Should we discontinue EVMs?

[Raja Bose]

The prime reason I support EVMs is that it eliminates the rigging probability at the local level. That is the major concern, not a country wide coup, in a scenario more fantastic than my own spy story.

<100 posts to go for Oldie, BTW.

Dileep saar, No fair! You really opened up the throttle in the past 2-3 weeks taking chankian advantage of me having busier time than usual at work.

[Pranav]

You do have photos of many of the voters on the voter's list, and you do have the people themselves.

And what are they going to do when the face of the voter actually gets texture mapped or made grainy enough to prevent a 100% positive id - have you seen what surveillance cam footage looks like? Morever, faking videos even ones showing authentic people is not a big deal anymore - just look at the fun Rooskies are having with some Amirkhan diplomat where they generated a video of him frolicking with a forbidden lady.

The guy making the fake video would also have to get the order of the voters right, otherwise it would be detected. Like the fat lady who voted right after the thin gentleman.

Actually as Dileep stated, the EVM eliminates the grassroots methods of rigging which our netas have mastered so well. Hence, one is extremely suspicious of the real reasons why many netas (much bigger fish than RM) are about the EVM.

So called grass-roots methods are quite easily detected and repolls can be held. That is why Indian elections until 2004 have been generally considered free and fair, with any fraud being too negligible to affect the outcome.

Hence, if you want to show that EVM based voting has been rigged (and I am not saying it is or it is not), it has to be proved through sane scientific methods. Rhetoric, fantastic conspiracy theories and considering everybody under the sun corrupt is not what is considered acceptable proof.

What we do have is a pattern of suspicious behavior of EVMs, coupled with extreme stone-walling by the EC to prevent any transparency. Burden of proof of supposed unriggability is on EVM supporters.

[Raja Bose]

The guy making the fake video would also have to get the order of the voters right, otherwise it would be detected. Like the fat lady who voted right after the thin gentleman.

Actually the faking is not done by some guy, it is done automatically online by a computer hence all those type of dynamic inputs get taken into consideration.

So called grass-roots methods are quite easily detected and repolls can be held. That is why Indian elections until 2004 have been
generally considered free and fair, with any fraud being too negligible to affect the outcome.

Please explain that to voters of Bihar, U.P., parts of West Bengal and see them go .

What we do have is a pattern of suspicious behavior of EVMs, coupled with extreme stone-walling by the EC to prevent any transparency. Burden of proof of supposed unriggability is on EVM supporters.

That is the exact reason why Dear Neta RM ji was requested to go to Delhi. Instead as usual he weaseled out claiming evil EC officials and engineers will use djinn energy to wiggle out of tough questions asked in public. Whats the use of about EC stonewalling - why not take the proactive step to ask them why they wont release EVMs to the public - how it hurts the credibility of the EVMs etc. Instead people like RM seem to be content with the current status quo (and secretly hope that the EC nevers allows public access to EVMs) since it provides them an issue to about and gives them full licence to indulge in fantastic conspiracy theories and milk it for other benefits.

[ArmenT]

What we do have is a pattern of suspicious behavior of EVMs, coupled with extreme stone-walling by the EC to prevent any transparency. Burden of proof of supposed unriggability is on EVM supporters.

I suggested a method to settle this here and posted my justifications for this here and here. Think they sound pretty reasonable?

[Pranav]

What we do have is a pattern of suspicious behavior of EVMs, coupled with extreme stone-walling by the EC to prevent any transparency. Burden of proof of supposed unriggability is on EVM supporters.

I suggested a method to settle this here and posted my justifications for this here and here. Think they sound pretty reasonable?

No, transparency of elections is a fundamental right. People should not be asked to pay exorbitant amounts. Nor can people afford the testing equipment - scanning electron microscopes, focussed ion beams, lasers etc, not to mention the cost of the labour.

[ArmenT]

That's why I suggested that amount. It is not unreasonable or exhorbitant at all and will pay for the lab-costs + widespread publicity of the test results as well (TV and newspaper coverage), so that every citizen can benefit from it. There is no need to purchase a scanning electron microscope, because good labs already have one and this will easily pay for renting one (or multiple labs, if you don't trust a single one). It also pays for technical people to come by and conduct the tests (engineers and scientists don't work for free ), scientific people who are nominated by all parties, preferably from private institutions, so that no one can later say that those scientists were bribed as well. Bigger parties can easily afford that amount and smaller parties can pool together to get the amount and they can share the costs together.

On the other hand, the amount is just big enough that no jokers can clog the system up. Only serious people who genuinely believe that the EVMs are tampered with will apply to have the tests done. Right now, it is entirely possible for people to print totally unfounded accusations about the EC with no repercussions whatsoever AND that achieves nothing. It is time to put-up or shut-up about the EVMs and this will let the truth come out. Incidentally, with the proposed scheme, the stakes are higher for the EC because if any evidence of tampering is found, those persons will be unceremoniously fired and jailed and will never be able to find a job in private or public industry again.

[Pranav]

^^^ The reason that EVMs were banned by a Court decision in Germany recently is that the average voter cannot verify for himself that the system is not rigged.

So it's not enough for a small number of "experts" to be convinced - each voter is important, so there has to be a system that meets that standard of transparency.

That is the difference between civilized countries like Germany, where humans have value, and brutal societies like India, where we have Election Commissioners with a history of baking people alive.

[Rahul Mehta]

.... All it takes is one peon or one clerk to blow the cover.

Somehow, that did not happen in Satyam for 8 years, Telgi scam for 5 years etc. And I can give 100s of examples where cover got blew only because of counter-copy. eg Raju claimed that he had deposits of 1000s of crores and banks said none. In Telgi scam, the fake stamp paper itself became proof. In case of EVM, the "fake" vote can never be proven as there is no counter foil.

So all this talk that "one clerk will blow up the cover" is nice novel story and movie script where a small guy beats the whole gang of thus with one hand tied behind his back. But it never happens in practice.

=======

By the way, that $220 million that you quoted is quarterly revenue, not annual. Here's their financial statements:
http://www.google.com/finance?client=ob&q=NASDAQ:MCHP
They have a market cap of $4.97 billion.

The company was in financial trouble. Also CIA has ways to designs from company and CIA can also get 100000 duplicate chips made with same functionality, same ROM code and a different microcode with trojan but otherwise same functionality. Now only 2-5 engineers inside BEL will be authorized to open the chip and scan the microcode, even if that is easy to do in a deterministic way. The other 100s of employees who directly come in contact with EVMs dont have powers to open the chip and dont have means and skills to scan the microcode. So if CIA has placed trojan in microcode and boxes are replaced after chip engineers scan the microcode, then trojan will never get detected.

====

Many moons ago, I used to buy chips. They were enclosed in plastic ruler like slides, where in chips would slide inside and come out. I dont recall the name of such cover. The 8051 chips IIRC were 5 cm long and some 1 cm wide. So 1 ft long ruler will have 6 such chips and that would be 0.75 inch wide. So 1ft * 1ft of such surface will hold 16*6 = 96 chips. This rulers height will be about 0.75 inch. So 1 cu ft box will store 96*16 = roughly 150 chips. So if each box is 2ft * 2ft * 2ft, it can hold 1000 such chips. So 100,000 chips would fit in 100 boxes or at most 200 boxes.

The store area is managed by semi-skilled people and is away from techie people. The store people go by labels and dont have faintest clue of what boxes have, and surely dont open the chips to test them. So finally, if BEL chief, Banglore head and BEL store keeper are compromised, they can somehow ensure that 100-200 boxes contained duplicate chips are placed closed to boxes containing actual chips. And swap the boxes once QA has cleared the chips. For Rs 100 cr, this is cery much manageable.

============

Re: The personnel in the plant can't read the binary: this has already been disproved multiple times and Mr. Mehta was even given links on how to do this + names of companies to do this (and he promptly ignored all the references and claimed it could be done for about 5 pages). You can read the details from around page 33 of this thread on how anyone can do this. Please don't repeat that assertion -- it doesn't make it any truer.

ArmenT,

The boxes are swapped AFTER ROM code is verified and lock bit is set. And later, the staff will NOT be authorized to open the chip and scan the ROM or microcode. So the 100s of employees who come into contact with EVM after chip testing will not be authorized or equipped or trained to scan the microcode.

[Rahul Mehta]

^^^ The reason that EVMs were banned by a Court decision in Germany recently is that the average voter cannot verify for himself that the system is not rigged.

So it's not enough for a small number of "experts" to be convinced - each voter is important, so there has to be a system that meets that standard of transparency.

That is the difference between civilized countries like Germany, where humans have value, and brutal societies like India, where we have Election Commissioners with a history of baking people alive.

The German courts gave a "nice" looking reason. But real reason was that population was anti-EVM because German population, which has far more tech information than commons in India could not be manipulated to believe that EVMs are unhackable. Now Supreme Court cannot officially say that "EVMs should be banned as they are hackable" because such reason though real doesn't look nice. So one has to make a diplomatically correct reason and give a nice reason.

In India, the elite openly says "who gives a damn about what citizens think". (Just look at the neighboring thread where many BRites are shamelessly opposing demand that citizens be allowed to put their complaints on PM's web-site -- it reflects the state of mind of Indian elitemen). And So whole idea of whether citizens approve/disapprove EVMs has been neatly bypassed by self appointed experts in everything from microcode to box placement.

And citizens of India dont give a damn about what experts say. Experts in Satyam claimed that Satyam had 1000s of crores of FDs, and there were none. Experts claims that stamp papers cannot be forged at large scale and they were liars. Experts in Supreme Court give bail to convicted pedophiles. Experts also claimed that Pokharan-2 was success and now we know the truth. And there are experts like Indirsen who without opening randomly chosen even 10 EVMs claimed that none of the 10,00,000 EVMs have any trojans !! Suffices to say that experts are no different from traffic constables when it comes integrity --- the price may be higher.

So lets keep so self appointed experts' opinions out of debate. They have no value what-so-ever as their integrity is doubtful.

[ArmenT]

=======

By the way, that $220 million that you quoted is quarterly revenue, not annual. Here's their financial statements:
http://www.google.com/finance?client=ob&q=NASDAQ:MCHP
They have a market cap of $4.97 billion.

The company was in financial trouble. Also CIA has ways to designs from company and CIA can also get 100000 duplicate chips made with same functionality, same ROM code and a different microcode with trojan but otherwise same functionality. Now only 2-5 engineers inside BEL will be authorized to open the chip and scan the microcode, even if that is easy to do in a deterministic way. The other 100s of employees who directly come in contact with EVMs dont have powers to open the chip and dont have means and skills to scan the microcode. So if CIA has placed trojan in microcode and boxes are replaced after chip engineers scan the microcode, then trojan will never get detected.

The first statement about the company being in financial trouble is total horse-puckey. Have you even checked their financial statements for the last few years. They've always been profitable. As for the rest of the conspiracy theories, those are your theories and are yet unproven. Please provide proof that any of what you've asserted is even remotely true. For instance, say that I confidently state that Rahul Mehta is an agent of <insert your favorite spy agency here> by taking your line of reasoning. See what I mean?

Re: The personnel in the plant can't read the binary: this has already been disproved multiple times and Mr. Mehta was even given links on how to do this + names of companies to do this (and he promptly ignored all the references and claimed it could be done for about 5 pages). You can read the details from around page 33 of this thread on how anyone can do this. Please don't repeat that assertion -- it doesn't make it any truer.

ArmenT,

The boxes are swapped AFTER ROM code is verified and lock bit is set. And later, the staff will NOT be authorized to open the chip and scan the ROM or microcode. So the 100s of employees who come into contact with EVM after chip testing will not be authorized or equipped or trained to scan the microcode.

Again, your statements are completely leading. How do you know that the staff will not be authorized to open the chips or scan the ROM or microcode?

[Rahul Mehta]

The first statement about the company being in financial trouble is total horse-puckey. Have you even checked their financial statements for the last few years. They've always been profitable. As for the rest of the conspiracy theories, those are your theories and are yet unproven. Please provide proof that any of what you've asserted is even remotely true. For instance, say that I confidently state that Rahul Mehta is an agent of <insert your favorite spy agency here> by taking your line of reasoning. See what I mean?

But RM did not make any EVMs. Here, Microchip made EVMs' chips and given that Microchip needs and gets lots of contracts from USG, Microchip cant deny a "request" from USG/CIA to make extra 100,000 chips with a different microcode.

How do you know that the staff will not be authorized to open the chips or scan the ROM or microcode?

BEL documents say that only 4-5 people know the source code. If 10-20 people were authorized to open the chip and scan the gate level design and also read the ROM code, then BEL would have said that 10-20 people knew the source code. Also, it is not standard procedure to open the chips AFTER chips are tested and put on the PCBs.

====

Dileep, Raja Bose, ArmentT etc

Pls put the names of the equipment by which I can scan the gate level design and obtain the VHDL code for what the gate level design does. IOW, you claim that reading gate level design and microcode is piece of cake. Pls show how it is done and how easy/difficult it is. And can any of you guys do it? Do you have equipment in places you work or worked to scan the chips at gate level? Have you yourself scanned the chips at gate level?

[ArmenT]

The first statement about the company being in financial trouble is total horse-puckey. Have you even checked their financial statements for the last few years. They've always been profitable. As for the rest of the conspiracy theories, those are your theories and are yet unproven. Please provide proof that any of what you've asserted is even remotely true. For instance, say that I confidently state that Rahul Mehta is an agent of <insert your favorite spy agency here> by taking your line of reasoning. See what I mean?

But RM did not make any EVMs. Here, Microchip made EVMs' chips and given that Microchip needs and gets lots of contracts from USG, Microchip cant deny a "request" from USG/CIA to make extra 100,000 chips with a different microcode.

Now you're talking trash again. Look, these guys make lots of chips a year and they ship them out to large suppliers around the world. Microchip do not often know where the chips are going because people purchase the chips from the large suppliers, not necessarily directly from Microchip. For instance, I get all my PIC chips from Digi-key. Also, Microchip have no idea whether particular chips are being used for an EVM or another project because people purchase large supplies of chips for other projects from them already. These are general purpose microprocessors, not specialized chips with only one function, but are programmed by the buyer according to their need. That puts a very huge damper on your theory.

By the way, I'm pretty certain this has been mentioned before and you're simply repeating something that has been debunked already.

As for RM not making EVMs, that has nothing to do with you being an agent for <insert your favorite agency here>. You could be paid by some agency to create mayhem and try to discredit honest elections for instance. This assertion is entirely in line with the logic you've been using so far.

Dileep, Raja Bose, ArmentT etc

Pls put the names of the equipment by which I can scan the gate level design and obtain the VHDL code for what the gate level design does. IOW, you claim that reading gate level design and microcode is piece of cake. Pls show how it is done and how easy/difficult it is. And can any of you guys do it? Do you have equipment in places you work or worked to scan the chips at gate level? Have you yourself scanned the chips at gate level?

This has been already given to you many times already. Equipment, books, theory, companies that do this for you, whether we have access to the equipment, who has scanned, the whole 9 yards. See pages from around page 30 onward for multiple times that all this has already been answered. No one can help you if you simply choose to deny that information.

[Rahul Mehta]

... these guys make lots of chips a year and they ship them out to large suppliers around the world. Microchip do not often know where the chips are going because people purchase the chips from the large suppliers, not necessarily directly from Microchip. For instance, I get all my PIC chips from Digi-key. Also, Microchip have no idea whether particular chips are being used for an EVM or another project because people purchase large supplies of chips for other projects from them already. These are general purpose microprocessors, not specialized chips with only one function, but are programmed by the buyer according to their need. That puts a very huge damper on your theory.

Someone mentioned that these chips are custom made for EVMs. If these chips are off the shelf and only ROM is put in the chip, then putting trojan in EVMs by replacing boxes became easier !! CIA doesnt need any extra co-operation from Microchip. They can get details of this off the shelf chip without any hassle and then create another chip with same ROM code, same microcode but an extra function in microcode which increases the votes of the preferred candidate number. They need to provide 100 such boxes, each box containing 1000 chips. Once the testing is over, the top 5-6 guys now need to only replace these 100 boxes which came from Microchip with 1000 boxes which came from CIA. And no one will know, as functionally the chips are identical.

This has been already given to you many times already. Equipment, books, theory, companies that do this for you, whether we have access to the equipment, who has scanned, the whole 9 yards. See pages from around page 30 onward for multiple times that all this has already been answered. No one can help you if you simply choose to deny that information.

I could not find any way to read the gate level design. You are welcome to put one consolidated articles on "how EVMs can be verified". IOW, say you have 1000 EVMs and all the documents you need. Now which equipment , procedure etc would you use to ensure that

1. gate level design in the document has no trojan
2. gate level design of the chip in given EVM is same as the one in the document
3. ROM code has no trojan in random bytes
4. ROM code in chip is same as what it is in the document
5. PCB is same as that in the document.

How much will it cost to verify each of these 1000 EVMs, mere 0.1% of the sample?

[Dileep]

Let me make some clarifications here.

1. The controllers are NOT custom ASICs. They are standard products, with the ROM alone is customized. The gate level design is owned by the manufacturer, and is not made available to the user.

2. If you want to verify the chip, all you need to do is to buy an off the shelf version chip, and compare the silicon. Only the ROM area will be different. This is NOT a production practice. I mention it here because it is a means of forensics. IF someone want to verify, they can.

3. The chips are NOT packed in the tubes (chute is the correct name). They are packed in trays. There are different standards for trays, but 1000 chips per carton is a normal practice.

As of now, the only standing argument is that the corrupt chips will be smuggled into the BEL stores and swapped with the good chips. RM have conveniently ignored to explain how this will be done, but what else is new?

1. It is impossible to smuggle the 100,000 chips.
2. It is impossible to fake a project for these chips. A large volume order needs a backing purchase order from a customer.

[Dileep]

.... All it takes is one peon or one clerk to blow the cover.

Somehow, that did not happen in Satyam for 8 years, Telgi scam for 5 years etc.

Ref to the context of that claim. We are talking about a number of DC offices and a number of people. Satyam and Telg scams were done at a totally different level. There is no comparison.

The company was in financial trouble.

No, they just had a "lower than expected" sale in that quarter. They still made a profit and paid dividend. Their reserves and surpluses figures are good.

Also CIA has ways to designs from company and CIA can also get 100000 duplicate chips made with same functionality, same ROM code and a different microcode with trojan but otherwise same functionality. Now only 2-5 engineers inside BEL will be authorized to open the chip and scan the microcode, even if that is easy to do in a deterministic way. The other 100s of employees who directly come in contact with EVMs dont have powers to open the chip and dont have means and skills to scan the microcode. So if CIA has placed trojan in microcode and boxes are replaced after chip engineers scan the microcode, then trojan will never get detected.

You have not yet resolved the trivial issue of getting the 100,000 chips into the BEL stores first.

Many moons ago, I used to buy chips. They were enclosed in plastic ruler like slides, where in chips would slide inside and come out. I dont recall the name of such cover. The 8051 chips IIRC were 5 cm long and some 1 cm wide. So 1 ft long ruler will have 6 such chips and that would be 0.75 inch wide. So 1ft * 1ft of such surface will hold 16*6 = 96 chips. This rulers height will be about 0.75 inch. So 1 cu ft box will store 96*16 = roughly 150 chips. So if each box is 2ft * 2ft * 2ft, it can hold 1000 such chips. So 100,000 chips would fit in 100 boxes or at most 200 boxes.

Well, I said 1 cuft for 1000 chips. What did you prove here?

The store area is managed by semi-skilled people and is away from techie people. The store people go by labels and dont have faintest clue of what boxes have, and surely dont open the chips to test them. So finally, if BEL chief, Banglore head and BEL store keeper are compromised, they can somehow ensure that 100-200 boxes contained duplicate chips are placed closed to boxes containing actual chips. And swap the boxes once QA has cleared the chips. For Rs 100 cr, this is cery much manageable.

Well, if they can "somehow" do it, someone else also can "somehow" figure out what they are up to, right?

You have not established how those chips to be swapped would come to the stores. Please do that first.

The boxes are swapped AFTER ROM code is verified and lock bit is set. And later, the staff will NOT be authorized to open the chip and scan the ROM or microcode. So the 100s of employees who come into contact with EVM after chip testing will not be authorized or equipped or trained to scan the microcode.

[/quote]
You have not established how those chips to be swapped would come to the stores. Please do that first.

[Dileep]

The German courts gave a "nice" looking reason. But real reason was that population was anti-EVM because German population, which has far more tech information than commons in India could not be manipulated to believe that EVMs are unhackable. Now Supreme Court cannot officially say that "EVMs should be banned as they are hackable" because such reason though real doesn't look nice. So one has to make a diplomatically correct reason and give a nice reason.

This was already explained many times, that the architecture of the EVMS in EU is different than what we use in India.

(Just look at the neighboring thread where many BRites are shamelessly opposing demand that citizens be allowed to put their complaints on PM's web-site -- it reflects the state of mind of Indian elitemen).

You yourself said that you present it as a means of citizens be allowed to put their complaints on PM's web-site for the sole purpose of Anyone who opposes this EN can be now portrayed as anti-RTI and pro-corruption

That reflects on the state of mind of YOU.

And So whole idea of whether citizens approve/disapprove EVMs has been neatly bypassed by self appointed experts in everything from microcode to box placement.

And you call ad-hominem attack on you and cry foul.

And citizens of India dont give a damn about what experts say.

Well, then why do you bother about the experts? What are you doing here?, instead of going to the citizens directly?

So lets keep so self appointed experts' opinions out of debate. They have no value what-so-ever as their integrity is doubtful.

Well, if the opinions have no value, why bother?

And you claim the integrity is doubtful. Your own post, quoted below, establishes yours.

I have added the buzzwords such as "RTI application" and "complaint against corruption". But, yes, any damn text will go in the affidavit and that is intended. And I have not added buzzwords to please you --- many of my activists proposed that adding words like "RTI application", "complaint against corruption" would make it difficult for 4-digit IQ+ walla intellectuals to oppose this proposal. And I think they are right. Anyone who opposes this EN can be now portrayed as anti-RTI and pro-corruption. And so I decided to add these buzzwords in the text of the proposed EN.

Maybe, the concept of integrity is different for you.

[Dileep]

Pls put the names of the equipment by which I can scan the gate level design and obtain the VHDL code for what the gate level design does. IOW, you claim that reading gate level design and microcode is piece of cake. Pls show how it is done and how easy/difficult it is. And can any of you guys do it? Do you have equipment in places you work or worked to scan the chips at gate level? Have you yourself scanned the chips at gate level?

It was never claimed that the gate level design could be read from the chip. What WAS claimed are:

1. It is possible to compare two samples of chips, to verify if they are different in ANY aspects, down to the elvel of a single transistor or a single interconnect. You need a microscope. A good chip inspection system would make it easier to do.

2. It is possible to read the Mask ROM code from a chip. All you need is a microscope.

3. It is possible to read the electrically programmable (flash/eprom) data from a chip. You need a chip prober equipment for that.

[Dileep]

Someone mentioned that these chips are custom made for EVMs.

Later information shows that they are OTP controllers. These are semi-custom products where the processor system is common, with the ROM area programmed by modifying the mask.

If these chips are off the shelf and only ROM is put in the chip, then putting trojan in EVMs by replacing boxes became easier !!

These are NOT off the shelf. You need to order them with the supplier, along with the code to be programmed.

CIA doesnt need any extra co-operation from Microchip. They can get details of this off the shelf chip without any hassle and then create another chip with same ROM code, same microcode but an extra function in microcode which increases the votes of the preferred candidate number. They need to provide 100 such boxes, each box containing 1000 chips. Once the testing is over, the top 5-6 guys now need to only replace these 100 boxes which came from Microchip with 1000 boxes which came from CIA. And no one will know, as functionally the chips are identical.

It is possible to custom fabricate such an IC, but a chip inspection would readily show that. A power drain test will also show the difference.

We come back to the issue of getting the corrupt chips into the BEL stores, and you have not answered that yet.

[ArmenT]

^^^^
Saar, just out of curiosity, where does Hitachi come into this now? I'm assuming that the BEL and EC folks ordered the processors from Microchip (or a general purpose electronics supplier, along with a bunch of other electronic components for better discount) and get the ROM from Hitachi? This is still correct, right?

I was assuming they were using Super-H processors because they were sending the chips to Hitachi .

[Dileep]

ArmenT, BEL said the chips are from MicroChip, and ECIL said they are from Microchip and Renesas.

We also know that ECIL was the primary contractor, and BEL designed their own box independently.

So, my take is that BEL chose MC and ECIL chose Renesas for their implementation of the design. ECIL, being the primary contractor, refers to both in the RTI reply. BEL deals with only MC, so they mentioned that in theirs.

Both are controllers with OTP ROM. There is no separate code memory, as clear from the Indiresan report. So, BEL orders the chips from MC, most probably through Avnet India, using their OTP order system. ECIL orders their chips from Renesas, through their Indian rep (not sure who. Never dealt with them). Renesas has published their ROM order system in great detail on their web site.

[ArmenT]

^^^^
Ah ok. By the way, for the benefit of those who are a bit lost, Microchip is a manufacturer that makes a family of processors called PIC series. Renesas is a joint venture of Hitachi and Mitsubishi and they make their own processor family called the SuperH. Per RTI, BEL uses only Microchip processors (PIC processors) and ECIL uses Renesas (SuperH processors) and Microchip processors (PIC). Note that these two processor families are not binary compatible, so code for one will NOT run on the other.

What this means is that to reliably affect EVMs at the microcode level, an attacker will first have to infiltrate three different companies (Microchip, Hitachi and Mitsubishi) that are headquartered in different countries and this is before the problem of smuggling all those chips into BEL and ECIL, not to mention the problem of evading detection at the manufacturer level at BEL and ECIL.

In non-electronics terms, it is like a car rental place buying cars from two different manufacturers from different countries (say, Toyota and BMW). Both cars will get you from point A to point B, but they cannot interchange parts.

[Dileep]

^^^I do not think ECIL makes two kinds of EVMS. The more plausible route is BEL making their design based on PIC and ECIL making their design based on SuperH. So, there are two hardware designs and two softwares, which are totally different, as PIC and SuperH are NOT source compatible.

Even the mechanicals are different, so the ballot paper inserts and labels have different dimensions.

And BOTH companies manufacture them at their respective facilities that makes defense production which are under utmost security.

[ArmenT]

Yep that's what I implied, though I didn't explictly mention it. Didn't know about the different ballot paper inserts and labels though. Thanks for clearing that up .

[vera_k]

Video demonstrating low-tech way of tampering with EVMs.

http://www.youtube.com/watch?v=2F0UHrBW ... r_embedded#

[Muppalla]

Video demonstrating low-tech way of tampering with EVMs.

http://www.youtube.com/watch?v=2F0UHrBW ... r_embedded#

This is not possible unless the agents of the parties are stupid. They know their symbol's position well before the start of voting.

[Dileep]

Read the RO's manual and understand how the printed ballot is placed in the machine and sealed.

[vera_k]

Video demonstrating low-tech way of tampering with EVMs.

http://www.youtube.com/watch?v=2F0UHrBW ... r_embedded#

This is not possible unless the agents of the parties are stupid. They know their symbol's position well before the start of voting.

All I can say is that this was reported to have happened in the Arunachal elections recently, but the tampering was detected by the people when the polling was in progress. It's the EVM equivalent of booth capturing, but in this case, it must have been before the EVMs ever got to the booth.

Repolling in 28 Arunachal booths

At the Hari polling booth of Ziro-Hapoli constituency in Lower Subansiri district, the poll process was suspended abruptly following damage of EVMs by people after detecting a fake ballot paper pasted over the original ballot paper of the EVM. Only 38 persons had exercised their franchise when the machine was damaged. Three AITC workers suspected to be involved in the incident have been arrested. In Hija, also under under Lower Subansiri district, polling was adjourned midway after detection of fake ballot papers.

[ArmenT]

^^^
Not necessarily. One of the early voters could have brought a duplicate ballot sheet containing the symbols and pasted it on top of the machine. The real ballot sheet is actually enclosed in a transparent plastic shield and sealed, so they can't easily access it. That strategy depends on the fact that the following voters, never having seen an EVM before, would have thought that the duplicate ballot sheet is the real one. However this strategy has a huge flaw that a person who has seen the EVM before and knows that the ballot sheet should be under a plastic shield can easily detect it. I'm guessing that is why the deception was detected so quickly.

[vera_k]

The plastic shield doesn't seem like much of a defence against this tactic because it would be easy enough to add a laminated ballot paper in place of plain paper. Plus we don't know how and when the fake ballot paper was added or how it was discovered to say if the security measures were adequate.

What I would say is that if so many people can be fooled by such an easy trick, it strengthens the case for adding paper trial functionality to the EVM.

[Dileep]

Well, what exactly are the perps going to achieve by this? Send the votes for your opponent to an independent? Doesn't make much sense.

I don't think it is possible to "paste" a ballot while the polling happens. The BU is kept on a table, and there is a shade around it that is just a foot or so in height. A voter can not paste something on the machine without being observed.

I believe this was just a stunt. Note that the EVM was damaged by the mob, so the evidence of the action is gone. How convenient!!

[vera_k]

Well, what exactly are the perps going to achieve by this? Send the votes for your opponent to an independent? Doesn't make much sense.

No, the intent would be to redirect your opponents votes to your EVM key in areas where you knew your opponent has more support than you.

I believe this was just a stunt. Note that the EVM was damaged by the mob, so the evidence of the action is gone. How convenient!!

Tampered EVMs were detected in two booths. Only one was damaged, while the other is safe.

In Hija, the polling was adjourned midway after detection of similar fake ballot paper pasted over the original one, however the EVM is safe. The local police, is enquiring into the fake ballot paper episode.

I don't think it is possible to "paste" a ballot while the polling happens. The BU is kept on a table, and there is a shade around it that is just a foot or so in height. A voter can not paste something on the machine without being observed.

Okay, so we can rule that possibility out.

[Dileep]

[quote="vera_k"]
No, the intent would be to redirect your opponents votes to your EVM key in areas where you knew your opponent has more support than you.

If the place have more support, then there will be agents of your opponent very active. So, it becomes almost impossible to pull this off.

Copies of the ballot paper are given to all agents, and the PO does the demo and mock poll, so unless the agents of the opponent are absent, double agents, or really stupid, this can not be pulled off.

[vera_k]

If the compromise happened due to the involvement of someone in the election staff, then all of the safeguards will not work. Because if this person can add a fake ballot paper by accessing the EVM, then he can also give the agents the same fake ballot paper.

I doubt that the local election staff was involved in this. Because corrupt election staff has easier ways of adjusting the votes polled by the EVM after the fact.

Can you trust the vote machine?

Nothing stops a corrupt election official with access to the seal from simply erasing the data, punching in an equal number of votes for the candidate of his choice — which will take a maximum of about five hours per machine — and reaffixing the seal.

The agents present at the time of counting the votes are not the same as those present when the machines were sealed.

But then, maybe these were newer model EVMs with the timestamp functionality, and thus the riggers had to resort to this method of "fixing" the devices.

[Dileep]

Well, in that case, stuffing ballots is easier, because it will not take 5 hours to do.

A single election official can not manipulate the ballot paper. There are a number of check points for that.

[Tanaji]

More fodder for Rahul Mehta

http://news.rediff.com/report/2009/oct/ ... ys-bjp.htm

Naqvi had told reporters that EVMs had turned into Electronic Victory Machines of the Congress and this had led to the BJP's defeat. Later, he retracted his statement, saying the causes of the loss will have to be analysed once all the results are declared.

Of course Rahul Mehta will claim that the above is proof positive that they are fixed and AWMTA

[Rahul Mehta]

More fodder for Rahul Mehta

http://news.rediff.com/report/2009/oct/ ... ys-bjp.htm

Naqvi had told reporters that EVMs had turned into Electronic Victory Machines of the Congress and this had led to the BJP's defeat. Later, he retracted his statement, saying the causes of the loss will have to be analysed once all the results are declared.

Of course Rahul Mehta will claim that the above is proof positive that they are fixed and AWMTA

I am not with BJP on this issue.

For all I know, BJP is NOT anti-EVM. If BJP were anti-EVM, they could have forced EC to use paper ballots by fielding 64 independents. But they didnt do it. So BJP is not anti-EVM.

And I am not surprised. The "agency" which got duplicate chips made in US, and got BEL store-keeper and BEL chief to swap 25-50 boxes inside BEL store rooms is powerful enough, and has deep roots in BJP as well. So BJP will never ever dare to oppose EVMs.

[sanjaychoudhry]

This "Congress wave" that India is witnessing only began after EVMs were introduced. Before that the party was in decline everywhere and BJP was ruling a majority of the states. The wave -- which has survived multiple bomb blasts all over India and anti-incumbency -- is actually arising from Chawla's office. It seems we will have to live with "shock victories" of Congress and "shock defeats" of BJP as long as EVMs and Chawla are around.

The results are totally artificial. Delhi was the main theatre where EVM manipulation was first practised under the safe nose of Congress HQ. This explains victory after victory of Sheila Dixit while BJP kept winning in every other state where elections were held at the same time as Delhi elections. Once the courtiers perfected the technique in Delhi, then they rolled it over to other states.

BJP should start an aggressive and violent campaign to ban EVMs if it ever hopes to come to power again. Otherwise the courtiers will run away with Indian "democracy." The goras are fully behind Congress in this to keep "Hindoo nationalists" away from power (just in case they introduce anti-conversion bill in other states).

The BJP's coming to power in the centre had spooked the Goras as well as the courtiers and they knew where things were headed if they did not do anything soon. This EVM thingie is the result of their collective brain-storming.

[Dileep]

And I am not surprised. The "agency" which got duplicate chips made in US, and got BEL store-keeper and BEL chief to swap 25-50 boxes inside BEL store rooms is powerful enough, and has deep roots in BJP as well. So BJP will never ever dare to oppose EVMs.

Well, please bring forth proof that at least in a few constituencies, there was a rush of cancellation at the last moment.

Then, please bring proof that BEL or ECIL ordered chips for 'another project' for replacement.

[KarthikSan]

I'm not claiming a conspiracy theory of any kind or am I siding with EVM rigging or not. But to a layman observer from the outside the victory of a relative nobody like President Patil's son against Sunil Deshmukh suggests there is something fishy going on. Especially, considering the fact that even Congress workers were working against Shekawat

[Rahul Mehta]

1. Well, please bring forth proof that at least in a few constituencies, there was a rush of cancellation at the last moment.

2. Then, please bring proof that BEL or ECIL ordered chips for 'another project' for replacement.

1. DCs dont publish which person withdrew at what time. All they publish is total number of withdrawals. In almost every seat, number of withdrawals are quite large, sometimes as high as 50% of those who file.

2. BEL does not disclose all projects details on web. And BEL must be having 100s of projects for which they order chip. One can use a new project or existing project to order the chip and with co-operation of store-manager, boxes can be swapped within 30 minutes. I dont have a degree in "box placement", but I know from my experience that box replacement is TRIVIAL task.


----

Essentially, your argument is :

1. There are 15,00,000 constables in India (a fact)
2. provide proofs that over 1500 of constables take bribe
3. See you have no proof that 1500 constables are corrupt.
4. Hence 99% of the constables are non-corrupt.

---

I have proved that EVMs can be rigged by putting a trojan at gate level design in the chip, and then implementing the chip in such a way that power consumption increases by less than 1%. I have checked with people who work in PCB design as well as manufacturing. They NEVER check gate level design and they only occasionally they do power checks, but as long as chip is within range (1% to 5% of rated consumption), they dont bother. So if rigged chip consumes power that is just 1% more than actual chip, no box-replacement is needed. CIA can simply ask the chip supplier to send rigged chip and since gate level design is not scanned in BEL, the chip will sail thru. And in case BEL has procedure to scan the chip, then and then only CIA has to resort to box swapping. And that too is a trivial task, needs only 30 minutes to replace 25-50 boxes.

====

1. This "Congress wave" that India is witnessing only began after EVMs were introduced. Before that the party was in decline everywhere and BJP was ruling a majority of the states. The wave -- which has survived multiple bomb blasts all over India and anti-incumbency -- is actually arising from Chawla's office. It seems we will have to live with "shock victories" of Congress and "shock defeats" of BJP as long as EVMs and Chawla are around.

2. The results are totally artificial. Delhi was the main theatre where EVM manipulation was first practised under the safe nose of Congress HQ. This explains victory after victory of Sheila Dixit while BJP kept winning in every other state where elections were held at the same time as Delhi elections. Once the courtiers perfected the technique in Delhi, then they rolled it over to other states.

3. BJP should start an aggressive and violent campaign to ban EVMs if it ever hopes to come to power again. Otherwise the courtiers will run away with Indian "democracy." The goras are fully behind Congress in this to keep "Hindoo nationalists" away from power (just in case they introduce anti-conversion bill in other states).

4. The BJP's coming to power in the centre had spooked the Goras as well as the courtiers and they knew where things were headed if they did not do anything soon. This EVM thingie is the result of their collective brain-storming.

sanjaychoudhry,

IMO, you are more loyal than the king. Every second BJP/RSS worker I meet in Ahmedabad is anti-EVM. But no BJP leader is against EVMs. And if BJP leaders are against EVMs, they do not need violent and aggressive campaign to get rid of EVMs. All they need to do is to put 64 independents and ECI will be forced to use paper ballots, as existing EVMs can take no more than 64 candidates. The EVMs are rigged by Agency and NOT Sonia/Congress or Mahatma Manmohan. Sonia, Rahul, Mahatma Manmohan etc do not have expertise or strength to rig EVMs and keep it a secret. And the same Agency has enough deep roots in BJP leaders. BJP leaders were nationalists, and today they are "multi-nationalists". and so no BJP leader, including Modi, is daring to oppose EVMs. A campaign against EVMs is not a campaign against Congress, but would be a campaign against those who are in-charge of EVM-rigging (aka agency in my lingo and Gora in your lingo), something no BJP leader would dare to initiate. So if you are anti-EVM, following BJP leaders would be waste of time and energy. BJP leaders will in the end sell out and settle for "audit by PwC", and we know what kind of audit PwC did.

I'm not claiming a conspiracy theory of any kind or am I siding with EVM rigging or not. But to a layman observer from the outside the victory of a relative nobody like President Patil's son against Sunil Deshmukh suggests there is something fishy going on. Especially, considering the fact that even Congress workers were working against Shekawat

I dont know about this specific case, but I get many mails citing too many such examples.