It's a trivial calculation in LibreOffice.

Enter the two dates, then subtract the earlier date from today's date.

83251 days, some 228 years.

Not sure if your objective is to use a cli function specifically or just learn how to do something, but if you just need to make the calculation, LibreOffice is a very easy option.

Couldn't help myself! My prediction based on the past data is a release between April 17 2013 and February 11 2014 with a 99.7% certainty.

allend,

It seems I’m slightly more cautious than you because my guess which bases on the previous results is: no less than 7 days and no more than 2 years after Slackware 14.0 release:

date --date="September 28, 2012 +7 days" +"%Y-%m-%d"
date --date="September 28, 2012 +2 years" +"%Y-%m-%d"
Certainty: 99.99%.

My sister left on vacations some time ago, not before telling me she would come back in about three weeks. Knowing the date of departure, and today's date, I wanted to know in how many days I should expect her arrival. That was all. But firs I had to do 28 + 14 to find today's date with respect to february (assuming february is infinitely long), and I wondered if there could be some linux command ad hoc. The answer is just 'date'. It does all the work. By the way I found a new linux command: calc.

cal

However you look at the data, 520 days should be considered an outlier:

Looking at your analysis it seems that the most probable is the range 400–499.

date --date="September 28, 2012 +400 days" +"%Y-%m-%d"
date --date="September 28, 2012 +499 days" +"%Y-%m-%d"
It isn’t my prediction.

I prefer 99.99% of the practical certainty as I wrote about that in the post 33.

(It seems to me that the term “practical certainty” was coined by Descartes.)

This is balanced by the lower extreme of 7 days. Statistically speaking, there is no valid reason to drop either or both data points.

There is also a trend to a longer time between releases.

Next Slackware release will occur sometimes between now and never with 100% certainty.

No algorithm was hurt to make this prediction.

Not one release took 400-499 days. The dot frequency chart suggests that the time between releases is more like 200 + 200 days.

You're right. If you ignore the two outliers, then a straight line regression would give a high correlation. From visual inspection we would be looking at 200 - 400 days.

16 occurrences in the range 0–199 and 17 occurrences in the range 200–399. So none of these ranges is privileged. On the other hand the older releases periods are usually shorter (or very short) and the newer releases periods are usually longer (or very long). So there’s the tendency to prolong these periods. The next range after the mentioned ones is 400–499 and it’s empty so far. According to the overall tendency sooner or later some releases will appear in that range as well. I assume that it’ll happen rather sooner than later. That was my point from post #37.

***

In fact to be fair I should take into consideration the range of the equal size as the previous ones: 400–599. It’s almost empty (one occurrence). Sooner or later more releases will appear here. I say: rather sooner than later.

Totally agree. Each time someone asks me how to calculate days
between 1784 and today (yesterday an old lady at greengrocer's
asked me exactly that), to calculate the beauty of a Leonardo
painting or make some prediction based in numerology, like "When
a big asteroid will impact with Earth and put an end to all this
idiocy?", I advice to install LibreOffice. Compile it from
source for a better result.

(Just to remember people that Microsoft Office Suite is
indispensable, essential, irreplaceable software).

Nice program, troyaine! Thanks.

Using the C program made by troyaine I got 734942 days from 01/01/0001 to 03/14/2013...???
What is the catch?

The red line is the trend obtained from simple linear regression. The X axis is voodoo, and there is no reason to suppose a linear model would work here anyway, but if it did, then the next release is some 340 days after 14.0.