Site Sponsors:
Intergalatic Julian Dates 

It's A Date!

As you might imagine, someone who writes robots designed to automatically mange opt-in content or daemons - designed to run unattended for decades - might have some relatively interesting time and date calculation requirements. Indeed, when tracking appropriately-recurring day-content containing holidays, events, and other falderal is more important than hours, many folks like to use Julian Days.

Day Julian?

The problem is, whenever we try to universally define exactly *what* any computationally useful encoding might look like, opinions can differ wildly. While some insist that 'day zero' is important, others disagree. -Speculation also rages over just how far back, and forward, an official year and / or day-count might reach. Rather than haggling over the exact time of a technically unmeasurable origin, how about coining a more event-relative time-reckoning?


Living close to NASA, like others we have always been inspired by the mathematical possibilities of the "T - 1" form of event-counting. -Being computer geeks, we also recognize that using negative numbers can also have some decidedly plotable graphical & computational advantages.

Concepts Encoded

For the above (and other) reasons therefore, we decided to come up with our own definition of what a better Julian Date should look like.

-Negative Years

Honoring the mathematicians of The Roman Empire, only "Day 1, Year 1" will be encoded in event origination; Honoring the needs to align both computational & graphical symmetry, years before "Day 1", are negative. (We feel that using negative and positive numbers might be a better way to mathematically express concepts such as "BC" or "AD"?)

-Counting Days

Next, while incrementing / decrementing days AFTER an event are logical, things like decrementing the time BEFORE an event must be logically identical, yet mathematically inverted. (Time, after all, indeed has a somewhat forward bias?) Hence increasing or decreasing a pre-event calculation should ever maintain the same linear motion as do those same operations in the event's positive complement.

Consequently our new Julian year # day series before & after any event is:

... -1#3 ... -1#2 ... -1#1 ... [event] ... 1#1 ... 1#2 ... 1#3 ...

-Leap Years

For 'galactic reasons' (not all planets share the same number of days in their year after all!) note that the number of days & years around any event are completely customizable. -For the sake of conveyance to us 'Terrans however, by default the number of days in a leap-year remain as presently calculated.

-The Potentia

Finally, across all celestial bodies, adjusting for leap-years remains a somewhat arbitrary process. Because some years are allowed to have more days in them than others, we refer to the potential maximum number of days across all years as being the potentia. By way of example, while the usual number of days for our orb is 365, the potentia for Mother Earth is 366.

(While one might be tempted to apply Earth's n+1 formula to calculate the potentia everywhere, some conditions may require the scheduling of an event every n ''years'' yet - just as predictably - routinely lasting much longer than +1 days. Because the length of any schedulable, mathematical-rounding additive can obviously vary, the potential must be customizable.)

Try It

Therefore, submitted for your consideration please find our DateJulian class. -The guiding basis for all that our robotic endeavors deliver to-you on the Internet, I hope that you might find our "Julian Day with +/- Year" event-management concepts as invaluable, as we do.

Here is where you can download our JavaDoc, Test Case, & Source Code.


-R.A. Nagy

[ view entry ] ( 3699 views )   |  permalink  |  related link
ClassIO 1.2 Released 
Just a quick note to let everyone know that we just released another version of [ClassIO]. Still a Java tool, but we plan to create one for C# as time permits.



[ view entry ] ( 2057 views )   |  permalink  |  related link
Generic Object Placeholders in Templates 
I am reminded that it is good to remember that we can use templates to take advantage of "isa:" In C# and Java specifically, since all classes are reference types (object pointers), the universal object-placeholder is Object. From a template's point of view, Everything 'isan' Object.

What does having a universal Object for absolutely all classes, mean? -Well, much like using "void *" in C/C++, Java and C#'s Object offers us all of that polymorphic goodness, as well as some sophisticated base operations.

To demonstrate the former, given the following template + interface combination:

public interface OmniAdapter<T> {
// Adapter: Convert one object, to another
Object xForm(T ref);

-We can support a simple child 'Object'-conversion:

public class Event3 extends Day2
implements OmniAdapter<Day2> {

public MyNewClass xForm(Day2 ref)
return new MyNewClass(ref);

What is interesting to keep in-mind here is that, because MyNewClass is-an Object, that we can xForm one type, to another. -Even if MyNewClass is only used by a single, solitary function - as long as the compiler can find it - then we can return it.



[ view entry ] ( 2608 views )   |  permalink

<<First <Back | 81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | 89 | 90 | Next> Last>>