the benefit of the vendor community who continue to pursue development of solutions for NonStop.

High in the Sierras this past week, my thoughts were solely on survival and while that too is the sentiment of some within our community, it is a bit of a shame. Yes, there are choices and yes, there are many more options – but isn’t it an encouraging sign for all of us that we do retain control and that facing the mix of platforms we have today, there continues to be robust offerings for NonStop?


MartyTInOZ said...

Regarding TACL and alternate non-Tandem scripting languages, just try using PERL! You can NOT use it on the Guardian personality (unless you try and use an old ITUG version) and the OSS PERL is almost as old as Larry Wall (sic)! Perhaps NonStop (NED) should rehire some of the folks they've laid off so that an important/powerful scripting language like PERL (a really recent version like 5.10 mind you) is REALLY supported on BOTH the Guardian and OSS personalities.

Harry Scott said...

As usual, another terrific post…
Keep it up!
Our experience with NonStop customers echoes what you and Standish Group have been seeing.
Customers see the NonStop as expensive and proprietary until they start to avail themselves of the open technologies that HP has enabled for the platform. As Marty Edelman, formerly of The Home Depot, has said many times, “They have a NonStop but they are running [and blaming] a Tandem [for being expensive] when it doesn’t have to be that way.”
By our estimates, HP has invested 100’s of millions of dollars in making the NonStop capable of supporting many of the latest industry standards and being able to use these capabilities absolutely will save the customers money – in most cases, huge amounts of scarce money.
The gating factor in most customer’s adoption of these new technologies is not desire, lack of talented / committed staff or even budget, it is a pure technological chicken-and-egg issue of having an inflexible, non-industry standard database that is holding them back. The net-net is that Enscribe is the anchor that is preventing “Tandems” from becoming NonStops.
In the past 15 plus years we have seen and helped many customers overcome this massive obstacle with some sophisticated but intriguingly simple technology for replacing the database as the first step. Getting the application database in proper shape before starting more structural enhancements such as user interfaces, is analogous to shoring up a building’s foundation before repairing it’s exterior windows, roof and siding. Step one has to come before step two, it’s true in software and in construction.
Unfortunately, some NonStop customers are being encouraged to “mask” their Enscribe data by front-ending it with technology that attempts to make Enscribe look like SQL. Richard Carr, my business partner, calls these attempts “Lipstick on the pig”. No matter how much any tool tries to dress up a non-conforming database, it’s still non-conforming and still limiting the organizations from truly modernizing their environment. It’s not that these tools are useless, far from it. They are great at aggregating data and transforming foreign data sources that cannot be modernized but because pigs exist, doesn’t mean you have to dress up yours!
There is a simple test for modernization projects…If you can’t start using all the new technology for software development, such as Eclipse, there’s no arguing, you’ve still got a pig!
Eclipse and all of the rest of the 1000’s of tools and technologies that are available for people using industry standard database, will never be available to those that refuse to actually rid themselves of Enscribe. It’s not a bad file system, it’s just proprietary. No self respecting open tool will ever support the Enscribe proprietary database with its convoluted data structures, nor should they. Doing so, they too would become entangled in the web of proprietary support and expensive maintenance our customers are trying to rid themselves of.
So do keep up the good work and please help the NonStop community to be aware and before “masking” their database. Encourage them to take a hard look at curing the root problem, rather than applying a masking lipstick, only to leave the real problem to fester and continue to have the NonStop be viewed as expensive and proprietary, when it’s simply not necessary.
Wishing you and the NonStop community long life, good health, fulfillment and prosperity!

Anonymous said...

@MartyTinOZ, the Guardian PERL was ported by Carl Adler years ago. If I remember correctly, it was contributed to CPAN. As with most things open source, if you can find a developer to work on it, the port could (and should) be updated to reflect the newer PERL versions. Carl himself has been away from this stuff for a long time.

If I had the time and C skills, I'd be tempted to take a crack at it myself, unifying both the OSS and Guardian implementations on the same base version.

(Is anyone in HP watching? Here is an opportunity...)

MartyTInOZ said...

Anonymous: I have the C skills, and though I'm not PERL guru, enough skills to test build and test PERL (I used PERL at my most recent employer and I keep PERL on my Linux system up-to-date with its git repository on a weekly basis just for "fun").
My point is, and always has been, if NED (and HP in general) was/is REALLY serious about NonStop as a product line, then they'd fund these woefully-behind technologies: by virtue of the fact they haven't, then, they're not.
What is needed is a real commitment to NonStop by HP (i.e. in action rather than just words).