Dear Green,

In the early days they do binary mathematics on paper. After they manufacture electronic machines, they try make these machines do the same calculations but on hardware.

They face the problem of decimal point, there were two choices for decimal point, the fixed decimal point (fixed point) and moving point (float point), look we don't have operating system yet we are in machine code.

In the fixed point there are the same quantity of integers and fraction digits for inputs, and output. The decimal point is fixed too, nothing change, the output must be contained into these digits separated by fixed dp.

0000.00000 input

0000.00000 input

__________

0000.00000 the result

In the moving dp (floating point) the decimal point could move to left or write depending on the result, so we can get a long fraction (very small) or a short fraction (bigger).

Micro Soft in the early days of VB, the machines were weak, so they use many kinds of variables with noisy names like integer, long, single, double, decimal and many other.

I think (maybe I am wrong) the guy who ask the question mix double from MS and float from the meaning above, he trying to ask how much smaller fraction can LB handle. The idea of Stefan was the best.

For you, if you are not going to make programs in assembly or machine code the above information is sufficient. And don't worry our chief scots Carl and Stefan maybe others will take care of modern variables in LB5.

If you are young man the future is yours, I hope you will see an amassing developments in high level programing languages.

Sincerely yours

Sarmed N.

________________________________

From: green8819 <

green8819@...>

To:

libertybasic@yahoogroups.com
Sent: Friday, June 29, 2012 3:14 PM

Subject: [libertybasic] Re: big number

"...what is the biggest double floating point number Liberty Basic can handle?" was the question.

Frankly, I didn't know what was "floating point number". I think I know a little now, after some search.

The example I gave last can be considered as answer for INTEGER numbers (eg: 11, 275, 384094585). At least I know now that LB can display astronomical numbers.

A floating-point number is a decimal number.

(eg: 1.1, 27.5, 384.094585)

How much is 22/7 ?

LB gives 3.14285714 as answer (8 decimal digits)

So probably 8 decimal digits is the answer to the question. It's a measure of precision. A 64-bit computer gives a higher precision.

'----code start

a = 1

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.01

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.001

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.0001

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.00001

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.000001

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.0000001

print "a ";" = ";a

print "a/3";" = ";a/3

print

a = 1.00000001

print "a ";" = ";a

print "a/3";" = ";a/3

print

END

'----code end

[Non-text portions of this message have been removed]