Jump to content


Sensei

Member Since 04 Oct 2013
Online Last Active Today, 01:43 PM
***--

Posts I've Made

In Topic: Trying to pick a good laptop

Today, 01:42 PM

 

Yes, that's true.  I have also had a very good experience with a Samsun Series 9 unit (Samsung's answer the MacBook Air - very similar form factor, good but not quite as good battery life, etc.)  I used it for a while with Windows on it and then for a longer while with Ubuntu.  I managed to crack the screen and that  made the touch screen misbehave badly enough to make the machine unusable, but I was able to disable the touch screen and it was still perfectly usable.

 

Did you try to get replacement screen?

https://www.google.c...lay replacement


In Topic: Do you think, you can magnetize a rotating disk with the centrifugal force???

Today, 12:37 PM

Use the same method to magnetize the needle with silk, fur or hair. Pull the needle 50 times over the object to magnetize it.

 

This sounds to me like method of electrostatic charging..


In Topic: Just an idea about AIDS cure.

Today, 12:06 PM

We can't change the CD4 (which I guess you are aiming at) in our T lympho's, since then they simply won't work as desired anymore.
 
The way HIV infects and destroys CD4-positive T-lymphocytes is by their CCR5-receptor.
 
If you find a way to modify T-cell stem cells to lose their CCR5, be my guest. Fun fact: around 1% of the total population is naturally CCR5-negative, so they won't be able to develop aids from HIV.


How about making "fake modified T-cells", which have CCR5, like original one, but everything else in them is disabled.
So once they are introduced to human body (drip), they are capturing majority of viruses from blood.
Later they can be identified by unique part on surface (due to modification), and expelled from human body.

In Topic: Gauging the scale of impact on large prime factorization

Today, 10:56 AM

You don't understand.
There is nothing to skip in internal loop (except "2" in entry list[0], because I incremented i+=2 in outer loop (instead of i++), in "optimized version").
Computer does not know whether f.e.
i=%1 0010 1011 1110 0101 0110 0001
is dividable by 5 or not,
until cpu will reach line
if( ( i % 5 ) == 0 )
5 is in entry list[2], reached very quickly.


Instead of using built-in type "int" programmer can make custom class Integer, with overloaded operator of modulo,
and implement the whole math dynamically.
If it's in range 32 bits, use unsigned int,
after exceeding 32 bits, use unsigned long long,
after exceeding 64 bits, use dynamically allocated buffer with size limited only by free memory.

In Topic: Gauging the scale of impact on large prime factorization

Today, 04:35 AM

BTW, I think you can skip every other 5th integer as well since they all end in a 5. i.e. all primes > 10 end in 1,3,7,9 only. I guess that means every 5th iteration after the odd check.


It's easy for human because of using decimal system, but computers use binary system, and easily they can skip only 2,4,8,16,2^n etc.
f.e.
255 in binary is %11111111
250 in binary is %11111010
How do you want to skip numbers ending up 0 or 5, without using modulo (at least)?

i modulo 5 will be 3rd in internal loop (2,3,5,7,11,....)
and in my optimization, it'll be 2nd in internal loop.
Very quickly found in array and skipped.