‘Twas a Three Dog Night

So it came to pass, back in my old Timeplex days, that I was conducting an evaluation of the new Quad Synchronous Port (QSP) modules. In the Link family of multiplexers, it was the first use of ASICs (Applicaton Specific Integrated Circuits), so DIP switches and FPGAs (Field Programmable Gate Arrays) were eliminated. The product release that supported it had been out for about a year so that software was all ready to go. I encountered a “bug” whereby the ability to force the DSR (Data Set Ready) signal *ON* did not function. Timeplex data channels looked like a giant cross-over cable. DTR (Data Terminal Ready) at one end came out as DSR on the other side, and vice versa. Ordinarily, this didn’t present any problems.

However, I was aware of a condition that existed on IBM Mainframe Mainframe circuits, where IBM would not assert DTR unless DSR was present. The actual EIA specifications are ambiguous on this point. Did it really matter what came first, the Chicken or the Egg? Strangely, this wasn’t a problem if a FEP (Front End Processor) was was used at either (or both) end(s), or if one side of the connection was set to Half-Duplex in IBM’s channel programming. So, I wrote up an Engineering Change Request (ECR) and sent it on it’s way. The very next afternoon, I got hijacked by the Field Service Division‘s Technical Assistance Center (TAC). Here they were, working a problem with a customer that had IBM mainframes at two locations and using the new QSP modules. Yes, the same QSP modules that *supposedly* hadn’t been released for GA (General Availability). I then modified my Change Request to Priority 1, invoking a mandatory “Stop Ship” order.

I couldn’t get in touch with the H/W Engineering Lead on the project, but put a phone call into his boss, the Director of Hardware Development. That was a guy named Bruce B. Surprisingly, Bruce took my call right away (instead of blowing me off for a day or so) and asked me to shed some light on the IBM Mainframe problem. I did so, and he explained that Development Engineering would be accepting my P1 status change (also surprising). The problem, it seems, was that the Revision “A” ASICs had this problem. A subsequent Revision “B” never made it off the drawing table. Another new Revision “C” ASIC was ready to be sent out to the vendor (LSI Logic) to build but his problem was one of cash. It took about $100K to get an ASIC tooled up by LSI Logic for manufacture. He assured me that he was going to be on a conference call with both Sales and Marketing to discuss the issue. He felt the earliest we could get new ASICs built would be 30 days. I hung up the phone and updated my boss (Ed A.), the two TAC Center techs that were working the problem, and their manager.

About 8:30 that evening, I get a page from my boss. “I need you to come in early tomorrow, at 07:30. There are some things to go over, but the great news is that Development is going to get the new ASIC version built immediately”. OK, or so I thought.

The next morning I arrived early, and Ed gives me a flight itinerary to Houston on a plane leaving Tampa at 9:30. “Nothing serious”, he said, “You just have to explain the situation to the customer and the Regional Service Manager.” The customer was a bank called “Texas Commerce Bank“, or TCB for short. It was early summer, 1992, and the plane sat on the tarmac for an extra hour (due to weather conditions somewhere) before takeoff. I finally landed at Houston’s Hobby Airport. The Service Manager, Phil M., met me at the airport and drove into downtown Houston for the meeting. One of his star techs was already at the customer premises, awaiting our arrival.

Once we were escorted into TCB’s lavish conference room, I knew there was going to be trouble. Around the oval table were five people from Texas Commerce Bank, and 23 people from IBM. Phil nudged me to sit on the bleacher seats, off-to-the-side.

One-by-one, the IBM people raked me over the coals. “No, I do not have any further updates.” “Yes, the product should be available within 30 days.” “No, I don’t know why this product was shipped to the customer.” Etc, etc, etc. I was pretty much repeating myself over and over again. Finally, the twelfth IBMer cut to the chase, “I don’t care about Chickens. I don’t care about Eggs. I don’t care about DTR or DSR.  I want to know when my customer is going to get a working board from you people.” Phil intervened at this juncture and explained that when the first of the corrected boards rolled out of Manufacturing, they would be FedEx-ed directly to the customer. In the meantime, arrangements were made to have the Field Service Division build some adapter cables that forced DSR high so that IBM’s project could get underway. There were a couple of other parting shots from IBM chiding me on the “irresponsibility of Timeplex to build such a faulty product”. Blissfully, the meeting finally came to an end.

I was extremely irritated. In the elevator on the way down, Phil put his hand on my shoulder and said “Wow. You really took one for the team.”

Yep. I felt like my sphincter was about three sizes too big. He took me to lunch in the trendy Galleria section of Houston, and his staff got me back on a flight back to Clearwater from Houston’s Intercontinental Airport.

The Field Service guys gave me the cold shoulder for a while since they had to absorb the cost of the adapter cables. The “Tech Pubs” and “Educational Services” departments weren’t pleased either since they had to document the new “Adapter Cable”.

Three weeks later, I’m up in Woodcliff Lake, NJ for a few days worth of my usual meetings. On the last day, I’m joined at the Marriott bar by Bruce B. from H/W Development and Peter M. from the Carrier Marketing group. I was still sore over the whole Texas Commerce Bank issue and retold my tale. Bruce explained that it only cost them $25K for LSI to retool the new ASICs, and that they were being shipped that Friday. He was happy. The Field Service guys were a little happier also.  All of their techs were getting a special tool that would allow them to replace these ASICs in the field. Now, Bruce was a funny guy but he was also a battle-scarred, fast talking, New Jersey-ite. He could spew out technical information at a rate that makes the Gilmore Girls banter seem like a Southern drawl.  But quick with a joke, he wrapped up our beers with a story:

Last week, an Engineering guy, a Marketing guy, and a Sales guy started bragging about whose dog was the smartest. So they arranged to meet up on Saturday at the nearby Meadowlands Sports Complex with their dogs.

They had their dogs stay on the 50 yard line. The Engineer sprinkled 9 dog biscuits on the ground and said, “Watch what my dog can do.” The dog’s name was Slide Rule, so he called out, “Come here, Slide Rule, show these guys what you can do.” Slide Rule wagged his tail, raced down to the end-zone and aligned the nine biscuits into a perfect Right Angle.

The Marketing guy said, “That’s nothing. Look at what my dog does.” His dog’s name was Spreadsheet. “Come on Spreadsheet, do your thing”. The dog happily yelped, pranced down to the end-zone, and arranged the nine biscuits into to three columns of three biscuits each.

The Sales guy swaggered up and said, “You guys are idiots. Allow me to demonstrate.” His dog was huge, and aptly named Expense Account. He yelled to his dog, “OK Expense Account, show them how it’s done.” Expense Account leisurely walked down to the end-zone, screwed the other two dogs in the ass, and ate up all the biscuits!

The end.