• Welcome to Valhalla Legends Archive.
 

Trinary Code

Started by MetaL MilitiA, December 30, 2004, 01:08 PM

Previous topic - Next topic
|

MetaL MilitiA

Quote from: rabbit on June 16, 2005, 10:35 PM
Quote from: MetaL MilitiA on June 16, 2005, 10:02 PM
Now, let's review! I prove my theory for multiple binary channels wrong before anybody else does, yet other people are still trying to prove it wrong, like you! So uhh, don't tell me to drop it, I've totally dropped even the possibility of it a long time ago. You guys drop it. You're the ones that keep comming and trying to prove it wrong again when it's already been proven wrong.
1. We proved you wrong.  Repeatedly.
Please, just explain it to me. How the hell could you or anyone prove my 3 consecutive posts wrong when I proved it wrong myself before anybody else posted? Also, how could you even involve in proving me wrong when nothing you said is even relevant to anything that I said in those 3 posts (which might just be a sign that you're too damn stupid to comprehend it in the first place)?
Quote2. My last post was May 11.  You came June 15, more than a month later, and posted again, so don't tell us "we came back and keep coming".
I guess you couldn't comprehend that either. After my 3 posts (which I already proved myself wrong inside of the third one), 3 people (Adron, MyndFyre, and yourself) came back and tried to prove my theory wrong (yes, the one I already proved wrong before anybody posted about it). Even after I explained it to you, you came back just to have the last word (which makes me wonder what your next moronic post is going to be like, because it seems that you'll just continue blowing random shit out of your ass just so you don't look bad. Oops, I guess you didn't notice that makes you look even worse).
Quote3. If you've dropped it, stfu.  YOU TRIPPLE POSTED.
4. See #3.

@Mod: Lock please.
Hmmm, let's look at something I said in my second post out of those 3.
QuoteNote: If double posting is against the rules, tell me, I'd be glad to combine this with my above post.
Just shut the fuck up already.

R.a.B.B.i.T

<INSERT MORONIC LAST WORDS HERE>

STFU MetaL, it's no wonder everyone hates you.  YOU'RE STUPID.

MetaL MilitiA

#92
Why do you think I would care if people hate me over the internet?

Statements without facts aren't taken very seriously.

EDIT: I decided for fun, I would take apart your post that has "proven me wrong" and show everyone how... I think I don't even need to finish that sentence.\
Quote from: rabbit on June 16, 2005, 04:44 PM
Trinary is only more efficient in a mathematical sense.

Not if you have a system to properly implement it into, which happens to be exactly what my last post was talking about, the impossible "perfect scenario". On a side note, you can't base trinary code off of the system it runs off of, trinary code is more effecient than binary code.

Quote from: rabbit on June 16, 2005, 04:44 PMActually determining the values would degrade the quality severely.  The point of binary is that it's either "on" or "off", with trinary there is "on", "off", and "that place in the middle...you know the one".  Creating circuits/components which could accurately and speedily determine the state (1, 2, 3) with no margin of error would cause said degredation.

If you actually read and comprehended my post, you would decipher that I was talking about quadratic code, not binary. I also wasn't talking about having quadratic code on a single channel, yet 2 parallel channels. This means that there would still be the "on" and "off" state, yet just on 2 channels that are simultaneously transmitting combinations of "ons" and "offs" to emulate the effect of four different values on a single channel.

As you can see, your post has almost no relevance to my 3 posts, yet have no direct relevance at all to my idea of 2 parallel binary channels. This means that you did not comprehend my post, yet you still decided to say I'm wrong. You didn't know about what, but that doesn't matter, I'm just wrong, and you helped prove that!

Now, I guess you're not really an idiot, and you're probably actually intelligent, yet what irritated me is the fact that you tried saying that I was wrong about something you don't even understand.

Adron

Quote from: MetaL MilitiA on June 16, 2005, 02:33 PM
Quote
Aside from that, your calculations on the efficiency of systems as a function of the number of possible functions are flawed. You talk about storing 256 possible values or storing 16 possible values. 16 possible values correspond to 4 binary bits (2**4 == 16) and 4 trinary bits correspond to 81 possible values (3**4 == 81). Using your quarternary bits, you get 256 possible values though (4**4 == 256).
I also discovered my calculations were flawed last night, yet wasn't able to post the following. This is how trinary code relates to binary code: 1 bit of trinary (3**1=3) is 50% more effecient than 1 bit of binary (2**1=2). 2 bits of trinary is (3**2=9) is 2.25x more effecient than 2 bits of binary (2**2=4). 3 bits of trinary (3**3=27) is 3.375x more effecient than 3 bits of binary (2**3=8). Now, let's jump to 8 bits. 8 bits of trinary (3**8=6561) is 25.62890625x more effecient that 8 bits of binary (2**8=256). The number will just keep growing exponentially like that, so the performance increase will always depend on the workload.

The efficiency doesn't grow exponentially, if it does, it means your calculations are flawed. Calculate the number Y of binary bits required to represent the same number that you are representing with X trinary bits. The difference in efficiency between binary and trinary given the same resource usage for implementing a binary and a trinary bit will be Y / X, and that will be a constant except for rounding. That is what I have been trying to tell you in at least two previous posts.

The performance increase will not depend on the workload, because the ratio Y / X describes how many more binary bits you need to store a certain amount of data, or how much more time it will take to process a certain amount of data. And that is a constant!

MetaL MilitiA

That makes sense. The only way to really calculate it though would be to put 2 different sums of the possible combinations for one 1 bit of trinary (3**2=9, 3**1=3) into a ratio (9/3=3) and compare it to that of the binary (must be corresponding) (2**2=4, 2**1=2; 4/2=2) (might not make sense to you, but it does to me). Trinary would be I guess only 50% more effeicent than binary (what I originally thought, but I just had to make things more complicating).

|