I have a guilty and shameful confession to make.
It happened years ago, and it’s been troubling me ever since. I didn’t understand my mistake until recently. It’s a little hard to explain, so bear with me.
Sometime around 2007, I happened to be at the old DISA Skyline-7 building, when a project called “Button Two” was unveiled.
Button Two was part of an acquisition program called Net-Centric Enterprise Services (NCES). NCES had many piece-parts, but one of the major functional areas was “collaboration“. The Director of DISA (now retired) had decided to pursue a fairly clever, “two-button” acquisition strategy – essentially fielding two different collaboration services, and allowing the users to choose which one they liked better.
When the DISA personnel first demonstrated “Button Two”, it was a web-conferencing solution based on Adobe Connect, and an Instant Messaging solution based on Jabber XCP. I remember being a little confused at the demo, because the Adobe Connect web-conferencing tool seemed to have an embedded instant messaging capability, and Jabber XCP is an instant-messaging platform.
So I asked, somewhat naively, “You’re calling this a single service, but it seems like it’s two separate things. Does the chat window in the web-conference tool connect somehow to the instant-messaging tool?”
The answer was, “No. Those are two separate functions.”
I became disturbed. “Wait… you’re fielding a single service that includes two chat services that don’t even interoperable with each other? There was already a chat capability in ‘button one’. Does the either of the instant-messaging/chat tools in Button Two bridge to Button One?”
The answer was, “No. Those are two separate systems.”
I asked, “But some of the combatant commands are already using Button One. How are they supposed to communicate with people who use Button Two? Won’t that be confusing and decrease the value of the whole system? Why don’t you build a bridge between Button One and Button Two?”
And the answer was, “There’s no requirement for that. If users need to communicate, they need to be on the same system.”
It got worse: “Hold on…,” I say, “one of your other services is the User Access Portal, Defense Knowledge Online’. DKO is just Army Knowledge Online (AKO) re-branded. AKO has a chat tool too, based on Bantu Instant Messaging – which NCES is thus paying for indirectly. Is Button Two going to interoperate with AKO IM?”
“There’s no requirement for that.”
By now, I’m on a roll… “And NCES pays for Intelink to run the Enterprise Search service. They’ve got an IM tool too, called ‘Intelink Instant Messenger’ (IIM). Can we bridge to that? Is that interoperable? I know IIM is based on the same XMPP technology you’re using – I think it’s even Jabber XCP, the exact same software.”
“There’s no requirement for that.”
Me: “It’s all just text messages…! How hard can it be to bridge the systems?!?”
“There’s no requirement for that.”
Here’s the part I’m ashamed of: I was disgusted. I shook my head and I walked away, grumbling about incompetence and lack of vision. I walked away and let it happen.
In my previous post, I pointed out that Title 10 says the proverbial buck stops with the SecDef. Now, I ain’t the SecDef and I won’t ever be the SecDef and I don’t speak for the SecDef. But I do work for the SecDef, and if I had the insight to see what the right thing to do was, it was my responsibility to get it done. I was not responsible for oversight of NCES, and it was easy to feel like it was someone else’s problem.
I’m naturally conflict-averse, and I was even more so back then. I let it go, and I’m sorry. I am trying to cultivate the courage to change the things I can, and the wisdom to know the difference, as they say. It is not always easy. I will breathe deep, learn, and live in the future, where I can do better.
Postscript: This also illustrates my previous post about CAIV, in the sense that, if we had had a more CAIV-like approach to NCES, then it would have been harder to hide behind a wooden repetition of the phrase “There’s no requirement for that.” If the requirement was, “Provide great collaboration tools,” we could have done the right thing without a risky, 24-month trip through JCIDS.
[update] Post postscript: As it turns out, the first disconnect I noted – the one between button-two chat and button-two web conferencing embedded chat – actually doesn’t make any sense. The other three did make sense, and the last one still does.