Should we discontinue EVMs?

Pranav · Post by **Pranav** » 15 Jul 2009 19:51

Dileep wrote:Pranav, I have no interest in personal attacks. You asked for "mathematical proof". It takes time and effort for me to present the technical information. I don't want to waste it on someone who can't understand it, the "willingness to understand" notwithstanding.

I am pissed off on people with hardly any knowledge on the technology (and it is not rocket science either) making expert comments (like your millions of transistors and interconnects) Which will then be taken by others who have still less knowledge and propagated. I wouldn't be surprised if the phrase "millions of transistors and interconnects" itself appear elsewhere in the context of EVMS (Or you yourself might have copied it from somewhere).

Rahul Mehta at least makes claims based on what he understands.

Dileep, I have no desire to talk about myself, but you force me to remark that my ability to comprehend mathematical proofs is probably enough to dissect whatever you can produce. If you have a protocol that provably, fully characterizes a black-box, let us know.

Even if you can provide a link to a peer-reviewed Journal paper that explicitly makes such a claim in a theorem, that would be useful. I'd be happy to accept any method that guarantees full verification, if and when such verification is actually permitted by the EC. Meanwhile, the electronic voting mechanisms remain suspect.

As regards transistor counts, all that matters is that the systems are complex enough to make detailed verification an expensive proposition. Let's not go into irrelevant tangents.

Rahul Mehta · Post by **Rahul Mehta** » 15 Jul 2009 20:09

Tanaji,

Does EVMs have clock in them?

Also you claim that Totalizer cannot change CU data. Have you run any test? Has anyone you PERSONALLY know run any test? How do you say that Totalizer cannot change CU data?

.

Rahul Mehta · Post by **Rahul Mehta** » 15 Jul 2009 20:29

pgbhat wrote:In all this debate about fraudulent EVMs .... can anybody give the expected number of people that have to be involved in perpetrating this fraud?

Depending on how EVMs were rigged.

If Totalizer was used and rigged, I need just 20 persons (CEC Chawala, next 10 top EC guy and 10 guys to change the code in 5000 Totalizer in the last week before dispatch.)

If Totalizer is not used, and EVMs are stored in 1-4 locations in India, then I need co-operation of top 10 guys and about 10-20 field agents to replace the cartons containing good EVMs with cartons containing rigged EVMs. Each carton contains say 20 EVMs. So to replace 700,000 EVMs, I need to replace 35000 cartons over a period a month i.e. replace 100 cartons a day or say 200 cartons a day. Do-able with 10 field agents. And replace these EVMs with real ones when they come back into warehouse. Also, after the counting, the real CU will be fed with electronic signals so that they will have same data as rigged one had disclosed on the counting table.

So say an EVM's serial number is say 100898, just for an example. Then rigged one will have same serial number 100898, and we call it R100989. Now R100989 give 5% more votes to candidate no. 2 (Congress guy) and reduces votes of rest. Now say in reality, 1000 voters voted for BJP and 800 voted for Congress. Then R100989 will register 950 votes for BJP and 850 votes for Congress. And thats what it would show on table. And the field agents in the secure location needs to give electronic signal to real CU so that real CU has 950 votes for BJP and 850 votes for Congress. This would need a PC with several ports that simulate protocol between CU and EVM. One PC can easily feed 1500 CUs with appropriate hardware designed in advance. And then when rigged EVMs come, they need to be replaced back with real ones containing data fed by PCs, not voters.

The cartons will contain rigged EVMs which look like original EVMs and have same logo, sticker and keys. This would need large engineering facility off shore and would cost about Rs 500 cr for making some 700,000 EVMs.

----------------

Dileep,

To change paper ballots, how many guys would you need? Why are you refusing to compare the number of people needed to rig EVMs vs number of people needed to rig paper ballots?

.

suryag · Post by **suryag** » 15 Jul 2009 20:38

Whatever Dileep has proposed in the earlier page is by far the most comprehensive way of making sure that the code does what it is supposed to.

Pranav JTAG is a boundary scan protocol that the micro processor manufacturer puts into the One time Programmable(OTP) microcontroller. We could always suck out the binaries from an EVM and then check it for correctness. Anyways, Dileep sir has given all possible safeguards for the s/w part of the EVM. What i am worried is the PCB on the ballot unit which basically converts the button input to a unique sequence. With PCBs being multiple layers you could always do some interesting things with them, but the easiest way to protect the EVMs would be to allow the Executive officer of a polling booth to pick up the button-party assignment on the polling day. That should dumb out all PCB related maneouvres.

Raju · Post by **Raju** » 15 Jul 2009 20:40

If Totalizer was used and rigged, I need just 20 persons (CEC Chawala, next 10 top EC guy and 10 guys to change the code in 5000 Totalizer in the last week before dispatch.)

this is the most likely route then.
Also Gujarat didn't have totalizer because the state IB and PD are not in a terminal coma like the others.

Pranav · Post by **Pranav** » 15 Jul 2009 21:01

suryag wrote:Whatever Dileep has proposed in the earlier page is by far the most comprehensive way of making sure that the code does what it is supposed to.

Pranav JTAG is a boundary scan protocol that the micro processor manufacturer puts into the One time Programmable(OTP) microcontroller. We could always suck out the binaries from an EVM and then check it for correctness. Anyways, Dileep sir has given all possible safeguards for the s/w part of the EVM. What i am worried is the PCB on the ballot unit which basically converts the button input to a unique sequence. With PCBs being multiple layers you could always do some interesting things with them, but the easiest way to protect the EVMs would be to allow the Executive officer of a polling booth to pick up the button-party assignment on the polling day. That should dumb out all PCB related maneouvres.

What is actually needed is full verification of both software and hardware. If EVMs have to be used, then what is needed is:

1. Publishing complete specifications of software and hardware of EVMs, totalisers etc.
2. A methodology that can provably and fully verify the software and the hardware.
3. Allowing full post-poll audit by experts nominated by candidates. Also, chain-of-custody issues between the poll and the audit have to be resolved.

If and when all these conditions are satisfied we can consider accepting EVMs. Until then EVMs have to be considered suspect.

If we can't satisfy these conditions, then we should follow the example of the many sensible nations that have dumped their EVMs.

Tanaji · Post by **Tanaji** » 15 Jul 2009 21:27

Mehtaji,

Go to the nearest public library.
Remove 100 books
Replace them

Then report to us if you were able to put them back in the same shelf as they were before, how much time it took to do the replacement and were you observed in this exercise. Also report on how many times errors were made when the book landed on the wrong shelf.

Then compare to 700000 or whatever number you are using these days.

ramana · Post by **ramana** » 16 Jul 2009 04:19

Do we have a list of places where there was potential for EVM rigging and compared the final tally?

Muppalla · Post by **Muppalla** » 16 Jul 2009 04:47

Raju wrote:
If Totalizer was used and rigged, I need just 20 persons (CEC Chawala, next 10 top EC guy and 10 guys to change the code in 5000 Totalizer in the last week before dispatch.)
this is the most likely route then.
Also Gujarat didn't have totalizer because the state IB and PD are not in a terminal coma like the others.

Are totalizers used in LS elections? There is no news that either says yes or no. However, most of news are saying that they are tested prior to LS elections.

Rahul Mehta · Post by **Rahul Mehta** » 16 Jul 2009 06:44

Tanaji wrote:BTW, those who accuse of a mass conspiracy: here is a suggestion. At your place of work, try bribing the accountant/cashier whoever, to transfer Rs. 10 to you on a regular basis from your company's account illegally. See how long that conspiracy lasts, with just 2 people in the loop. And then compare it with your proposed conspiracy where 100s are involved.

Tanaji,

If your theory was true, then Raju-Satya scam would not have happened, where in forged fixed deposit certificated of Rs 1000s of crores were used. But Satyam-Raju scam did happen, and hence your theory is wrong.

Essentially, people do small theft, doesnt mean that large thefts are not possible.

-----

Tanaji wrote:Mehtaji,

Go to the nearest public library.
Remove 100 books
Replace them

Then report to us if you were able to put them back in the same shelf as they were before, how much time it took to do the replacement and were you observed in this exercise. Also report on how many times errors were made when the book landed on the wrong shelf.

Then compare to 700000 or whatever number you are using these days.

In the steps I proposed, the cartons are replaced by another cartons. One carton has say 20 EVMs. Then to replace 700,000 EVMs, I will need to replace 35000 cartons over a period of one month. This is do-able with 100 field agents.

Dileep · Post by **Dileep** » 16 Jul 2009 07:26

[quote="Rahul Mehta]]If Totalizer is not used, and EVMs are stored in 1-4 locations in India[/quote]

The EVMS are stored by the District Election Officers at the district headquarters. That throws out 90% of your arguments.

What remains?

Dileep · Post by **Dileep** » 16 Jul 2009 07:29

Pranav, you don't want to talk about yourself because there is nothing to talk. That kind of establishes the weight of your arguments isn't it?

Also, whoever understands how things work, got it already. I am not wasting any more time on that.

Moving on...

Dileep · Post by **Dileep** » 16 Jul 2009 07:31

Raju et al.

The totalizers were not used in this election. Give that a rest please!! When someone tests a system, would they use a RIGGED one?

Let us discuss the future scenario where totalizers will be used.

Dileep · Post by **Dileep** » 16 Jul 2009 07:34

Surya, any static modification on the PCB or cable would show up readily in the mock poll. The only plausible thing is the "key activation" and that needs a compromised binary. The audit report posted by KVRao details them.

Dileep · Post by **Dileep** » 16 Jul 2009 07:43

Totalizers:

Totalizer is a machine, very similar to the EVM. Most probably using the same microcontroller system. The only way to compromise it is CIA replacing the machine.

We are yet to see the security procedures in usage and handling of the machines. They will be published in the handbooks of the elections where they will be used. Looking at the affairs right now, I am sure those will be debated to the death when they are (if at all) introduced.

A compromised totalizer can mess the count results. But to change the counts that is stored in the EVM, you need a compromised EVM. We have established how difficult (or how easy, depending the way you look at it) it is to compromise the EVMS.

One totalizer machine can conceivably count upto 8 (most likely 4) EVMS at as time. Taking a reference to the existing procedures, I would imagine that the totalizers will not be re-used. So, there will be one totalizer for every 8 EVMS, under the control of the DEO, and issued to the ROs along with the EVMS. EVMS will retain the counts, and totalizer will retain the total counts indefinitely

I would be happy with that setup.

Pranav · Post by **Pranav** » 16 Jul 2009 07:46

Jeez, you have not even provided a reference for a protocol that fully and provably verifies a black box, with no prior assumptions. Let alone making any arguments yourself. Do move on, I do not see what you have to contribute.

Dileep wrote:Pranav, you don't want to talk about yourself because there is nothing to talk. That kind of establishes the weight of your arguments isn't it?

Also, whoever understands how things work, got it already. I am not wasting any more time on that.

Moving on...

suryag · Post by **suryag** » 16 Jul 2009 07:55

The audit scheme need to verify the software code, and the functional operation of the box. There is no need to verify the transistors and interconnects within the microcontroller.

No, let us take the audit discussion further. The following would completely establish the integrity of the system.

1. Get the complete source code of the system. It is a simple embedded unit, with a single executable image. That makes it easy to verify.

2. Verify the code for proper function.

3. Verify the switches/lamps on the unit that provide the user interface.

4. Compile the source into binary.

5. Compare the production binary with the compiled binary, to see that the executable is exactly the same.

6. Program the OTP and verify the checksum via JTAG. This checksum should match on all units that use this binary.

7. Field audit the units for this checksum.

Pranav Dileep has already provided this

Pranav · Post by **Pranav** » 16 Jul 2009 08:24

suryag wrote: Pranav Dileep has already provided this

Suryag, all that's happening in the field is a checksum of of some binary code that is given out by a PROM chip. That is far from what is needed - a full, provable verification of all hardware and software, with no prior assumptions. You have to imagine an adversarial scenario - the manufacturer is trying to conceal, and you are trying to provably verify.

Muppalla · Post by **Muppalla** » 16 Jul 2009 08:24

Dileep and others:
In the election threads, I have provided similar docs to those who have doubts but the doubts are still persisting. In fact the EC has the procedure handbook pdf that talks about the randomization of the election officials. Then the new question that sprangs up is what about the randomizing software hack? The possibility of hacking of software is never ending. Questioning of everything is never ending. This goes on until EC publishes the audit reports of all the millions of EVMs. Then atleast we can concentrate on fudging the audit reports. My belief is you cannot do everything to that level and nothing comes out.

Atleast folks should desist comparing to other countires. The EVMs in other countries are OS based gizmos and hence it is sensible to not allow them. Indian EVMs are too simple.

Coming back to the topic: Any more info on what software is actually outsourced? What is the need to do such things? I don't see any voluminous program in these machines that requires outsourcing. This is where I am still confused. Is it just a rumor?

Rahul Mehta · Post by **Rahul Mehta** » 16 Jul 2009 10:12

If totalizer is used, one would need no more than 5000-10000 Totalizer, about 1-2 per Assembly seat. Replacing 10000 machines with rigged ones and re-replacing them next week is possible with mere 20 trained field agents and by bribing out top 10 EC officials. Child's play for CIA, Rajmata, MMS.

If totalizers were used, you pro-EVM people have no case. . Lets wait to see how many seats used Totalizers and how of them favored pro-MNC candidates from BJP or Congress. Also, if CIA is rigging totalizers, they would favor Congress as well as they would favor pro-US guys in BJP like Yashwant, Jashwant etc.

----

Dileep wrote: The EVMS are stored by the District Election Officers at the district headquarters. That throws out 90% of your arguments. What remains?

The link you showed only says that "EC desires that EVMs be stored at district head quarter". It does not say where they were actually stored before polling day. Also, the Gandhinagar RO on April-15th had mentioned in a meeting that he is yet to receive the EVMs, while polling was due on April-30th.

If EVMs are stored in District Headquarters, then that reduces CIA's ability to replace EVMs to those areas where CM and Collector agree. So it is possible in states in AP, TN etc where CMs and too many DCs are now Missionary/MNC agents and not possible in Gujarat where CMs and many District Officials are still anti-Missionary. So that reduces "rigging by replacement" to some 50-100 seats, but does not make it zero.

-----

Dileep, Tanaji, EVM-lovers, anti-paper, anti-RM-elements,

I must have asked you this question 5 times, and you refuse to answer. Can you give me estimate on how many people you need in field and how many middle/senior officials you need to bribe to replace paper ballots after polling and before counting?

I gave my estimate for EVM riggings. Why do you refuse to give estimates for paper ballot rigging? The mechanism which requires more field agents and requires more co-operation from middle/top officials is the one which is more difficult for CIA/anyone and is the one we should prefer. Your repeated refusal to give estimates on number of people needed to replace paper ballots and boxes between polling and counting date makes me believe that you really dont have stomach for comparing EVM-rigging with paper-rigging. Am I right?

So far, the debate has been really one-sided. I gave listing of code by which CIA can replace say 700,000 EVMs with Rs 500 cr, 100 field agents and support of top 20 EC officials. You have been busy with finding flaws in my procedure code. But I see no effort from YOU guys in giving procedure code by how YOU will replace paper ballots and boxes with fake ballots.

Dileep · Post by **Dileep** » 16 Jul 2009 10:37

Pranav wrote:
suryag wrote: Pranav Dileep has already provided this
Suryag, all that's happening in the field is a checksum of of some binary code that is given out by a PROM chip. That is far from what is needed - a full, provable verification of all hardware and software, with no prior assumptions. You have to imagine an adversarial scenario - the manufacturer is trying to conceal, and you are trying to provably verify.

Yea, Yea, the almighty CIA would influence the microcontroller manufacturer to change the JTAG module in the chip, to send the "correct" bits out instead of what stored in the OTP area of the chip. Is that technically possible is no question. It is the mighty CIA right?

Then you would go inspect the sand from which the silicon comes. You are no different from the stupid ISO auditor I encountered some years ago. The difference here is, there I HAD to satisfy him by doing some equally stupid thing. Here I don't have to.

Pranav, what you allude is STUPID, with a capital S. If you still don't get it, read up on JTAG, cyclic redundancy check, Security Hash algorithms etc.

The brighter part is, I started respecting Rahul Mehta. At least he is not promoting stupidity here, only outrageousness.

Dileep · Post by **Dileep** » 16 Jul 2009 10:48

The link you showed only says that "EC desires that EVMs be stored at district head quarter". It does not say where they were actually stored before polling day. Also, the Gandhinagar RO on April-15th had mentioned in a meeting that he is yet to receive the EVMs, while polling was due on April-30th.

Please read the WHOLE paragraph.

As a general policy, the Commission desires that all EVMs available within a district shall be stored at the district headquarters under the direct control of the District Election Officer. It may be possible that due to want of storage space the EVMs may be stored in a decentralized manner in different locations. Even in such cases, for the purpose of first level checking and randomization procedure, all EVMs available in the district shall be brought to the district headquarters under proper escort.

It means, the machines could be stored further decentralized within the district, probably under the sub collectors. All of them must be brought to the headquarters for randomizing.

Anyway, you have contacts in your locality. Go find out WHERE EXACTLY the machines WERE stored before the election, and where the ARE stored now, at least in your own district. In your own style, provide conclusive proof.

Then we will talk about it.

I don't think the totalizers will be re-used, so it will be one totalizer per 8 EVMs. We will know when the procedure comes out ultimately. My take is that the totalizers will be used on one set of 8 EVMS, and kept with the counts intact for record.

If that is the case, it comes to the same level of security as EVM.

Pranav · Post by **Pranav** » 16 Jul 2009 10:53

Dileep wrote: Yea, Yea, the almighty CIA would influence the microcontroller manufacturer to change the JTAG module in the chip ...

Here is someone who believes that intelligence agencies would never, ever influence chip manufacturers.

Dileep · Post by **Dileep** » 16 Jul 2009 10:58

So far, the debate has been really one-sided. I gave listing of code by which CIA can replace say 700,000 EVMs with Rs 500 cr, 100 field agents and support of top 20 EC officials. You have been busy with finding flaws in my procedure code. But I see no effort from YOU guys in giving procedure code by how YOU will replace paper ballots and boxes with fake ballots.

Why should we do that? We are not proposing to rig ballot elections!!

Serioudly, if the difficulty level in replacing ballot boxes is twice as replacing an EVM, obtaining a fake ballot paper is 1/1000,0000 less difficult than getting a fake EVM. You want code, here is the code:

1. CIA prints ballot papers.
2. CIA gets specimen signature of the presiding officer from the CEO office.
3. CIA use a robotic machine to put signature on the ballot, put the stamps, and vote stamp.
4. CIA fabricates ballot boxes to the same spec.
5. CIA pre-fills the ballot boxes with 95% of expected polling number.
6. After polls close, CIA gets the polling numbers and puts the required number of ballots in the box.
7. CIA seals the boxes.
8. CIA replaces the sealed boxes en-route or at the wearhouse.

Looks stupid? HAve objections? Your scheme looks equally stupid to me, and I have valid objections too.

Dileep · Post by **Dileep** » 16 Jul 2009 11:02

Pranav wrote:
Dileep wrote: Yea, Yea, the almighty CIA would influence the microcontroller manufacturer to change the JTAG module in the chip ...
Here is someone who believes that intelligence agencies would never, ever influence chip manufacturers.

Here is someone who hangs onto the tail-hair of a lost argument.

Can you explain how a chip can send out data that is unknown at the time of manufacture? Or do you fall to the same kind of people who aren't constrained by the laws of time, gravity, conservation of energy etc?

Raju · Post by **Raju** » 16 Jul 2009 11:05

those arguing that totalizers were not used in these elections at all please back up with some link or article.

Because as per one article they were not used in Gujarat for some reason. And that's it, and to be presumed that they were used everywhere else.

Dileep · Post by **Dileep** » 16 Jul 2009 11:12

'The EC has considered the proposal to use totalizer machine during the counting of votes and steps have also been taken by the EC to make appropriate amendments in election laws to pave the way for its use. However, I don't think we are going to use it this time,'' the state's Additional Chief Electoral Officer, T Natarajan told UNI here today.

The law is not changed, hence totalizer can't be used. Read the ECI website. It is only a proposal.

Read the full article. There is nothing Gujarat about it. Totalizers were tested, and some DDM asked if they will be used in this election. Gujarat happened to be on the last batch of elections. EC said no it is not going to be used.

There was no proposal to use it, and there was no "no use" in Gujrat. The whole news is made out of context.

If anyone want, file a PIL, or talk to the collectors to prove the use of toalizers.

Rahul Mehta · Post by **Rahul Mehta** » 16 Jul 2009 11:32

Dileep wrote:1. CIA prints ballot papers.
2. CIA gets specimen signature of the presiding officer from the CEO office.
3. CIA use a robotic machine to put signature on the ballot, put the stamps, and vote stamp.
4. CIA fabricates ballot boxes to the same spec.
5. CIA pre-fills the ballot boxes with 95% of expected polling number.
6. After polls close, CIA gets the polling numbers and puts the required number of ballots in the box.
7. CIA seals the boxes.
8. CIA replaces the sealed boxes en-route or at the wearhouse.

1. How many field agents will be needed for step-8 IYO?

2. How many constables, Tahasildaar level officers need to be bribed in step-8 IYO?

3. And how will your machine robot in step-3 know the color of the pen returning officer was using?

---

4. And you claim that CIA can print 70 cr ballots within 30 days, can make 700,000 boxes in 30 days. But you claim that CIA cant make 700,000 fake EVMs off shore in years of time?

5. Also, are you printing ballots inside India or outside India? Are you making boxes inside India or outside India? And your field agents will be taking fake boxes stuffed with fake ballots to 30,000 rooms all over India? (One room has some 300 boxes and there are 700,000 boxes in India). Is that YOUR plan? Is that what you call a PLAN?

6. Pls mention which operation is done off-shore and inside India. The off-shore vs inside-India is key point. If 70 cr fake ballots are printed in India, your plan is non starter. And if 70 cr ballots are printed outside India, you cant get them inside India within 30 days. Ditto with boxes. Where as 700,000 Control Units take only 1/100th of the space of 70 cr paper ballots plus 700,000 boxes and bring them inside India is easy, given that we have so many naxal infested areas where policemen dont even dare to go.

---

So when it comes to replacing ballot boxes, your CIA is mighty powerful. And when it is comes to rigging EVMs, CIA is not even 1/10th as strong.

Pls keep a level field. You have assumed that CIA cant make 700,000 boxes in US. You have assumed that CIA does not have field agents in India to replace 20000- 35000 cartons, each carton having 20-50 EVMs. You assume that CIA cant rig 5000 to 10000 Totalizers. And now you claim that CIA is strong enough to sneak in 70 cr fake ballots, 700,000 fake boxes, get them stamped, and get them replaced at 20,000 assorted rooms located in 500 places all over India !! You claim that 20 top CEC officials will not sell out. And now you claim that 10,000 PIs and head constables guarding 30,000 rooms will sell out !!

Pls keep a level field. Pls provide detals on how CIA or anyone will rig paper ballot

----

The point is : paper ballots are 10-1000 times more difficult to rig. And instead of accepting this TRIVIAL fact, EVM-lovers are giving hare brained schemes that one can make crores of fake ballot papers and lakhs of fake boxes in mere 2-4 weeks !! There is a limit of arguing for sake of arguing.
.

amit · Post by **amit** » 16 Jul 2009 11:40

Dileep wrote:
Pranav wrote: Or do you fall to the same kind of people who aren't constrained by the laws of time, gravity, conservation of energy etc?

Dileep,

The other day I was watching a TV programme which gave a name for this particular ability. You had this guy (with the ability) boasting: "I can bend time onto space..."!

You don't need totalisers, CIA, Evangelists (in Hyderabad) ECIL etc if you have that ability.

And yes the TV serial's name: Heroes.

We seem to have quite a few of them (Heroes) on this thread who refuse to be cowed down by reason, logic and facts. More power to them and all conspiracy theories!

Talking about conspiracy theories, here's mine: Any entity (CIA?) which has the power and ability to control the thousands of variables which it would need to in order to rig the elections via EVMs would be powerful enough to control whichever party comes to power at the Centre. Then why go through all the effort, cost and trouble of rigging in the first place? And why go through such a torturous debate, we're doomed in any case?

Tailpeice: I'm no techie and the word totaliser sort of conjurers up a vision of some sooper doopar laser gun manufactured by aliens. Is it a wonder then that the totaliser angle is championed on BRF by the one man who believes aliens control our destiny!

Dileep · Post by **Dileep** » 16 Jul 2009 11:42

RM, I didn't say it is doable. I say both are equally ridiculous. You start with the false axiom that "CIA want to rig the Indian Election by direct action" and asking for ways to do it. I say direct action is impossible in either case, so CIA doesn't take that route.

They are, despite all the screwups we hear, not stupid to try it.

Raju · Post by **Raju** » 16 Jul 2009 11:46

totalizers may not have been used in Maharashtra also. Because Shiv Sena candidate Mohan Rawle was seen mentioning that he got only 5 votes from Shivaji Park booth EVM. And info is not possible if totalizers were used to mix up EVM results and count them together.

And Shiv Sena considers Shivaji Park as its base.

niran · Post by **niran** » 16 Jul 2009 11:54

Gentle Gurus, let me if i may,Sir Rahul Mehtaji what Dileep means is that both
is absurd, have as much a chance of happening as much of you spitting on President
Obama face.

Pranav · Post by **Pranav** » 16 Jul 2009 12:00

Are you still beating the "candidates unknown at the time of manufacture" drum? That has been addressed long back.

Dileep wrote: Can you explain how a chip can send out data that is unknown at the time of manufacture? Or do you fall to the same kind of people who aren't constrained by the laws of time, gravity, conservation of energy etc?

Dileep · Post by **Dileep** » 16 Jul 2009 12:29

I don't have time to deal with stupidity. You have no clue how an embedded system work, You have no idea how the executable code burned into a microcontroller can be verified to be the same as a master copy. You pretend to not understand, and demand "mathematical proof", as if you are an expert in cryptography. What can I say to you?

I say it again, I have outlined how the system could be audited here. Those who have the requisite technical background did understand it. I can't help you if you don't. I have no incentive to make you either.

Now, as gently as I can put it...

SHOO.....

AjayKK · Post by **AjayKK** » 16 Jul 2009 12:31

.

Totaliser: Posted without comment.

BANGALORE: March 11, 2009

The totaliser machines which can maintain confidentiality of the votes to be cast through EVMs, will not be used in the state {Karnataka} during the upcoming Lok Sabha polls. The Central Election Commission has been mulling over the idea of using totaliser machines only in sensitive polling stations in order to avoid the possibility of any intimidation and victimisation of electors. In order to not disclose the voting pattern in a particular booth, the totaliser system is used, where the votes will be mixed, before counting.

http://www.expressbuzz.com/edition/prin ... O3Ljqb5AQ=

Chandigarh May 16, 2009

With the general elections scheduled in the fifth and last phase, Chandigarh has had to wait for a long time for its turn to go to polls. The counting process will, however, compensate this, with the verdict for this Parliamentary constituency set to come in two hours flat.

In all, 3.4 lakh people—around 65 per cent of the total voting population—cast their votes on Wednesday to elect their representative from Chandigarh to the 15th Lok Sabha. As many as 14 candidates were in the fray. Their fate will be sealed on Saturday.

The four counting centres are Government College for Boys, Sector 11; Chandigarh College of Engineering and Technology, Sector 26; Government College for Boys, Sector 46; Government College for Girls, Sector 42. There are 92, 113, 98 and 120 EVMs stored, respectively, in these centres.

IAS officers Niharika Rai, Prerna Puri and R S Verma, and PCS officer P S Shergill are the officials in charge.

Results from all four counting centres will be sent to the returning officer-cum-deputy commissioner, who will further compute all figures and declare the final result from his office in Sector 17.

There will be 14 counting tables at each counting centre. As many as 16 representatives—one at each table and two additional—from each political party will be allowed to witness the process.

Each round of counting will take up 14 EVMs — the number cannot exceed 14 under the rules set by the Election Commission of India. Seating arrangements for counting agents will be made in such a way that they can clearly watch the machine and its display.

The ‘result’ button will be pressed and the machine will display the total number of votes polled at a particular polling station. This will be followed by the number of votes polled in favour of every candidate in a serial order. Then comes the role of ‘Totaliser’. The device can be connected with several control units (EVMs) at a time. It will indicate the total number of votes polled in each polling station where these EVMs had been used. It will also display the grand total of the votes polled in those booths.

http://www.expressindia.com/latest-news ... te/460471/

Dileep · Post by **Dileep** » 16 Jul 2009 12:42

I think the DDM mixed things up. They picked up the totalizer experiment and thought it is a done deal. ECI doesn't work like that. Laws need to be changed to do even simple things.

When EVMS were first used loong time ago, the court annulled the election based on some law point.

Pranav · Post by **Pranav** » 16 Jul 2009 12:52

Dileep wrote:I don't have time to deal with stupidity. You have no clue how an embedded system work, You have no idea how the executable code burned into a microcontroller can be verified to be the same as a master copy. You pretend to not understand, and demand "mathematical proof", as if you are an expert in cryptography. What can I say to you?

I say it again, I have outlined how the system could be audited here. Those who have the requisite technical background did understand it. I can't help you if you don't. I have no incentive to make you either.

Now, as gently as I can put it...

SHOO.....

You have outlined your checksum proposal and it is thoroughly inadequate. All the blabbering and name-calling will not help you.

Dileep · Post by **Dileep** » 16 Jul 2009 13:01

Pranav, prove that the checksum proposal is inadequate. Specify the exact vulnerability.
YOU are the one who babbles.

niran · Post by **niran** » 16 Jul 2009 13:17

IMVVHO it is IB4TL here. before it develops into fisticuff and multiple bannishment.

Sanku · Post by **Sanku** » 16 Jul 2009 13:18

Pranav wrote: You have outlined your checksum proposal and it is thoroughly inadequate. All the blabbering and name-calling will not help you.

As an practicing Engineer in this field I can vouch for the fact what Dileep is saying is correct.
The system has two parts
1) Hardware
2) Software

The hardware itself can be sufficiently tested in the method that Dileep has outlined. Yes there are paranoid methods of checking the hardware (such as under electron microscope and structural testing) however they are manufacturing tests for manufacturing flaws. Quite different from testing the functional correctness of the chip as per a pre defined spec.

So we can assume that the testing itself if enough to validate the spec.

------

The software is another part and the doubts there are at least valid if through discussion we can assure ourselves that they are far fetched.

-----

Of course it is entirely "possible" that the Chip itself has a Trojan functionality which is not documented and is used by some Trojan code in the software and is enabled by some Trojans.

However given the varied nature and unpredictability of candidate and machine distribution etc (all discussed before) using any such feature in a planned manner would require lot of people in the know at various levels and a massive impossible to hide trail.

I would not say that the above scenario is probable.

-----

Pranav, I am really sad to see you getting into a solid disagreement with Dileep. You are both good guys on forum. Please chill.