Thursday, 8 January 2015

The Leap Second - 1 Second or 1000 milliseconds?

Given that Rs. 144.00 is charged of a mobile customer in a day, for usage of talk-time minutes, every mobile provider can lose Rs 60,000.00 for every lakh of customers but the other way of looking at it, is the 'half-empty or half-full' theory - every customer is likely to pay 1/6th of a paisa for a second not earned!

So, how does one look at the extra 'leap second' in 2015? If I charge $ 0.33 per minute for my work, does that mean that I can charge $ 0.33/60 extra on that one day when the leap second will be adjusted? Of course, it may seem too much ado about nothing when applied in an individual context but when stock indexes work on such minute deltas as margins to work on, the resultant 'gaps' can be a busy, merry dollars that may leap around during that one second!

Do economies include the full second and adjust it as 23:59:60 or should countries go Google's way of spreading the second across the full day? The latter may just not be productive for services dependent on time but at the same time, will governments look at regulations over the second? Because, after all, the government is not going to increase wages for the extra second nor will workers work the extra second.

So, what exactly is the extra second's worth?

This is where the 'half-empty or half-full' theory comes into play.

Is it like a gift to all the mobile providers to make that extra buck if they manage to squeeze out the extra second of talk-time or internet usage from the customer's pocket? Because it is not necessary that every customer talks during that extra second as it is not necessary that the providers should or can ignore the possibility of ignoring the eventuality.

But what about the overall cost of implementing the adjustment into algorithms or web applications that calculate the user's talk-time? Will it be worth the while for the service providers to invest into accommodating the change in time for one second?

Databases will have to adjust the time stamp so as not to 'trip' over the extra second and cause either dirty data or loss of integrity in data. Will the easier way be to stop all data operations for that extra second?

More importantly, what is the need to accommodate the extra time? Does it make business sense to do so?

It is not like Y2K, where the adjustment affected a whole year or at times, a whole many number of years so investing into the algorithms made immense sense! And, even if operating system manufacturers oblige by providing a Leap Second update for all their installations across the world, application logic may not find it all that simple to sync with the computer clock of different time zones so easily because it must also have to conform to any audit controls of a region without which there may be violations of money laundering on a mass scale.

The 'Leap Second' will be a huge test of maturity in governance for countries and in management sense for organizations to know how to ensure audit controls are in place to be able to track any money laundering done during the period.

For the technical side of it, it may be noteworthy only for database vendors and time-critical applications to take care of the second.

Will the web servers and the Cloud follow any convention especially when many servers are physically located in different locations that obviously will have their own anti-money laundering laws?

It will be interesting to know how governments respond to the extra second in 2015. 

No comments: