Mr Deka's Nightmare – Part 3

Click for Part 1 of the Story
Click for Part 2 of the Story
That night the three of us did not get to bed as Anirban took apart every card, every peripheral and reassembled the machine and the cycle of testing continued. I learnt to use a multi-meter that day as I helped Deb and Anirban test every chip on every board. In the early dawn pinkness, the computer had been re-assembled multiple times and Deb’s iterative additions were still running with the occasional burp.
When a wan-looking Mr Deka came in, we were forced to concede defeat. The matter had to be referred to R&D at Head Office. We could do no more. Deb and Anirban flew back to Calcutta. My 5 installations at the bank kept me in the city and Mr Deka took to checking in with me for news.

Out in Bangalore, the R&D crew dedicated to the problem huddled and heated discussions were presumably held. The apparently arbitrary nature of the imprecision did not provide any pattern for a line of attack to be created. Into this little maelstrom of activity appeared one Kailash, who happened to be passing by the room on his first day back after a spot of leave and was attracted by the hubbub. And Kailash it was who cracked the problem. Apparently, the operating system lost precision when the logical storage used to hold the figures crossed the physical boundaries of the the memory chips.
This was tested and an OS patch produced and Bani and John were dispatched to install and test the solution. I was there too as the tests were pronounced successful after millions of iterations from John’s programs raised nary a burp. Mr Deka, smiling from ear to ear, pronounced himself satisfied and I left with Bani and John for a satisfying lunch. Shortly thereafter they left for the evening flight back to Calcutta.
I was in my hotel room at 6:00pm just beginning to start on my cup of tea and paneer pakoras after another day haranguing assorted local air-conditioning technicians, carpenters and electricians in a fruitless pursuit of site readiness for my 5 bank installations when there was a hurried knocking on my door.
“Mr Sharma, it is happening again!”, Mr Deka burst into the room.
This was the days of telexes and I had to run down and shoot out telexes to Calcutta after I had calmed Mr Deka down. The mood in Calcutta was grim and I suppose strong words must have been exchanged between the sales guys in Calcutta and R&D types in Bangalore. Tempers had been severely tested over the last few weeks and I too, far away from the action, but the one nearest to the customer, had been wondering when I would see an end to the regular visits from Mr Deka. The paneer pakoras, probably the only item on the menu of the Hotel Nandan that I found edible, turned to ashes in the presence of Mr Deka.
There was a happy ending however, a week later OS v 7.2 was sent out to replace the abortive v 7.1 and Mr Deka’s imprecision was finally laid to rest and so ended a stressful time for a whole lot of people.
The image of my initial reception by Mr Deka, the night spent in that computer room  and the shock of seeing a visibly disturbed Mr Deka bursting into my room barely 4 or 5 hours after he had signed off on the system and the engineers were on the flight home has endured over the years.
I wonder if Mr Deka still remembers this part of his career, Deb certainly does!