MASTER VOTE MACHINE USES FLOAT INSTEAD OF INTEGER !

Try harder, CTR. We're techies here, we know why you should be using unsigned integers instead of floats.

*her website

There's more to it, I didn't had time to listen to the whole interview yet, the 11 minute clip upload doesn't explain everything indeed, she was mentioning something about, but I have to relisten it again when I will get to the whole show later, just saw nobody posted this so I did it quickly.

Drudge links to the article which has more other stuff, it's not just the float thing.

The alex jones show is known to be very tech-unsavvy, alex jones is kinda making his big huha-dupa but he probably didn't get a thing, yes I agree simply stating that it uses float vs integer is really NOTHING, but wait, I am no stranger to these radio shows, they kinda sometimes don't get the whole detail over since it's a hurry with radio time and commercials, so it's kinda all missing bits and pieces here.

The detailed video by Bev Harris here:
youtube.com/watch?v=Fob-AGgZn44

EDIT: No embedd, since file exists, someone just beat me to it.

While that's possible, int goes up to 2147483647 (2^32 - 1).

So my money's on

a 64 bit unsigned integer (you don't need negative #'s of votes, obviously) holds 4294967295 (0xffffffff) or 4.3 billion.

yep, 2.1 billion would be enough to hold all the votes I believe.

Thanks for that, the way I understood him was the way i said it in the previous post, IMO simply having float vs integer by it-self may indeed not be changing anything, wouldn't it stay 95438.0 right ?, unless they do something more with that. That something, is what I didn't know when i made the OP.

Now I'm techy, but I'm not a real developer so correct me and don't take this 100%. I knew it was big when I saw the headline, i just wasn't really sure completely after listening for 10 mins what were they doing ontop of having fractions.

Because he's trying to discredit it…

Using floats makes it easier for the voting tabulators to do the D+10 adjustment.

Float is perfectly accurate with full numbers, such as when counting the number of votes. This is perfectly acceptable though an (un)signed non floating point type (such as a 32 or 64 bit int) would be more appropriate given how the value would be used.

Float can be inaccurate when dealing with fractions. If they're doing numerical analysis then it would be a problem. Realistically this inaccuracy could occur at the 3rd or 4th decimal place. A better fractional number type would be double (64 bit accuracy) or decimal (128 bit accuracy and suitable for financial applications).

No. Floats are not perfectly accurate. Sheesh. Leave it to tech people to talk about this.

I don't see how this really matters. The programs efficiency is not maximized. This does not give any candidate an advantage.

...

I'm guessing Pajeet coding, but he forgot double is a thing.

Either that, or the voting machines are some insanely archaic shit that can't process them?

Quiet dumbass, your nigger tier understanding of tech isn't worth sharing.

You would make CTR blush.

I said: Floats are perfectly accurate for whole numbers.
As soon as fractions enter the picture, the stored data may become lossy. And again, only for the fractional part.

Wrong, it's accurate up to a certain amount of significant figures. If you use very large numbers, you end up losing precision in the units.

I read what you think is being done. Note that you could do that with integers as well. Just multiply the numbers by 10.

And you're wrong. Singles only have 23 bits of mantissa. GTFO with your disinformation.

2^64 = 18,446,744,073,709,551,616

Oh yes, actually youre right. Most of the time when I use them its for graphics programming so its usually clamped to 0..1 and also accuracy doesnt matter. Come to think of it ive never actually used floats above 10m.

Sorry.

What you chalk up to disinformation is just bad memory.

Yeah, you use shit like that in basic graphics and mathematics. The moment finances come into play, float is fucking terrible. Double for math, decimal for finance.

t. banking software developer

Do it pedo faggots. I dare you to try to rig the vote, though I doubt you'll make it through this week without being exposed regardless.

bump because I'm a retard who only vaguely understands the conversation

Float has 23 mantissa bits, that's 8 million counted accurately before it starts losing them. 32 bit int has 31 bits, that's 2 billion. Double had 53 bits, which would be enough, but there's no reason to program these in Perl or JavaScript.

...

if (batch[Hillary]>0.5)
{cumulative[Hillary] += batch[Hillary];}
if (batch[Trump]>1000000.0)
{cumulative[Trump] += batch[Trump];}
else
{for (float CtR=batch[Trump]; CtR>0.5; CtR -= 1.0)
{cumulative[Trump] += 1.0;}
};
// this code verified by Pajeet

Because there's more to it, they need floats to make advanced recaluclation to rig the elections, you need to watch the full video youtube.com/watch?v=Fob-AGgZn44

I have to say the explanation and the presentation is a bit confused it's not done by real techy people, so I finally wrapped my head around what's going on:

For example the woman says "overlay" nobody uses such terminology, she means GUI, the point is, the fractional float counting is supported by the voting machines, but the voting machines don't have the hacking software, doesn't have the GUI, they simply support it or in other words are designed with a backdoor in mind, someone then uses a hacking program on another computer entirely, but not over internet but on the same local intranet, so that's why in the end of the video the people who can do fraud are IT guys, ballot printers, the employees there who can get close to machines.

This Fraction Magic is i'm still not sure since they didn't explain it well enough, either a simulator or just the same GUI being brought up, which in the video is something the guy made from the leaked files the organization

The black guy is just a programmer that was hired by Bev Harris, he's not hacking anything, his demonstrating it.

It's also old info compiled, they know why this is real is because they have leaked emails and court testimony of programmers talking that they need fractions to make "weighed races" Now some of you here were quick, I am totally new to this so it took me some time.

Next time she needs to hire a techy to make a video, because this is such a big deal, but many people won't understand it if they just gloss over.

Holy shit, this is the motherload.

Tripdubs confirm this is real source code.

In programming, there's many ways to store data. int, short for integer, is for whole numbers (1, 2, 3…) and in theory should be used for tabulating votes, but instead in the master vote counting program it uses float. Float, short for floating point, is for fractional or decimal numbers (1.1, 1.2, 1.3…).

In addition to that, all the tools necessary to alter votes in any county or precinct are available to whoever uses this master count program.

Example of how it can be abused and why having the variables(data set) as float is bad.

Except sense its float, the numbers will actually be:

I can add these numbers together:

And then redistribute them however I like based off of proportional algorithm:

Scrape off those decimals on the end to conceal the fraud:

This would all be practically impossible if they had just stayed with an integer, int, value because there's no way to predict how many voters actually show up. Fraud has to be proportional to turnout, hence the need for fractions/decimals and float.

i hate alex jewnes, but i gotta admit
the sound effects and intro animations are pretty good.

If you don't use float how are you going to count 3/5 of a vote :^)

There was a shitposter on half chan a few months back that said basically Clinton would win the election, but that the electronic voting machine code would get released and would through the whole thing into question.

I'm starting to think he was not a larper.

What the fuck?
I cannot think of a single legitimate reason to use floats for something that is collected as an integer.

Ah I didn't finish that statement good enough …

The Fraction Magic program it self shown in the video, is to my current understanding, a program (GUI) that is built along of everything that the leaked code files contained, (zip folders), the leaked code didn't contain the executable GUI program.

So Bev Harris, didn't understand what this was, and hired a programmer to make sense of it, and the programmer simply recreated the GUI application for the features and abilities specified in the code.

The Big Smoking Gun: The program wouldn't work at all if thousands of voting machines wouldn't support float precision numbers, or fractions. So all of the voting machines are compromised by intent for this purpose. This whole thing wouldn't be possible with integers only.

int on its own is tricky because it can be 16 or 32 bits depending on the compiler, which is obviously something to avoid. I'm not saying the answer is to use float but I can see how someone might have thought so.

They legitimately need to rig the elections to get the results they want.

You don't need the data to be stored as floats to be able to do float arithmetic. Then again, they probably did that because they're too retarded to be able to 1. take control of code execution 2. do floating point arithmetic 3. save data and restore execution

Well, it could be the case, it could be something else, but Bev Harris hired programmer made this GUI from a code that's from 2003, so this was out there, at least with Bev, for a long time.

Ofcourse she may have something else from what that poster meant, ofcourse they may have other code, other stuff.

Or….

The likes designed them to be flawed in their favor

I found the video a bit unclear as well. You explain it well tho.
Any sane programmer would use ints everywhere for this problem, except when calculating the final outcome in percentages.

Platform and compiler are known, so using ints would be no problem here.

Any sane programmer would store a ledger of all votes for later auditing if needed.

Its so they can shift votes from one candidate to the other without messing with the number of people who actually show up to vote.

Its about percentages.

lol, no they can't have a record goiym,

I wonder (((who))) hired the coders

As a programmer, I am skeptical. It would make sense to use float instead of int as higher numbers can be stored before overflowing. Knowing the architecture would be relevant (not sure what it is or how old these machines are).

I skimmed over the blackbloxvoting website and it sounded borderline retarded tbh. Most of it was nothing… just them demonstrating how they could skew the votes, no proof that that happens.

I fully agree that electronic voting should not be used (at least not without a lot better transparency), but I this line of attack is not a strong one.

They could cast to float, make the calculation, and cast back to int.

we all know they are cheating as the exit polls have not lined up with results for several elections now.

The possibility to audit each vote would be neat. If each vote on the paper trail has an unique id it would be possible.
That way you can have three stages, pure digital quick results, a number of random samples to check for irregularities, and a full manual count if needed.

It makes no sense to use a float. As other anons said, there's only so many numbers a float can represent, and it's much smaller than an unsigned integer. It is a very strong line of attack, CPU floating point units don't usually do floating point math properly, i.e. they can give wrong answers.

This is really 100 level stuff, if you call yourself a programmer, I highly doubt you've been educated to be one. Choosing the right data structure in this situation is absolutely basic, and the best answer is an unsigned integer.

Even if, by some chance, not done out of malice, it's still a massive sign that the developers are incompetent. The fact they are using floats only proves that either they are idiots, cheating, or cheating idiots. Any programmer who isn't some self taught pajeet flunky knows you only use floats when you absolutely have to.

I agree, that is how it should be done. Your voting card should have your unique ID on it.

You can take that number to a website and see how your vote was recorded.

A float loses whole number precision after 16.7 mil. I can't think of what fractional operations you'd even want to do on a vote count.

You have to keep in mind that the voting process is supposed to be secret. You cannot take any information about your vote with you to the outside of the booth.

2+2 = 4
2+2 = 4.000000000001

see where this is going?
additionally how can you fuck this up? this is highschool 101 shit. it's an input tallier with some tamperproofing. a godamn 10 minute vbs script would take care of it.

breddy gud

There is a type that is used specifically for accuracy, they use it for stock trading, money and scientific programming, but you already know that the kikes probably contracted some outside pajeets to program it.

that is pretty much bullshit these days

they can look up your registration and see your a republican, so that pretty much indicates how you will vote.

too much fraud to let that shit go on

That's bullshit. It's to protect against """voter intimidation"""? It's infinitely more useful as a protection against vote rigging.

aka double?

Only at the very end, you might need to calculate say 60% for some referendum passing requirement. But that will be done by humans anyway.

IMKAMPFY IS A FUCKING SLAV DICKSUCKER
HE SUCKS DICKS
whew

Visit 420|^han anytime m8s
Keywords: Trump Clinton

this
hundred million times this
this is what you get when you outsource programming to poo worshipers

No they have a thing dedicated for real numbers, I don't know what the thing is called, but it's not a double or float. CAD programs use it as well.

Double or decimal, yes.

Or, if you want to go really precise, decimal128 which is used for taxes and finances.

Yeah so the headline is now also off. Sorry guys.

I'm not blaming Bev and the team, they got this thing out, but the 11 minute clip was just a bad idea for me to post this, whoever put that out in a hurry at IW does more bad than good.

ALL* THE VOTE MACHINES USE FLOATS

ALL* THE VOTE MACHINES CAN BE HACKED BY A PROGRAM THAT HAS THE ABILITY TO

Otherwise this type of fraction magic rigging wouldn't work if only integers were supported by the voting.

The reason for this type of rigging is because it's very stealthy. The hacking software doesn't even have to be buried in code inside the voting machine, so the voting machine it self is clean and would pass the test, so this big thing that they found is a semi-remote hacking capability, and most importantly it enables them to be very accurate without having to know how many people would get out to vote, and they don't need to flip votes which is

They're getting desperate that's why pretty obvious tactics such as basically vote machine GUI flip was seen in that video that circulated twitter, that's the most crude, the most obvious way of doing it, so what Bev Harris from BlackBoxVoting.Org found was a very unseen and elaborate way of rigging not known until now.

They (globalists) have a big problem because when Trump has so many supporters now it's way way over 50%, they can't simply use this software to flip the numbers, they need to use other features which make it what the public expects, but you can't win if you're not over 50% so even if the public expects 70% for Trump and 30% for Clinton what's the point, you won't win. So that's why all the corporate corrupt media is constantly reporting that the race is close, all the corrupt polls, because they can't flip and make hillary win by 80%, nobody will believe that, the MSMedia plays a key role in setting the psychological background for the public to accept shouch a turnout, but they aim to make her win by a small percentage of votes because they can't do more than that, so that's minimal thing they're targeting because of a such a big lead for Trump. They have a hard time battling alternative media, so they're getting very frustrated, and this is where their minions are going to resort to more obvious tactics which gives us more opportunities to expose them, when they get hammered they make mistakes, so keep your eyes peeled.

They can brag however they want, like they're so powerful, they don't know the spirit of liberty can turn the tide very quickly when the gloves are off, so the minions have no idea about the big picture of themselfs either, don't be demoralized when they brag how they don't care they're exposed, it's a psychological scare tactic too. They have no idea patriots inside military keep a good eye on them, and when they go too far they may get swift justice (hopefully non-violent)

It can also be used to buy votes, reward upon showing you votes for a candidate or proposition.
With a proper paper trail you can still check and track individual votes trough the system, and avoid this problem.
It should be an improvement on all accounts, and not introduce other problems.

I'm very bad at finishing sentences…

The second caps line should go "input the resulting percentage, they just put the numbers in and no matter how many voters on either side the person will still get the XX percent of the vote in the final report"

The first big paragraph should end "don't need to flip votes on the go at the voting booth, which is more detectable and has less accurate control"

So… the hacking software, automatically calculates the fraction of each vote, to provide a desirable outcome. So that's why they want fractions.

Bev Harris was around Infowars at the time of 2012 too. I remember her speaking about voter fraud. She is an expert at this. I haven't watched past 30 seconds yet but thought to relay that info in this thread. She is a regular during election season basically.

The only reason they would do this is because they are doing multiplication/division to the variable
I haven't watched the video yet.
Because they are going to be presenting the number as an integer, there is only one reason they would use a float: They are doing multiplication/division and don't want it to truncate. They are probably using a float to make sure that the end votes equals the number of voters after they do some shady math using the variable; if the votes don't match the number of voters people might start asking questions.

The best not-rigging explanation I could think of would be that some idiot wanted to get the percent of total, but the integer was getting truncated 3/2 = 1 instead of 1.5 and thus breaking the formula if Trump got 656 of 1000 votes and if both are integers, Trump's percent would come out to 0.656 which when truncated = 0.0 This problem should be bypassed by simply calling a float that is equivalent to one of the integers. However, it seems just as if not more likely that it was built to weight the votes. By putting a multiplier above 1 on Hillary and one below 1 on Trump. This is certainly nothing good. If it is a mistake, then it shows the incompetency of the people making closed source software we have to trust. If it is malicious, then we currently don't have a method to determine what it is doing exactly. Electronic voting is a terrible idea to begin with, but once you start running proprietary software to do it the issue grows exponentially.

That'd be 2^24 because the sign bit is only 1 bit.

2^24 = 16777216

...

It's all in the video it's just not presented that great, no offence.

If Hillary will get 1000 votes and Trump 4000 that's 80% for Trump and 20% for Hillary, if we just display these whole nubmers in percents.

With the hack software to get to let's say 60% for Clinton and that is 40% for Trump, it makes each Clinton vote count as 3.0 and then each Trump vote count as 0.5 – So Clinton gets the final result as 3000 and trump as 2000 votes.

With dobule float they can do this more precise with way bigger numbers.

THIS LEAVES THE TOTAL VOTER COUNT INTACT.

Okay I just randomly choosed numbers for simplicity, didn't get a fraction here but you get the point.

—————————

I was on hiatus (various life problems) in 2012 and early 2013 so I have no idea what was going on then, yeah I missed her, but I'm watching it all since April 2013 and I don't remember seeing her until now so I'm kinda totally new to this fraction thing, cause it seems like it was out there but this is a new revelation on top.

its fucking nothing; if they used int you'd be whining about integer overflow

Old Holla Forumsack here who's dealt with older shit and thus may have relevant info. Not that the use of FLOAT vs INT isn't lazy (FFS people it's a vote. make a custom number type to fit your data if you have to. It doesn't have to be fast, just accurate.) But it is worth mentioning that a lot of this crap is horribly archaic.

And in archaic systems (leaving out weird octal shit here,) INT is only 16 bits. Not 32.

Also, even shitpedia says a 32 bit floating point is only accurate to 9 significant digits.

Now, the reason your math doesn't come out like shit every time you add a number is that there's a LOT of layers to how this crap works. typically a floating point operation is performed in a CPU with MUCH more precision (I think in intel it's like 80 bits.) But if you take the value, save it, load it, add, repeat over and over, you can eventually run into some precision loss. I've run into this while doing precision calculations; changing the 'types' of data I used even from a 64 bit floating point (i.e. way more accurate than a 32 bit which is the most likely used one here) to a Precision numeric type (Meant for accurate calculations of large numbers or decimals, but much slower) showed an improvement in the data transformation.

I'm too lazy to figure out where the 'holes' in 32 bit floats are. But as an example, here are where you lose precision in a 16 bit float:

Integers between 0 and 2048 can be exactly represented
Integers between 2049 and 4096 round to a multiple of 2 (even number)
Integers between 4097 and 8192 round to a multiple of 4
Integers between 8193 and 16384 round to a multiple of 8
Integers between 16385 and 32768 round to a multiple of 16
Integers between 32769 and 65504 round to a multiple of 32

I'm guessing 32 bit floats have similar holes, I'm just not sure where.

Some computer heads here are arguing mainly about the computers not being accurate with simply having floats, as I said the OP is very crude and shound't be taken literally there's more to this than just having floats lying around there, they need floats to pull this off, and ofcourse purely floats wouldn't do this, it's an intentional script.

While I do agree there is risk by having floats at all is totally weird and isn't fullproof from a glitch but maybe couldn't be that bad I mean, if float calculations were that prone you would have your games and applications crashing all the time.

I'm trying to find that supposable leaked email where vote machine programmers said "we need floats for weighed races"

Games and applications crash all the time, have you never used GFWL?

can you find the cyanide pill that killary supplies all of her shills and let us know how it tastes? Should be in your welcome packet.

One of the things that makes electronic voting less secure is that bureaucrats can't help but constantly change the requirements of a system. Give me a full list of what access controls you want, what reporting you want, verification requirements, recount support features, etc., and I'll come up with a system (or tell you why your requirements are incompatible). But if you subsequently add something (like a new reporting requirement), then I might want to do a major rework of the system. These voting machine companies are obviously dealing with whimsical bureaucrats by providing almost no security. To provide flexibility, they could push in all sorts of directions, but I'm sure they're pushing for limited security.

Come on, GFWL isn't bad because CPU float calculations is off, what a crappy argument.

If the floats them selfs would be so inaccurate by the end of the voting then your game would crash in the first 10 minutes in all the time, everywhere, everything, I mean, you probably couldn't even use the Operating System for more than 30 minutes, a game does so much stuff it's way more than a whole election would do just summing up numbers.

- im not english
- the video provides only a claim

Sounds like an average Windows session to me. Really though I was just making a joke, lad.

How embarrassing.

These are the most retarded things I've read all year.

As long as you have write access you can put any fucking number you want there, regardless of variable type.

That's the point, if you weren't intending on rigging it then it'd be useless to cast it as float ever in the process.

No.

I can't discern some of these jokes since I already said I'm not a real programmer, I'm just on the surface of these things, I'm not a real guy who knows CPU hardware, maybe errors do happen but there's some kind of error correcting going on, that's how I feel with my existing awareness.

Yes I'm mainly on windows but heavily tweaked, I know this doesn't count, so I got myself a secondary PC with Linux Mint.

Relevant Captcha, nice! Security is almost always an afterthought in the IT world, sounds like you know it as well as I do.

Then I'll tell you why you're wrong. - Most of your apps don't use floating point for what they actually do. Per above, if the app is doing anything with money and not written by a complete retard, they're using a data type meant for handling decimal without shitting the bed. Typically whole numbers are used wherever possible, because unless you need decimal places integers are going to be faster to crunch through.

You will see lots of decimals around UI however, due to how elements are placed. BUT those never undergo the level of math involved to exhibit any of these problems.

See my post above. You can easily do a couple of calcs on a floating point without shitting the bed. If you get to the CPU level, you'll have 80 bits of FP precision as long as that value stays in the same register of said CPU. Save it to memory (i.e. goes out of CPU Register) and you drop down to 32 bits.

Have you ever intentionally introduced a bug in code (or dealt with an asshole who decided to do so for lulz?)

No. There is no point. You have no point.

It would still be completely useless. Using float doesn't make rigging any easier or harder. It's completely irrelevant. All it is is inefficient for storing this kind of data, nothing more, and even that is to a completely irrelevant degree.

It's basically nothing. Having everything stored as floats is extremely sketchy because it alludes to weighting votes. Other than that it doesn't really mean anything without emails.

Yes I'm not sure either, the way it was presented …

I'm not sure if it gets copied or not, i'm not sure if it happens live or not, it does seem to get copied but at first I thought it just the program runs and intercepts the votes and changes them, and then sends to the master computer.

I'm saying the WAY that the video is showing the way of this type of rigging, the thing you have with integers is another thing, so I'm just trying to make a case for this type to see if it holds, now if you just found a big , then blame the ones who made the hack program for being so stupid.

Maybe the big idea is in keeping the total vote count intact, yes you're right from purely technical point of view, but that can be detected because if there's only 5000 thousand total voters in the county but the numbers would show 15000 just for one candidate.

So they could do the same and are, but I guess float give them more precision when trying to make an outcome that the acclimated public excpect …. but this time around that acclimation isn't going so well, thankfully.

I'm not saying I'm right, I'm just trying to understand it if it's like this and what could it be, if it's proven that floats have some other purpose then so be it.

Using an inefficient data type is not a bug. Rigging the results can just as easily be done with integers.

Understandable, I was thinking about more in the GPU calculations that games do, not so much other apps yes, but even just calculating some distance precisely probably won't crash the game outright, but if you calculate some memory address wrong … but those numbers aren't complex or floats. So I guess the OS idea was off.

I more meant that if the float errors were as common to affect the elections which proportionally don't have as many calculations by simply adding numbers a few million times compared to what millions of people's CPUs do when they play games each day.

The proportion of that is so huge that games would just be crashing all the time then, let's just say for the sake of the argument IF they use floats for critical stuff, but okay maybe games is the wrong thing, just use some other thing that uses floats for critical work.

Yes. That's why you would never do that when rigging the votes.

Nope. Since there are no half votes you have all the precision you need with integers and the system storing everything as integers can just as easily be rigged to the same precision as one with floats because it's the same arithmetic. 327,394 x 0.5324 will give the same rounded result regardless of whether the vote count was originally stored in or will be written to an integer or floating point variable.

But it's not about me defending anything, I'm just not giving up until this is cleared out, maybe we missed a detail in the video or the presentation just wasn't good enough, I see there are articles going back to May of 2016, so this new reveal on the Alex show is something in the middle or a summation of all the research, and that 11 minute clip is such , they put a 25 minute clip out now, but I still didn't had time to listen, I'm going to now, I was writting for the last hour and a half I forgot.

Infowars is very notoriously unsavvy when it comes to tech (I say this with 4 years of experience), I can't understand anything from the article, not because I'm not techy, but because it's explained for "laymans".

"It uses floating point numbers to round up and sway the election"

It really means nothing, unless it's explained in deep technical detail along the way, but it may mean big to average people, which is good, I don't see this as deceptive, we techy people just don't speak the "layman" language and I'm a techy and I hate the dumbed down explanations people love to do everywhere, Alex himself is totally computer unsavvy, that's one downside of IW for years, whenever he talks about tech I cringe and hopefully they hire some patriot geek already there.

Whats the tl;dr? Can this be used to rig the elections? Yes or no.

It could make it easier to.

I guess you either did:
- corrected what blackboxvoting.org has thought floats are for
- owned the original programmers
- you're right but they also know this and for some reason had another reason to go with floats

For the third option I speculate:
Presumably to not calculate when they will hit a whole number result, maybe the tests have shown it takes a long time finding a whole number or the closest whole number is too far away from their target. Time is something elections don't have and I guess accuracy too. THIS IS ALL MY IDEA not trying to force my theory, just thinking ahead, this is healthy discussion I'm not arguing.

With float couldn't you also start with a negative number for one candidate instead of 0?

I'm right there with ya user. Faggots on Holla Forums just dismissed the idea.

Anything tech is 100% rigged and they know it.

Fuck off CTR!

Someone should seriously make some script to test this idea if it takes effort to find a whole number and how close it is, to see of a standard office PC can do it fast enough, less than 30 minutes to be conservative even if the video they claim it's done so that mere mintes and second count, if it takes like 3 hours we can say it's probably too long imo (okay I have a gaming rig but we can approximate with human eye how much longer would an office PC take just for sake of argument)

Unless you have already something to prove it without this test ? Well, I can only take your word I can't know for sure.

When you make a conversion from float to int, because they will do that when they get the vote count, any fractional part is dropped and the number always rounds down unless they have some custom thing for converting between types. It will probably distort the voting percentages as well. This is intentional goy.

We're kinda discussing what really is the big idea with floats that Bev Harris claims is the big deal vs integers. They're already rigging by other means such as vote flipping in real-time which also is so badly programmed it makes it obvious on the LCD screen. A video of that was out for days now, it has to do with congress votes, i forgot which candidate.

They could even break it down so that minorities have greater weight when they vote. If it's a machine, it could in theory use hidden cameras to get their ethnicity. Think about it.

Wouldn't even need to do that, you just change the weightings in areas where ethnic turnout is expected to be higher.

No. Integers can be negative too. Integer is just a full number without a fraction. Unsigned integers are full numbers from 0 upwards.

This user is correct. We teach in first year courses, for example, that you don't do money calculations with float, you use a custom or library type written specifically for the purpose. (In C#, for example, it's the decimal type.)

As for "programs would crash if floats were inaccurate", well, that misses the point. Floating point numbers are by definition not perfectly accurate, they have only a certain degree of accuracy. If you ignore this and pretend every digit of a floating point calculation is real information, you don't get crashes, necessarily, but you do get small amounts of error and these can add up.
it's sketchy and denotes either incompetence or malice on the part of those writing the software. And the easiest way to hide malice is to disguise it as incompetence.

Shouldn't this be sticky?

Not any more than it already could with integers. In the video they are essentially saying that you couldn't rig the system to give Clinton 53% of the vote if the vote counts were stored as integers which is fucking retarded.

Can you expand of what you are thinking? I can't understand what are trying to say.

let me give you an example that you can see fairly easily.
It's fairly common in video games (I use that example because it's the most common use of 3d graphics) for one item to be closer to the camera than another, but because of floating point calculations introducing small errors, the back item is actually in front in places. It's a "z ordering" or "depth ordering" error.
This kind of thing happened all the time in Playstation game, for example.

This is true. It might make it easier to disguise the rigging, since some level of inaccuracy is very easy to introduce using floating point math.

info on jeffrey dean, the alleged criminal mastermind behind this

blackboxvoting.org/bbv_chapter-14.pdf

That the machines are rigged and we need to ask for paper ballots at the polls?

Have you not been fucking paying attention?

floats end up with 5 if there is a fractional part.

This guy actually knows what he's talking about unlike 99% of this thread.

Reminder to keep your dumbass mouth shut if you don't.

Not for whole numbers it isn't, unless you do something plain retarded as using 16 bit float for something like New York. Even then the errors are proportional and are not going to give you even an error of a hundredth of a percent.

You need to rig it proper to get any worthwhile results and that can be done just as efficiently with integers.

I'm not a mod, not sure if a request can be made by OP.

————–

The big headline should be "the missing link" i just couldn't get it across at the time properly

The missing link is actually a master key that was discovered … but again it's a terminology trap I think

But the lady again uses very bad terminology for laymans, "some kind of a master key as I put it" she may be talking about this hidden program and not an actual private key in encryption so this adds more confusion as the alex jones show is mainly broadcasting to technically unsavvy people but it's actually Alex himself that's very stupid with computers, by dumbing things down it actually makes it even harder now all the tech gurus get confused as well.

So this floating point stuff was just another thing, but the initial breaking news as so confusing.

THIS DOESN'T MEAN SHE IS DEBUNKED - I'm simply pointing out that yes she may not be totally correct with the floating point idea unless I'm not getting the full story, so I haven't had the time to research all the material on the site it's a lot of it, so no conclusions on this, but TLDR yes rigging is going on none the less anyway.

If I understood you correctly, after the hacking program does the weighing and returns the float you said an additional procedure could find a whole number by skewing some more until it hits on a whole number, but I said that the additional time required to calculate the closest whole number may be too much or that the closest whole number may be too far away.

But I'm kinda taking that back now after I did more thinking, not sure why would it need counting and finding whole number, just forcefully hide the decimals, like they already do when they report them in official final reports as stated in the video, I kinda lost myself I'm not even sure what it is that I feel why I'm wrong, let me finish this post, I need to clean up tabs, I'll read your replys and get back.

CTR again showing that retards working in anything related to current politics are too dumb to know anything other than middle school level maths. It also shows that common core is pure shit that is made to rot brains.

I hope you do realize the data format would be completely useless if simple addition of single digit numbers gave you a 25% error.

If voting has to be done electronically (it doesn't), it should be done on a custom circuit with only:
- a mechanism to randomly assign a 32-bit register to each candidate
- a hardware queue to receive the votes
- a mechanism to increment a 32-bit unsigned integer once per queued vote
- the ability to uniformly reset all values at once
- a way to dump all register values on a screen

Anything more flexible is fraud.

Does a sketch of an algorithm using floats is shown in the video or in a document ?

Are you talking about rounding the float 345.90124 to just 346? Because that takes a few cycles. It's basically instant.

possible if your numbers are near 0

Float stores 8 sig figs, while double has 16. On the other hand, a 10 off the total votes wouldn't matter as no such EC district exists.

Yes I'm talking only (mostly) what the video shows unless I specifically said it was my own speculation.

How fortunate then that there are no single digit numbers close to 0.

Its highly suspicious if any historical voting data shows fractional votes, so I'm not trusting that this self-taught nigger really figured out the master key to voting fraud. Didn't anybody ever notice before that the votes in the tallies were rounded? And if a tool can be developed to plug in global percentages that too would look suspicious since all the precincts would have the same percentage, which is highly unlikely. I don't doubt they goof with the numbers at some point and nobody would know, and perhaps fractional counts is a red flag, but this 'magic' thing is incomplete.

W E W

For some reason I thought it would have to keep adjusting the vote count weight multiplier for example 1 x 0.44 (let's say it uses 2 decimal float) until the 345.940124 hits on the closest whole number, now, with the huge amounts of numbers, maybe the skewing of multipliers may not be able to make it to 346 with single or double float precision they have in the original numbers maybe they would need to bounce around but this is kinda a way to get the whole number out with purely using the multiplier now why would this help I totally don't know anymore , i don't really see what the heck why I think this is needed now that I had been thinking.

But I forgot the original reason why I went thinking into this way, just ignoring the decimal is instant yes, i thought maybe the reason was they would figure it was rounded like this, but with millions probably unlikely, So nevermind.

Nothing good ever comes from using floats.

Want to fuck up your code? Use floats.
Want to rig an election? Use floats.

uint64_t or go home

wrong, and wrong

fucking Holla Forums understands less than shit about computers. to start, the ps1 used fixed point math for vertice coordinates, not floating point. this is why everything existed on a tiny grid and appeared to 'snap' from place to place.

goddamn retard street shitting indians know better than all of you combined

Agreed. Honestly, it's probably Internet Exploder-based javashit "programming". As such, float is the only number type in the language.

All of the bugs have been in division. No one, to my knowledge, has ever fucked up addition.

Generally, what I call a "float" is a "double", and has 53 bits of significand. If you have a 32bit machine (POS [Point of sale! Giggle…]), then the double gives you a larger integral range.

The video I forgot to embedd

youtube.com/watch?v=GgbdMc3QUHg

The programmer that Bev has in the video has made this program to what he would want and wish if he was trying to make rigging, presumably also along the lines of old leaked code to give him general direction and features, to my knowledge.

Then he discovered that the vote machines have backdoors and functionality built in to support such features so there's no actual hacking involved as it may sound from headlines, there's no real hacking to break into the voting machine, the voting machine already machine allows the remote user to do this things via the program he made with educated speculation as it seems.

There is no master key, it's another stupid layman term, the program is the interface and the voting machine already has support for such an interface, it doesn't need any keys, it just works.

The term "precision" as used by Bev is also meant as precsion in voting location, "down to precint level in each county, early vote" so don't mix this with "float precision".

ta.twi.tudelft.nl/users/vuik/wi211/disasters.html

Funny one!

—

I forgot to embedd again

Their explanation is outright not plausible from a technical point of view. Not missing things to be plausible… but in itself claims and assumes things that don't make sense.

Everyone is so focused on the tech details that it seems like nobody noticed the biggest problem with their explanation.

They assume the attacker has unfettered access to the database of ALL votes

At that point nothing else matters. Game over.

If you can get to the record of votes, unencrypted and with write access, you can do whatever the hell you want to them.

That doesn't mean the floats aren't suspicious

Particularly the fact that they switched from ints to floats. If it was floats from the start, you could let it slide as a default they never bothered to change. But they deliberately changed to floats, allowing for storing fractions of votes.

I think that's the crucial bit. You can do whatever math you want, whether you start with ints or floats doesn't matter. But why would you want to STORE fractions of a vote?

I THINK THEY ARE ON TO SOMETHING, BUT DREW THE WRONG CONCLUSIONS

The only reason I can think of is to do something like images related.

Easier to show with code than to describe. Should be fairly readable for non-programmers, I hope. Lines in red are comments (obviously?). Output displayed in second image.

Only code labeled malicious is malicious. Everything else would be completely legit. In fact, if you set skew to 0.0, you'll get completely legit results. This could be used for passing acceptance tests. Malicious code could stay undetected unless code is audited by honest people. Alternatively, there might be a way to add that bias on the memory sticks. Previous efforts have shown that memory sticks can do more than just hold data.

Now you just need a way to provide a value for skew on election day. Might be hidden on the memory sticks they use?

What the value for skew should be, you can decide based on pre-election polls. As long as you don't make it too big, people will accept results as plausible.

Indeed interesting I think I came across this patriot missile thing some years ago, but didn't know of all the rest the cases, gonna have a good read.

This fucking year again.

The Florida count was a little too close wasn't it.

Updating my oc with the new content from the video.

Who is this guy she is talking about that simulated the needed elements and found them in the system?
This is her book blackboxvoting.org/BBVbook-all.zip
web.archive.org/save/http://blackboxvoting.org/black-box-voting-book/

The right data structure is neither int nor float but an interleaved binary table with metadata like current time, current polling station staff, admin, windspeed, cpu temp, and a cumulative hash of the database.

Hashing the entire table per vote means that you need to recompile the entire voting record to modify any single record: an exponential big O tire fire that hotwheels would be proud of.

Anything else is ripe for modification.

floor(trump)
ceiling(clinton)

Fuck this thread is stupid.

Let me tell you a fact about IEEE floating point:

Adding two integers always results in an EXACT integer, because the mantissa is a fraction in base 2 and the exponent is a power of 2.

In fact, doubles will only become inaccurate once you add integers on the order of 9 quadrillion, and even then, it will STILL result in integers, not fractions.

this is completely false

Give me two integers represented as doubles which, when added, will not result in an integer.

I think your right, I'm listening to the interview now, the way it sounds is that they're speculating but they seem so confident.

The whole "counting votes as money" is a speculation, it's not really confirmed.

I still don't think they want to be purposely deceptive, I get it that they wanna figure out and they do point out "they think" so you can't say Bev is doing shady but surely it's just a theory on their part.

Exactly, float thing may be for some other purpose and she said that one company even admitted this publicly albeit very subtly as some kind of a feature for something, but this is what I tried to say earlier, to not write this off completely, maybe Bev is mixing things up a bit, floats may not be intended for what she thinks now, but still for something else also about rigging, even if right now it doesn't make sense, it may just be a different flavor later and that could be spot on.

If you don't understand what's wrong with this:

then you don't know enough to be telling people that this thread is stupid.

In which language, on which architecture, using which standard?

Saving votes as a money (currency) type.

double or single?

I never said voter fraud couldn't be conducted that way. I'm commenting on how fucking stupid it is that people think floating point is indicative of fraud rather than developer incompetence.

I already said, but I'll spell it out for you: IEEE 754, specifically binary64. AKA basically the only floating point standard people use today. But the integer subset of floating point values being closed under addition (even if the result itself isn't valid) is true for any floating point standard, barring adding to Infinity.

x86 IEEE single floats

How about any?

This has nothing to do with the numbers being stored as floating point and everything to do with the code that rigs them. You can just as easily achieve the same thing with integers.

...

It's not a float, it's a VB Currency-type scaled integer.

...

so can this be fixed if trump pushes it?

So they're using fixed point? Then what's the point of this entire thread?

It still allows for 4.2 votes for a candidate?

Guys, I'm pulling back some of my stuff, not as if I stated it as fact, I'm pulling my endorsment back, the lady doesn't know any of the terminology, if she's not a tech person to being with calling the GUI and "overlay" and callign the made up speculative hacking program a "master key" it's also chances that she's concluding something inaccurately and

AGAIN, not saying that she is totally wrong or debunked, it's just that it may be a bit different as to what she is claiming, so this was ALL just speculation, nobody knows really what they do with these floats, the hacking program is made up, they don't use this program they use something else, it could be similar, it could be quite a bit different but the TLDR probably still remains "more evidence in the direction of mass fraud"

So this BIG reveal is ONLY what BlackBoxVoting.org THINKS MAY be happening. So this is the OP i should have posted if I could figure it all out that fast.

so a double precision integer I guess.
max value approx 922,337,203,685,477

Sounds like they made the right choice. What's your source though?

But fixed point never has any inaccuracies. Any decimal value of votes for a candidate would be immediate egregious evidence of voter fraud, so they would group fraudulent votes so they get counted as integers, correct?

So what's the fuss about using decimals in the first place, if they have perfect precision and aren't actually being used as integers?

aren't actually being used as decimals*

She says it's "what i think" but then says "ah I know it i'm sure it's not a theory" …

But don't count her out because there's more that makes her go here into this direction:

In the interview She claims:
- She knows the exact date when INT was changed to FLOAT.
- She knows that the change to FLOAT propagated to all other vote machine vendors in the US.

Now these are the details kinda quite important because you can claim all you want unless you have more validating material to support it, so I learned not to celebrate too early, I guess I celebrated too early.

These details aren't in the video, I only hear it now on the interview on the AJ Show.

Now you're changing what you're saying.
I suppose your girl Hillary also had no "intent" when asking for that personal email server too.

None. The point of this entire thread is to show how gullible Holla Forumsiticans are. As others I'm not saying these elections are not skewed (in fact I believe that if they can, and they can, they will), just that an unsubstanciated claim is just that, a claim.

No, that's exactly what I said from the start. I actually do think Hillary should go to prison, and that you're being retarded. Tell me where exactly I backpedalled - my only contributions to this thread have been to explain what floating point actually is and why it has no relevance to whether voter fraud is or isn't happening.

It's disappointing, watching everyone think they suddenly understand the floating point standard. Yeah, I don't trust the elections right now either, and though I haven't seen convincing (to me) evidence of large-scale automated voter fraud, I wouldn't be surprised if some came about. But until I see some evidence, I have no way to justify believing voter fraud is happening.

What the seer sees, the prover proves. Holla Forums has been having threads on electronic voting machines for years now, it's just that normies are starting to think that it isn't such a good idea either. Throw some lady some spooky and occult terms about how basic computers work and she'll say with certainty that this is how the elections are going to be rigged, since she's already convinced the elections are going to be rigged. Holla Forums is convinced that the elections are going to be rigged, so they will bite on anything.

The most vulnerable part of voting machines isn't what this thread is on but someone in the middle changing results on the fly. Unless the process is entirely clear, seeing the journey a single vote tally goes into being counted like a paper ballot can be physically tracked, then you'll never know just what happened from touching a screen to cast a vote to the final vote count.

The video was too long, didn't watch. However, of the part I watched I did hear that votes were counted like money. I also assume a stupid language, like Visual Basic or Internet Explorer based Javascript (you wouldn't believe how many cash registers are Javascript in Internet Explorer). Visual Basic has a type dedicated to currency, so it makes sense to me.

Double precision IEEE-754 is just as good as ints in handling integer data, and all of the bad things you can do with floating point math you can do with fixed point.

I thought I understood it, I don't see why floating point numbers are needed really with full access to manipulate anything at will, the whole ability to do this with the imaginary madeup program is a speculation so this kills it here, they may not have full access, who knows, we may never know so this part can't be taken as 100% and people shouldn't keep arguing around it.

She's mixing things with "one central computer" and I have no idea anymore what this program interfaces with, precision down to precint level but then the one computer is used to count, where is the skew made, at the precint result before the main computer reads it or after, so there are technical holes in this explanation.

Repeating again, that doesn't mean the whole thing is debunked, the lady just isn't concluding correctly, even tho she has all other things right, she has people on video carrying out PCs, opening up servers they shouldn't, angry officials who don't like cameras. etc, it's more in the miscalaneous there's more smoking guns, the smoking gun is more in the big bunch of smaller things put together.

I apologize for making this about floating points I thought this was "the missing link" as reported on Drudge.

This actually feels deceptive because nothing is confirmed, so how can you call it "missing link discovered", it's a theory really, now that I REALLY got my head around it, if there is full HW access to all the numbers, why are floats needed, this is the main question.

I don't hate nor fully endorse alex jones but I just learned to listen more and not be picky or ommit sources and try to check and analyze everything out there, with my 4 YEARS of experience listening to AJShow I can tell you that he is aboslutely TOTALLY uneducated about computers, he uses apple products because Android is probably too complex for him, it's been ages since he been close to a real computer doing DVD movies, so he's enthusiastically cheering and I sware he doesn't understand a thing what she's saying he's just hyping this up because he trusts her 100% ofcourse he can't check himself he has no idea about tech, he only understand it as others tell him how it COULD be. If he has a good tech guru properly explaining to him then he understands ofcourse, like william binney etc

Oh by the way, Kim Dotcom will be on infowars this week.

You really need to fuck off and go look at the past documentary before trying to discredit anything.

Its kind of hard for me to connect the dots with the black programmer's little project. It has a lot of things I don't understand - read/write manipulation of votes, is that possible with the actual voter machines (instead of a fake local offline project like he did) when connected to the main local server for the county/region? The long video 20+ minutes long gives an example in Florida where it looked reeeeally sketchy, you should watch that. The car did infact pull up to hide the license number, it was weird that they took the PC out of the warehouse into the car, there's a lot of questions. There was another short video in there showing a man opening the election server when it wasn't supposed to, so that is another red flag. Someone with insider knowledge of this industry & is involved in the software or someone who can correctly analyze this whole situation is needed. Too many open questions for me right now.

Well man, we're watching it all and we don't understand and don't fully buy some parts and you expect to

There is many details in the video as well as in the interview on infowars, and it's like totally something else what I thought it was when i started the thread, there is a bunch of things going on, video footage, surveillance footage, float is just one small part of this report today.

Suck it up you fucking nigger. You can watch the vertitas videos which are equally as long but can't watch the 25min Fraction Magic detail version?

This Comment on the 24 minute video is a decent summary about what the programmer did:
contains copy of GEMS software and actual vote databases used in real elections. Yes. GEMS uses MS Access. The attack was implemented in a real system using the real program.

Depending on the language, a float is all you have. However, as many anons have noted, it is a questionable choice. Good practice would be to use an unsigned integral type, as the anons have pointed out. My point would be (as a programmer apologizing for other programmers [even if they are designated programmers]) to focus on the other shady shit they've dug up.

Salty much ?

I just noticed it was Bev Harris that made this comment an hour ago, wew!
Also in the comment: http:// blackboxvoting.org/voting-system-technical-information/

/* compile secretly the program with a malicious c or assembly stub modifying the IEEE rounding mode on the fly, based on a number of ticks. Does not need to have access to any variables */
n = number of votes;
/* cound add up some artificial votes to maximize the error generated by 1/n */
for (i=1;i

one guy in comments on the video says this

Yes, if votes were weighted according to whole integer factors, the weighting would be obvious in the up direction, as the numbers would exactly 2 times, or more, greater. Without fractions, negative weighting doesn't work either, because you couldn't provide a factor of less than one, unless you used negative integers, which would also be obvious for the reason above. So fractions are absolutely necessary, to enable both upward and downward weighting, to obfuscate the manipulation, and to give fine control over the vote totals and percentages.

I get it now.

Then why store integers, risking for your method to be unveiled, instead of casting them on the fly ? Why use this method for voter fraud if you can just rewrite the votes directly in the database (which would be immune to a later verification by a third party, unlike this method) ?

Why is it legal to use anything other than 1 vote equals 1?

This.

That's the reason for using float. It allows to slightly skew results in a way that's not noticeable.
Most elections, especially in a bi-party system, are won or lost by a very thin difference in NUMBER OF VOTES.

Skewing each votes to be counted 1.13 for Clinton and 99.87 for Trump (or anything else really… it's to whoever makes the check I suspect) allow for complete control over the results without raising any questions.

60M VS 65M votes (last presidential election, less than 4% margin) means a centile skew (0.0X X being the skew) is enough to completely invert the results.

Splitting the 25min into mp4s for lazy ADD assholes.

...

It's not legal, thats why they hide the decimals when the votes are counted.

...

Democracy has been a joke from the start in this modern iteration. The old Greeks who made the system this shit is named after (but bears no resemblance to) would laugh their asses of. All of societies degenerate dregs join with the minority of normal citizens to be pushed around by a deceitful propaganda system to choose between a couple of pre-selected and vetted candidates with systems outside their means to inspect or know if works as intended at all. It's bullshit from start to finish and anyone falling for it should be fucking ashamed.

Yeah, because that's the problem here. I actually called my govt dept who handles this shit in my country (I'm not burger, but we use the same proprietary boxes for voting here,) and asked if I could check out the source code of these machines. Sure, it wouldn't fix all the HUGE AND GLARING flaws of "democracy," but everything helps, right? Nope, that would be a security risk. Ok, me knowing if my vote is counted right at least is a security risk.

The audacity of the autocrats of today and their obvious contempt for me is only eclipsed by the magnitude of pleb dumbness. That fat fucks like Alex Filters talks about it in the way he does just confirms his shillness to me. "Oh I have top men who inspected this shit and found muh floating points, now buy my benis pills goy."

Can I see that source code you somehow got access to please? No? Then fuck you, and fuck this gay earth.

Okay. Watched it. Programming nigger is self-taught, not bad, though. I don't need decimals.

Say I want 43.25% for Trump, 51.04% for Clinton, the rest for "Other"

Total_Votes_Cast = Trump + Clinton + Other;
Trump = Total_Votes_Cast * 4325 / 10000;
Clinton = Total_Votes_Cast * 5104 / 10000;
Other = Total_Votes_Cast - Trump - Clinton;

I never have to leave integers. Floating-point numbers are just two integers and a flag deep in the bowels of the computer. As most anons have stated, it would be easier and safer to switch a vote from A to B in the software before it hits the database.

Say that you have a district with 10000 voters. Polling indicates 48% Clinton, 48% Trump. If the polls are accurate, we would expect Clinton to get about 4800 votes and Trump to get about 4800 votes. If we flip 1 out of every 50 votes for Trump to Clinton, Clinton now gets 4896 and Trump now gets 4704.

...

(((Lex Luthor))) is behind this!

The claim was that they rather don't depend on how many people will actually vote, they rather have something that's better at adapting with whatever the total vote count is, it may seem easy here, but how many precicts are there in the WHOLE USA … ? how many thousand?

I'm just pointing that out, not trying to debunk your idea.

criminally underrated post

Funny, but that will only be applied to white people.

I get that: it's a centralized means of skewing the results that requires auditing potentially all of the polling precincts to discover.

The thing is, we don't need "fraction math" to do this. We don't need to baffle people with bullshit. It is a problem with a centralized tabulation system that automatically reports its results. People trust computers, and machines in general. People stupidly trust computers. They see no humans in the loop and assume that there is no fraud. They forget that a human programmed the computer.

There is a tug of war between private companies trying to prevent industrial espionage and the public which has a right to know that the voting machine is trustworthy (by inspection of the hardware and software).

Read whole thread. Muh talk about integres vs floats. But here is my observation:
- woman obtained some result from previous voting
- she probably saw results with floating numbers (1234,43 votes for Bush in xy county, FL)
- doesnt make sence, since should be 1234 only
- she went to the magic program nigger
- nigger figured out, this must be some manipulation
- nigger was somehow smart and did understand, those number is result of programmed skewing; he also understood simple trick (that Anons three line program insert fom above); he understood, that skewing would necessary lead to floated results
- nigger wrote program whitch skew results with nice GUI as a proove, but he is still too stupid to explain in simple English
- woman take those to Jones
- Holla Forums is arguing about integers and floats and never asks, how that thing started at the first place

Now. Real question here is, does those data from zips contain voting results with floating point. Because that would be direct proove of manipulation. Why else she became suspicious and take the data to the magic negro?
Am I missing something?

He didn't write the code you god damn retard he wrote the GUI which read the code the vendors had programmable code. It's literally in the video. This is not the first time that an executable code was brought up, it's just the first time a GUI has been made for it.

Don't usually post filterberg but this is a good interview, lady knows her stuff.

Actually I agree now it's possilbe to if you would really want to with integers only ofcourse.

The point is some kind of error and side effects happen which makes it more obvious and less precise, i guess …
I have to leave it here it's very late here and I can't keep eyes open anymore.

Yes the zips are somewhere on her organization website, maybe included in those files, there is a program package 300 MB large I just don't have enough juice today to go deeper, she claimed she has the information the vote machines first had INT then switched to FLOAT, among other things, night everyone.

Actually fallen asleep for 20 min forgot to post, but heck im not staying up for more than 5 min.

Yeah I'm not saying she's so wrong, more saying she uses non tech terminology in an attempt to make it more understandable to people but in turn makes it less understandable to tech people.

Vote machines increment a count by 1 (one) each vote. Or several counts by 1 each vote.

Floating point numbers eventually lose the ability to distinguish between numbers with a difference of 1. Meaning: n+1 still equals n. You increment the vote count and literally nothing happens.

In Forth, interactively:

300e6 1e f+ f. 300000000. ok 100e6 1e f+ f. 100000000. ok 50e6 1e f+ f. 50000001. ok

With per-district counts this doesn't matter, but it's yet another reason to use an appropriate datatype. There is no reason whatsoever to use floats, meanwhile.

Programmer here. Technically, they could still manipulate the vote with this method with integers, with fixed-point integers. Say you have a 32-bit unsigned int. Normally it can count from 0 to 2^32. But instead, if you treat the lowest 8 bits as a fraction, you can count from 0 to 2^24 with a fractional precision of 1/256. This is how programs typically did math that needed fractional precision before hardware floating point became popular. They could easily give Trump 224/256 votes per vote, and Hillary 1+16/256 votes per vote this way.

However, the fact that they are using floating point tallies very strongly suggests that they are manipulating the vote with fractional voting, and are too stupid to know how to obscure it. If Trump wins and demands an audit, it would crumble in their hands.

The hacking program was already part of the vendors system, just not in GUI form

And what happened when they did that?
Without going through all the code to rewrite stuff, that's just guaranteeing that you'll introduce a lot of bugs.

Ok. If those data from the video are original data from some previous votes, and there are results like 1234,43 votes in there, for me (retard), that is clear sign, those results were manipulated. Am I wrong? Beacause, you know, 3/5 votes or such doesnt exists anymore shame

They can do that regardless of the variable type, float or int.
This is literally nothing.

I would recommend, somebody looks at those zips and if there are voting results like 1234,55, that shit is manipulated. There is no fucking way to obtain such a result without skewing the numbers. Thats all from me, now back to difference between floats and integers…

(checked)
Yes.

The insidious thing about this is that it would be nearly undetectable. Even as you follow the votes as they are coming in. Even as you watch precinct votes being added to state totals. Because the adjustment happens in fractions of a vote, as the votes come in, rather than one big edit at the end.

T: 1.76 | H: 1.24

Two votes each, but Hillary is ahead. If you round, at this point it still looks legit. But as more votes come in, the difference gradually adds up enough influence the outcome.

You cannot do this if the votes are STORED as integers!

The machines could reload a legit firmware after the elections close and delete evidence of it.
This int/float thing is nothing in itself.
The bottom line is that with voting machines the result is ALWAYS going to be suspect.

You are missing the point. Floating point result somewhere is smoking gun of manipulation. It shows, how it was done.

Remember the soldier's maxim.
"The lowest bidder built your weapon."
Maybe Sandjeep at the Indian coding mill who the coding work was outsourced to simply made a mistake?

Yes. And only way how would you know is, final result will be in form like 50.9 H, 49.1 T. That is, what I am talking about. If that form of result exist in previous voting, that is proof of manipulation.

you retards will believe anything

What mistake? It cant be.
100 000 are voting. 60 000 for Trump, 40 000 for Hillary hope so
Result is 60 000 : 40 000, period. Never 60 000,12 : 39 999.88

You faggots realize that having code in the voting machine that would some do autistic shit like splitting votes like that would be just as obvious, if not more, than literally just adding votes on the end as an integer. The fact that the voting machines use float is not any sort of smoking gun other than that it was produced by a crack team of poointheloos.

If you have write access to the machine, you can do anything you want to it.

Not in real time. Not updated every 5 min for Blitzer with right numbers.
Also, you cant just add numbers. Votes count must be equal to the voter counts.
Also, you would need an army of manipulators online.
This shit is stealth and automated. What is needed is few more lines of code for self deleting surplus code (reset).

Programming user here, this is absolutely damning to me, just hearing the datatype and not watching anything. In order to evade even simple statistical analysis, you need to compute the total skew over a number of votes fairly near to the desired vote total. If you flip votes (which is basically what you're limited to with integer vote totaling) any modular sequence of vote flips will be detectable (via en.wikipedia.org/wiki/Benford's_law for example). Randomization of the interval between flipped votes will win you elections, but the deviation from exit polling will be too regular, which is also easy to detect. So you need controllable per-candidate margin/minimum settings (dobut this is difficult when you could have a long ecumented here?), but doing this over the open interval does not complete, so this too does not give the needed control without heavy-handed measures apt to leave a numerical signature.

Or you could just use a float and a secretive storage format, and let the inherent imprecision of IEEE floating point math fuzz all those artifacts out of your cleanly fudged numbers. These lazy people aren't even trying to look credible.

Except nobody can view that code. Except (((approved))) (((government))) (((auditors))).

Except you can check intermediate results as the election is going on. It might be a little bit obvious if a candidate suddenly loses votes at the end.

Except most machines USED to use integers. And then 98% of them quietly CHANGED to floats instead.

Except people can and do check the behavior of the machine with test runs. And you can't guarantee access to the machine between test runs and election. So you need to be able to do it having access only to the memory devices.

Reported. What did you expect?

America is dead.

I'm not sure what I expected, but I wasn't prepared for tumblr-level alarmism

I'm going to go to bed now, but I have one last thing that I want to tell you faggots:
IEEE floating point math is not imprecise when dealing with integers

With a single precision number, at 8 or 16 million, adding one no longer has an effect. The number will stay there, looking mighty suspicious.
With a double precision number, that happens at 9 quadrillion.

In addition, look for this:
12.5 rounds up to 13
11.5 rounds up to 12
Add them together and you get 25 votes from 24 voters.

The whole voter fraud thing started back when bush beat gore and people accused diebold of being in bed with the republicans and they stole the election for bush. I was interesting in all of this back then so I remember following it. If my memory serves me right, the leaked source code that Bev obtained was fucking Microsoft access - basically the votes were tabulated into an access database then someone would put it on a disc or transfer the mdb file over the network. So i did some research and that version of access had low bits for integers and really could not be used for a big election, so that was my guess why they went to floats.

Now, it seems a lot of you fags are getting bent out of shape that it's stored as a float. Let's say it's stored as a long or int - all you do iin your subversive method is covert the data types to a float or double, apply your weights, then just add them and convert back to an int and store it. You could see the results in the database as your int/long even though it was fucked with behind the scenes.

The real issue here is it's not open source when it's being used for public elections. If they would just drop their shit on github it would calm everyone the fuck down.

plus it makes sense that they used them because vote tabluators are clearly recycled quakeworld servers.

cataboligne.org/extra/qcmanual.html

I applied at ES&S years ago and declined the offer because they wrote everything in Pascal and I had no interest in a dead end job.

for the normies in your life.

No. Intentional by leftist white program managers because after all "numbers aren't definite, but flexible things".

The problem I have with believing that this will be an issue is the ship-jumping everyone is doing, the panic ensuing in the Clinton camp, and the fact that they could have just done this in the RNC primary. Why are they freaking out if it's just gonna lead to a victory anyway?

because you can only steal elections at the margins.

Democrats fiercely resented not stealing the 2000 election hard enough. It was so close.

They won't be able to steal a Trump landslide.

no.

i do get part of it. we are being told to say vote is rigged, and to do so via 'master float integer' meme. the bit about a blank faced idiot and an geriatric bobbin haired hippie finding the missing link may just have been a linguistic oversight. the 1980s style graphics and pictures of circuit boards certainly lend an air of technical to it all. then there's the bit about sending clicks to drudge and infowars. the sense of urgency and fervor sensationalized in to everything is quite… well… urgent. the ads within an ad are may get some clicks, we'll all be made to look like fools, and be doing useful work on several levels. maybe everyone will do as they're told.

but please, user. just go away when this is done. i know, i know, no 'muh safe space' shit, we have to be attacked, subverted, and granted no peace – and this vid, with all the reactions both real+paid+shilled+memed+spammed certainly achieves disturbed-disquieted. 'who's jewing who' achievement - checked. as if the trump-bane post 'ANTI-TRUMP JOURNO DOX DUMP 9000' whatever earlier did make that clear enough – but it also did dox, and maybe some utility will come of that. again, maybe. then again, maybe some asses just rolled in and posted here to generate content to sell their news wares – the speed of it all was quite… speedy, with snapshots of plenty headed for /gaschamber all ready to go. quite the monday-tuesday operations in all. much more effort this time around.

but please. that nig looks terrible, that bobbin hag looks ugly. alex needs to go lift and cut the vitalwater. drudge has been as disgusting as digg or slashdot or redice or renegade or redice (to be fair, each of them has had their moments of win, but i'm not sure i fairness in order after you SPAMED MY FACE assholes).

also, none of what ever 'master float integer' is a thing. god damn it. and double-god-damn-it that i'm pretty sure you all did a good enough job that there will be echos of this for months/years to come, whereby we'll all be subjected to the same master float spam again. i sure hope you actually know what you're doing while you're doing this grid, except fuck hope.

Hello?

This is where their whole argument gets very confusing. If you try to perform multiplication entirely in an integer space then yes you will have those problems but this whole thing assumes you either have read write access or your rigging code sits in the middle and fudges the numbers before they are permanently written.

If you have read write access to a tally you would just read the integers, convert them to decimal or binary floating point inside your black box, fudge the numbers any way you want, convert them back to integers and write them back to the tally. In this case using float or scaled integers would only harm you because it will only draw attention to the a process you are trying to keep covert by how retarded it is to use that data type for this application.

If your have black box rigging votes before the vote is registered permanently then you can do the same thing better with integers. That black box would simply register a new vote, look at what the target split is for that point in time and then forward it as 1 whole vote for either Trump of Clinton based on what will keep the split in the right place. In other words if five people in reality voted T C T T C in that order, what came out of the black box and being stored in the tally could be T C T C C.
Again using float or decimal would only harm you because anyone getting a look at a single fudged vote would immediately see you registered it as 0.94 votes for Trump. It would require you to also have control of the entire tallying mechanism and assure that no-one will ever be able to look at the raw data of any stored votes. Just keeping it spitting out integers at a preset rate would be so much more inconspicuous.

But that would leave the trace that you are fudging out fractions of votes. It's so much easier and clandestine if you just make the machine spit out more 1s for Clinton than it does for Trump.

Imagine a two-candidates election. You have chosen that candidate A would have ``p`` percents of the vote, and that candidate B would have ``1-p`` percents of the vote. Now program your voting machine so that, each time someone votes it draws a number ``s`` from a uniform distribution between 0 and 1. Your program now looks like this :

s = rand();
if (s

It really doesn't. You are counting and fudging whole numbers so integers will do just fine.

(checked)
unless kek steps in and rigs the election by manipulation over random numbers

Yes you can. When you reach 4 votes for Trump and 3 for Hillary you just store it as 3 for Trump and 4 for Hillary. Same result, no sudden jumps, no conspicuous floats that will draw attention to you.

What are you even talking about? If you use a simple bias, like Trumps votes only give him 0.88 and 0.12 to Clinton then it will round up to one extra vote to Clinton at the same time an integer system would decide to just give the next of Trump's votes to Clinton. It will just a regular.

You are applying it on top of a naturally fluctuating stream of votes with the constant increment of 1. A bias for Clinton inside the computer will not be distinguishable from a bias existing in the real world from her really being more popular. You will get the same fluctuations only they are more shifted to one side which is just what happens naturally when a candidate is more popular.

For any numbers we are realistically looking at you'll have like a ten decimal precision. How is that going to mask an irregularity in whole numbers?

And there's no guarantee that all the machines will be running this code. Not to mention the election fraud lookout this election cycle will be at levels that make expert fraudsters sweat.

No. this must be more sinister: a hardware level rigging by custom made floaing point ALU chip, so when they forced to release source code nothing incriminating will show up

Convert churches into holy houses of Kek.

No. this is more sinister: a hardware level rigging by custom made floaing point ALU chip, so when they forced to release source code nothing incriminating will show up

No. this must be more sinister: a hardware level rigging by custom made floaing point ALU chip, so when they forced to release source code nothing incriminating will show up

No. this must be more sinister: a hardware level rigging by custom made floaing point ALU chip, so when they forced to release source code nothing incriminating will show up

Hopefully this will stop it. Election rigging is probably a treason-level offense.

I wonder how President Trump will deal with these people/these companies?

My god you're dumb.

The reason they're floats is because you need to increment or decrement votes by fractions of vote FOR EACH VOTE in order to get the desired result!

You are 100% wrong.

They don't give wrong answers, some architectures have an off by one unit in the least significant position and both answers are correct. For instance is 1/3*3 equal to 1 or 0.999999999 different architectures may give different answers and both are correct. If you need an architecture to behave in a different way you can make it crank through floating point math using defined subroutines that twittle bits directly instead of using onboard logic.

Your ignorance combined with arrogance is showing.

Absolutely unequivocally false.

Still it is odd to use floats for counting votes but I can see voting systems where it would be desirable for instance if a tribal government encoded on voting cards your percentage of tribal blood and gave you a vote proportional to that ie a 100% native gets one vote but a 1/4 native only gets a 1/4 vote. But I don't know if this is actually done anywhere.

A long long (64 bit int) goes from −9223372036854775807 to + 9223372036854775807 an unsigned long long is 0 to 18446744073709551614

But not all architectures will be happy with it, still you can implement it manually if needed.

They gotta make sure there people get in or the jig is up. Trump as made some nice comments about 9/11 in the past, I think he's a truther but knows when to keep his power level down.

So much salt for autodidacts.

The best programmers are self taught, always have been, always will be, because rote learning is no substitute for talent and enthusiasm.

This is why (especially) female "programmers", but also those who chose CS as their degree purely in the hopes of a high salary, are so butthurt about "stereotypes" of programmers who spend most of their free time coding, learning about code, talking about code, coding some more and enjoying it.
They hate it because when the clock strikes 5 these "graduate professionals" are out the door and don't so much as think about a program until it's time to start work again the next day.
It causes them so much distress to realize that they aren't and never will be naturals, that they invented a fake mental condition in an attempt to make people stop drawing comparisons and induce guilt and pity.
"Imposter syndrome."
There is no "imposter syndrome", it is perfectly natural to feel inferior when all of your knowledge about a subject came from directed learning (i.e rote memorization) and progression was dictated by a curriculum instead of natural lust for knowledge.
Prior to programming being made "cool" and massive efforts made to recruit women, ALL programmers were largely self-taught and began their flirtation with computers and programming well before university and possibly even secondary school age.
Women and normal people scorned them, now they hate them and try to pull rank with impotent pieces of paper.

Carry on.

Alex actually hit the nail on the head when he used the phrasing "shaving votes"

I agree in priciple but we are mandated to have a secret ballot so any metadata on the votes that could be used to link back to the person voting could compromise that ideal. Still i think we otta throw it out and make it so you can audit the voting database and see how your vote was counted.

You can do it all in ints, just use modulus to account for any remainder that got scraped off or save your total before doing the skewing and then ensure your total still equals the saved total.

should we meme fix your site kike monkey?

Going to play devil's advocate for the sake of argument. The only instance I can think of in which there's a legitimate reason to use floats is if you need to store a very high number. Ints can go up to 2^32 which is 4 billion, but if you don't have those numbers memorized like I do you might abstain from using ints (though you'd just use a fucking long). Maybe the coder was an asshat and decided floats were best, they still retain their precision up to an extremely high value, much higher than ints. So maybe the GEMS programmer didn't know longs existed and used a float. Especially considering floats can go to higher numbers than longs and use less space. Also, gems isn't solely used for presidential elections. Maybe there's a case where they needed to count 8 trillion votes.

Anyways, that's absurd, I don't believe it was programmed without the intention of malice in mind. That was just a scenario.

10/10 post, except you didn't mention those who learned via rote memorization etc and do not have a passion for programming, will not generally go on to get a programming job and excell in that environment. Theyy will only get hiehired if tbeyre a mSeen too many of them when I was at uni.ority or a woman and there's affirmative action going on.

Sorry, in mobile so typing is shit. :^)

Isn't that a little odd? Only addition is required in the counting operation right? Surely this only requires an integer operation.

What is unsigned long long int? Integers in both C and C++ go to to 64bits user.

en.cppreference.com/w/cpp/language/types

0 to 18,446,744,073,709,551,615