How 100,000 Evms Can Be Tampered by Just 10-12 People at Top
How 100,000 Evms Can Be Tampered by Just 10-12 People at Top
How 100,000 Evms Can Be Tampered by Just 10-12 People at Top
This is not an allegation (as of yet) This article is NOT an allegation. It is to only show that EVMs can be tempered at super-industrial scale by few people at top. Essentially, I am making allegation against EVMs, that EVMs are temperable in an industrial scale to favor some party such Congress etc and that too with just 10 people at top. I am not making allegations against any party or EC or BEL in this article. Some 100,000 new EVMs were made in Dec-2008. I am only proposing one way by which people at top may rig 100,000 EVMs with mere 10-12 men. As of now, I am not claiming they did it. Because to claim that X did action-A, I need to first convince people that action-A is indeed possible and do-able. So for now, I only want to show that tempering100,000 EVMs is possible with few people. Once I have shown that is do-able, I may or may not show that they were in position to implement the tempering mechanism I suggested. In this article, I have used following fictitious names to show the process (resemblances to any real names is an unintended coincidence) CECX : A fictitious head of a fictitious body called ECX, in charge of elections CongressX : A fictitious party interested in tempering EVMs to favor its candidates. SoniaX : The fictitious head of CongressX BELX Chiefs : The persons in charge of manufacturing EVMs in fictitious body named as BELX This is a summary document and many details have been skipped as I wanted to keep it no more than 8 pages.
Remote control EVM machine EVMs use a custom imported chip which is made in USA or Japan. BELX and ECIX has no disclosed the name of the company which makes the chip and also not disclosed the owners of the company. So for now, we do not know whether the company is owned by CIA or US Military or SoniaX or Rahul Mehta (author of this article). BELX claims that EVMs cannot receive radio signals. The EVMs are not given to any candidates of public for examination. So it is not possible for citizens to detect if the EVM dont have ability to receive radio signals. But it is possible that some 10% to 20% EVMs are made in a way that they can receive and process radio signals. The radio technology has so advanced that a 2 mm by 2 mm chip can receive radio signals from a van standing 100 meters away (such small passive chip cannot send to a distance, but can certainly receive). If these EVMs use such chip, then candidate number can be communicated to EVM from a van anytime before counting and EVM can change the votes inside it. And even the van driver need not suspect. It could be some van used by media persons etc who are merely asked go near booths or collectors warehouses to give coverage, and the equipment in van can send the candidate number to the EVM. So even the driver may not know that his van is being used to send favorite candidate number to the EVMs. I assume that at least 20% of EVMs are radio enabled, unless proven otherwise. But for now, lets say EVMs dont have radio waves. Then can they be made pro-CongressX? 3 Why pre-programmed bias in EVMs was thought to be useless so far? We all know that a programmer can indeed write code in EVM, that favors a particular serial number. Pls note - a particular serial number only. And if top persons in BELX agree to put such tempered code in EVM, then there is no technological barrier to stop them. But till date, everyone thinks that no one can write a code that would favor CongressX, as the serial number of CongressX candidate is undecided when the coder is writing the EVM code and
undecided even when the EVM is being shipped to the District from BELX. eg say a coder and BELX Chiefs decide to favor CongressX. Now the EVM is to be shipped out in Jan-2009. CongressX candidate will get the serial number in Apr-2009. Then which candidate number should the program be coded to give more votes? The Congress's candidate number will vary from seat to seat. So if the programmer programs EVM to give more vote to candidate number 1, then only in some 20% cases only CongressX will be no. 1 . And so in 20% cases, it would benefit Congress and in 80% cases it would hurt Congress. Some anti-EVM people have hypothesized that some passkeys can be inputted in EVMs to communicate the favorite candidate number. But then inputting this key combination in 100,000 of EVMs need 100,000 men to be sent at the booth and is a difficult task. So this way was ruled out due to logistics. So till now, it was stated by "experts" and accepted by gullible that "BELX cannot make an EVM that would favor CongressX". 4 BELX can make pre-programmed EVMs that favors CongressX But I have shown that a "pro-CongressX" EVMs be made !! How can EVMs be programmed to favor CongressX? Well, pls read this article and if you have any questions, pls call me at 98251-27780 . 5 A brief description of election related dates Pls note that EVMs can support only 64 candidates, above which EC will have to use paper ballots. Now following dates play important role in "how EVMs can be made to favor CongressX". 1. Apr-2 : Say opening date to fill form. 2. Apr-8 : Last date to fill form 3. Apr-10 : Candidates will get accepted or disqualified by RO (Returning Officer, mostly District Collector). 4. Apr-11 : Candidates of CongressX, other recognized parties will get to know their serial number on EVM (*) 5. Apr-13, 3 pm : Last day , time to withdraw 6. Apr-13, 5 pm : Total number of candidates will be known at Apr-13, 5 pm. Pls note this : total number of candidates plays important role in how EVMs can be made to favor CongressX [* - See section-6 why recognized party candidates can accurately guess their candidate number] 6 The recognized parties' candidates are mostly in number 1-5 EC (Election Commission) divides candidates in a constituency into 3 blocks : National or State Recognized Parties, Unrecognized Registered Parties and Independents. In each block, the candidates are alphabetically ordered by their names in the local language. There are 7 Nationally recognized parties and different states have0-4 State recognized parties. So if all recognized parties put their candidates, the number of candidates in first block will go to 7-11. But the reality is that all recognized parties don t put candidates in every seat. eg CPM has not put candidates in many seats since ages. In general, having 5 recognized party on one seat is a rare event. And so a recognized Party is will get a number between 1-5 with 99% guarantee in general. And as a reality check, one can look at 2009 election. All Congress candidates got serial number between 1-5 on the EVM. 7 Step-One : Nature of the tempered code to put in EVMs Say BELX chiefs made 5 different types of tempered EVMs.
Type-1 favors candidate number = (nCandidates + 1) mod 5 + 1 (what is mod?) Type-2 favors candidate number = (nCandidates + 2) mod 5 + 1 Type-3 favors candidate number = (nCandidates + 3) mod 5 + 1 Type-4 favors candidate number = (nCandidates + 4) mod 5 + 1 Type-5 favors candidate number = (nCandidates + 5) mod 5 + 1
Now what is "mod"? It is remainder left after division, i.e. ( 16 mod 5 ) means the number you are left with when you divide 16 by 5 which 1. So
(1 mod 5 ) = 1 , (2 mod 5 ) = 2 , (3 mod 5 ) = 3 , (4 mod 5 ) = 4 , (5 mod 5 ) = 0 , (6 mod 5 ) = 1 , (7 mod 5 ) = 2 .
The *mod will keep repeating. There are only 5 values (n mod 5) can have. 2
The tempered EVM use the above formula for selecting favorite candidate. Eg1 say one constituency has say 200 tempered EVMs of type-3. Now say that constituency has say if nCandidates = 19. Then the tempered EVMs there will favor candidate number (19 + 2) mod 5 + 1 = 21 mod 5 + 1 = 1 + 1 = 2 . i.e. it will favor candidate number 2. Eg2. say one constituency has say 200 tempered EVMs of type-4. Now say that constituency has say if nCandidates = 11. Then the tempered EVMs there EVM will favor candidate number (11 + 4) mod 5 + 1 = 15 mod 5 + 1 = 0 + 1 = 1 . i.e. it will favor candidate number 1. So BELX made file types of tempered EVMs. A tempered EVM decides candidate to favor based on total number of candidates based on the following table Table tempered EVMs follow (Pls read column headings carefully) Row #
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Type-5 Type-4 Type-3 Type-2 Type-1
Type
6 , 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 , 57 or 62 8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 , 58 or 63 4 , 9 , 14 , 19 , 24 , 29 , 34 , 39 , 44 , 49 , 54 ...64 5, 10 , 15 , 20 , 25 , 30 , 35 , 40 , 45 , 50 , 55 , 60 6 , 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 , 57 or 62
2 , 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 ... or 62 candidates 3 , 8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 ... or 63 candidates 4 , 9 , 14 , 19 , 24 , 29 , 34 , 39 , 44 , 49 , 54 ... or 64 candidates 5 , 10 , 15 , 20 , 25 , 30 , 35 , 40 , 45 , 50 , 55 , 60 6 , 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 3 , 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , . or 62 candidates candidates candidates candidates candidates candidates
8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 , 58 or 63 4 , 9 , 14 , 19 , 24 , 29 , 34 , 39 , 44 , 49 , ...or 64 5 , 10 , 15 , 20 , 25 , 30 , 35 , 40 , 45 , 50 , 55 , 60
I would politely request the reader to read further AFTER answering following questions : 1. If EVM is of Type-3 and number of candidates are 12, whom will that EVM favor? (Hint : see row#12) 2. If EVM is of Type-2 and number of candidates are 24, whom will that EVM favor? (Hint : see row#9) 3. If EVM is of Type-1 and number of candidates are 19, whom will that EVM favor? 3
4. 5. 6. 7.
If EVM is of Type-4 and number of candidates are 21, whom will that EVM favor? If EVM is of Type-5 and number of candidates are 10, whom will that EVM favor? Will EVM of Type-1 always favor candidate #1? Will EVM of Type-2 always favor candidate #3?
Lets make converse of the above table. The following converse table will be used in Step-3 by SoniaX. I request the reader to reach each of the 30 lines in the following converse table. Converse table to be used by SoniaX in step-3 (Pls read column headings carefully) serial # of candidate SoniaX wants to favor
1 2 Type-1 3 4 5 1 2 Type-2 3 4 5 1 2 Type-3 3 4 5 1 2 Type-4 3 4 5 1 2 Type-5 3 4 5
Row #
Type
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
6, 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 , 57 or 62 8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 , 58 or 63
1, 6 , 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 candidates 2, 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 , 57 or 62 candidates 3, 8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 , 58 or 63 candidates 4, 9 , 14 , 19 , 24 , 29 , 34 , 39 , 44 , 49 , 54 , 59 or 64 candidates 5, 10 , 15 , 20 , 25 , 30 , 35 , 40 , 45 , 50 , 55 or 60 5, 10 , 15 , 20 , 25 , 30 , 35 , 40 , 45 , 50 , 55 , or 60 6 , 11 , 16 , 21 , 26 , 31 , 36 , 41 , 46 , 51 , 56 or 61 7 , 12 , 17 , 22 , 27 , 32 , 37 , 42 , 47 , 52 , 57 or 62 8 , 13 , 18 , 23 , 28 , 33 , 38 , 43 , 48 , 53 , 58 or 63 9 , 14 , 19 , 24 , 29 , 34 , 39 , 44 , 49 , 54 , 59 or 64 candidates candidates candidates candidates candidates candidates
I would politely request the reader to read further AFTER answering following questions : 8. If EVM is of Type-3 and your serial number is 5, how many candidates should that seat have to favor you ? 9. If EVM is of Type-1 and your serial number is 1, how many candidates should that seat have to favor you ?
10. If EVM is of Type-2 and your serial number is 3, is its must that EVM will favor you only if you have 5 candidates? Or will any other total number of candidates would do? 11. If EVM is of Type-5 and your serial number is 2, is its must that EVM will favor you only if you have 11 candidates? Or will any other total number of candidates would do? 8 Step-Two : Ensuring each seat has same type of tempered EVMs CECX has to ensure that 200 tempered EVMs go ton each of the 500 Constituency, in away that Type of each EVM is a constituency is same. How can he ensure that? He and BELX chiefs can ensure this as follows :
BELX chiefs send 1st shipment of 20000 EVMs between dates say Jan-01 and Jan-06 BELX chiefs send 2nd shipment of 20000 EVMs between dates say Jan-07 and Jan-13 BELX chiefs send 3rd shipment of 20000 EVMs between dates say Jan-14 and Jan-20 BELX chiefs send 4th shipment of 20000 EVMs between dates say Jan-21 and Jan-27 BELX chiefs send 5th shipment of 20000 EVMs between dates say Jan-27 and Feb-07
CECX asks staff to send 200 EVMs from 1st shipment to each Constituency-001 to 100 CECX asks staff to send 200 EVMs from 2nd shipment to each Constituency-101 to 200 CECX asks staff to send 200 EVMs from 3rd shipment to each Constituency-201 to 300 CECX asks staff to send 200 EVMs from 4th shipment to each Constituency-301 to 400 CECX asks staff to send 200 EVMs from 5th shipment to each Constituency-501 to 500
So each seat has 200 tempered EVMs of same type. And only CECX, SoniaX know the type for each Constituency. 9 Step-Three : Adjusting the total number of candidates I request the reader to look at the Converse Table given in Section-7. Now take some Constituency such as Gandhinagar as example. Say in Mar-2009, Gandhinagar Constituency got 200 tempered EVMs of type-2. Now on Apr-9, CongressX knew that its candidature number, say it was number 2. How many candidates should be in race so that EVMs of type-3 favor the CongressX candidate? Pls look at the converse table row#7. The answer is : if SoniaX can ensure that Gandhinagar has exactly 4 , 9 , 19 , 24 , 29 or ... 64 candidates in race, then all EVMs which are of type-2 will favor the no. 2 which is CongressX candidate. But how can SoniaX ensure that number of candidates in race in Gandhinagar are 4 , 9 , 14 , 19 , 24 , 29 or ... 64 and not some other number? She can tell the Congress candidate that astrologer has asked that the number of candidates must be 4 , 9 , 14 , 19 or ... 64. And CongressX candidate has to be told to put 4 spurious independents and have 0-4 withdraw the form at last moment, so that nCandidates 4 , 9 , 14 , 19 , 24 , 29 , or ... , 64. Let me explain in more detail. Say it is Apr-13, 3pm. The collector will generally close his office door at 3pm and ask a peon not to let any more persons come inside. Say the CongressX candidate has 4 spurious independent candidates named as say A, B C and D. They are in collector's room for withdrawal, along with 1-10 other people who want to withdraw. Now all the CongressX candidate has to manage is that these 4 people are at the end of the queue and not in-between. Now say all other candidates who had entered the room have either withdrawn or refused to withdraw (in which case they have left the room) and at the end only A, B C and D are left. Number of candidates, before A, B, C, D are called
20 28 32
Decision
20 1 = 19 28 4 = 24 32 3 = 29
31 A, B, withdraw ; C, D stay
31 2 = 29
IOW, using 0-4 candidates at the end of the withdrawal queue on Apr-13 , he can ensure that number of candidates in the race are 4, 9, 14, 19, 24. 29, or ... 64. This is not difficult as RO or anyone is not even suspecting that EVMs use nCandidates as any parameter to rig elections. And in states where Congress-X has CM, SoniaX can ask CM to ask RO to put the spurious independents at the end of the queue so that they have last word n nCandidates. In 5% cases, say CongressX fails to ensure that candidate number is what they need. Then they lose 5% EVMs to opponents. But in 95% cases, the tempered EVMs will support CongressX candidate. The RO (District Collector) is not a person in conspiracy. And the Congress X candidate too need not be aware of this conspiracy, for all he knows, he is doing what SoniaX's astrologer asked him to do. Or SoniaX can simply not use the candidate or use some other professional persons to adjust this count. All in all, ensuring that number of candidates in race is (4, 9 , 14 , 19 , 24 , 29 , 34 , 39 , ... or 64) is not a difficult task and is trivially easy where CM is also from CongressX 10 Where would this tempering be used? There are 543 Loksabha Constituencies. Step-One will work well as it is all in a few controlled factories and there is no field operation. And Step-2 will work as dispatching decisions are very centralized. CongressX will not send tempered EVMs to 70-80 sure seats. Say tempered EVMs are sent to 460 seats. And out of these 460 seats, step-3 works out in 410 cases and fails in 50 cases. Then CongressX or ally candidate will get all 200 EVMs in those 460 seats which will favor CongressX. If each EVM gives him extra 300 votes, he gets 60,000 extra votes. Enough to make increase final tally by 100-150 seats. Now please note that I am not making allegation that this indeed happened. I am saying that this can happen in future. And this is one way to rig EVMs, there can be 100s more. 11 Compact description of rigging A compact description of tempering is as follow 1. There are 5 types of tempered EVM BELX made. EVM-k favors candidate whose roll number is
= (nCandidates + k) mod 5 + 1 , k is between 1 to 5
2. CECX ensures that all seats have 200 rigged EVMs of same type and knows which seat has which type. 3. CongressX will know its roll number 2-3 days before last day for candidates to withdraw. Say Congress's roll number is p and it will be between 1 and 5. 4. SoniaX has to only ensure that, that seat has nCandidates = 5m + (p - 1) - k , m is any integer. 5. So in each seat, CongressX candidate or some agent of SoniaX has to put 4 dummy candidates and make 0-4 of the withdraw after everyone else has withdrawn, so that nCandidates is 5m + (p - 1) k 6. If nCandidates = (5m + (p-1) - k), then the rigged EVMs in that seat will benefit candidate number
= (nCandidates + k) mod 5 + 1 = ((5m + (p-1) - k) + k ) mod 5 + 1 = ( 5m + (p-1)) mod 5 + 1 = p - 1 + 1 = p = the roll number of the CongressX candidate.
12 Why such tempered EVMs simply cannot be detected? 1. A smart person would do tempering only if there are over 200 votes and time EVM was left on was above 7 hours. Hence mock polls and testing would never detect this rigging 2. A smart coder will do tempering only once in life time and never again. So if the same EVMs are now tested, all of them will be free of rigging. 3. The code inside ROM cannot be read after lockbit is set.. I dont know if BELX sets lockbits after writing program into EVM, but it is common practice in industry to set the lockbit. So if BELX has set lockbit, there is no way to know if the code was tempered. 6
4. One can read the hash of code. But if the manufacturer of the chip was asked to make a chip that would give a planted hash than real hash, then hash is not the hash of the code in EVM but some planted hash. Hence, there is no way to detect if EVMs used in Apr-2009 were tempered or not. 13 How to put such code in the microcode of the chip There are two place to put this tempered code in the chip. One is ROM and other is microcode. If this tempered code is in ROM, one may find out by opening the machine and scanning the ROM if lockbits are not set. If the lockbits are set, then ROM cant be read, only hash can be read. And if chip maker has agreed to make a chip that would give planted hash rather than real hash, then hash has no value. But if lockbits are not set, the code can be read. But one can also put this temperingin microcode. How? If the microcode writer knew the EVM's code, he can find out the memory locations in RAM that stores nCandidates and counts of each candidates. And based on nCandidates and its own type k, it calculates Favorite_Candidate_No from the table given in Section-6 (Step-One). So it can add votes to the favorite candidate and deduct from the rest. Also say there were one bit of on-chip flash memory which gets reset some 8 months after manufacture. And once that bit is reset, that part of microcode will never get executed and so the microcode will become will become "honest". This is technologically very much feasible. And microcode cant be read using any equipment. So now there is no technologically feasible way to know if such a given EVM is tempered or not. The EVM used a made in Japan chip. So if CongressX had asked that Japanese company to put tempered code in microcode of the chip, which implements the table I gave, then CongressX, BELX and CECX together could have tempered 100,000 EVMs in a way we will never ever detect. 14 Putting tempered code in EVM without co-operation of BELX The tempered code can be planted in EVM without co-operation from BELX chief and without their knowing it. How? The chip has been imported. EC says that chip is imported from Japan, but the company name is not disclosed. And we dont know who owns this company. For that matter, we dont even know whether SoniaX owns this company via Mauritius holdings. After all, if a company is making chip using which one can rig 100,000 booths in India, every entity like MNC, CIA etc would like to invest in that company. Once a chip is made, no equipment exists to know what exactly has gone in that chip. And scanning every chip before it goes into EVM is an expensive process. Now lets say the chip has been tempered as follows 1. Lets call actual program as A and tempered version of A as B. 2. Say B is already present in chip when shipped from Japan or wherever 3. If any program say C other than A is downloaded on the chip, it will execute C 4. If A is downloaded on the chip, it will execute tempered version of A i.e. B 5. If the chip, and chip is asked to give hash or verification signature, it would give signature of whatever program that is downloaded. Now this possible, if the chip maker has program A that BELX has written for EVM. Obtaining a copy of this program needs only one mole in BELX, one need not compromise the chiefs. IOW, it is possible to put the tempered program in chip right inside the foundry which makes the chip in Japan. And then no one in BELX will even know that chip already has the tempered version of the program. Can we trust the integrity of some Japanese or American or any company, whose name even we dont know? Whose ownership is not disclosed to public? I dont want to put such trust on a Japanese chipmaker, and I would let each citizen of India decide on his own. 15 Way out I propose camera in booth, a stamping machine with counter and built in delay of 20 seconds and paper ballots. The details are in http://rahulmehta.com/evm2.pdf .But that is proposal from just a common man, myself. The final decision has to be taken by the citizens of India. How can we know if they want EVMs to continue or not? So my demand is not remove the EVMs in first step, but to know what commons in India want. I have proposed a way by which EVMs can be tempered. But do commons believe that such a way is feasible for SoniaX, 7
CongressX, CIA or anyone? If commons believe that my theory is mere hoax, there is no reason to remove EVMs. And if commons think that my theory is plausible, then EVMs must be removed no matter what experts in ECI believe. So my only demand is to create a mechanism by which we can know what commons want. 16 Knowing if commons want EVMs or paper ballot I request the PM to sign a Govt Order whose draft is as follows, and I request ECI and all concerned citizens to make same request to PM :# 1 Officer Collector Procedure If a citizen submits an affidavit to the Collector and demands to be put on PMs website, the Collector will issue a serial number and put it on PMs website for a fee of Rs 20 per page. If a citizen comes with voter ID, and specifies Yes-No on an affidavit submitted in cluase-1, the Talati will enter his Yes-No on the PMs website with voter-ID and give a printed receipt for Rs 3 fee. The Talati will also allow citizen to change his Yes-No for Rs 3 fee. The fee will be Re 1 for BPL card holder The Yes-No count will not be a binding on PM, CMs, officers, judges etc
Talati
-----
The details of the above Govt Order are on http://rahulmehta.com/mrcm_demand_01.htm Now IMO, we should assume that citizens support EVMs, and those want paper ballots back have to prove otherwise. How can we anti-EVM pro-paper prove that over 40 cr citizen voters of India want paper ballots back? We can prove that (or fail and withdraw our demand) if and only if the PM agrees to issue the GO as above. If the PM issues above GO, and then I would submit an affidavit to demand cancellation of Section 61A of Peoples Representation Act and also demand to install camera in booth, stamping machine with 20 second delay and counter and paper ballots. If we fail to convince over 38 cr citizens to register YES on this affidavit, then there is no need to replace EVMs. But if citizens do not wish EVMs to continue, then above Govt Order can be use to prove that. 17 Finally, my proposals I want camera in booth, stamping machine in booth with 20 second delay and paper ballots. But I want that AFTER over 40 cr citizen voters have approved it. Now if PM agrees to sign the Govt Order I have proposed, then I will try to convince citizen voters to register YES on the demand to cancel EVMs. o If over 40 cr citizens register YES, I will start agitation to remove EVMs o If over 40 cr citizens register do not register YES, I will drop the demand to remove EVMs if PM refuses to let citizens register YES-NO on laws passed or affidavits of fellow citizens, then I will assume that PM is not interested in even knowing what we citizens want. In such case, I would appeal activists to become candidates so that number of candidates are more than 64 and EC is forced to use paper ballots. I request the concerned citizens to ask PM to sign the above GO so that citizens can register their decision on EVM issue and also many other issues. 8