Is pro-VVAT in the same category as anti-MMR?

williamgrogan

On the basis of its investigations and its review of expert reports, submissions received and other relevant information, the Commission has noted the following in relation to the chosen system:

the physical layout of the system is straightforward, contributing to ease of use by both voters and election officials;

the system eliminates many inadvertent voter errors as well as the need for subjective judgment by returning officers;

the system was deployed in pilot tests at previous elections and a referendum in Ireland;

a system designed and manufactured by the same suppliers is in use at elections in the Netherlands and Germany;

testing carried out by experts retained by the Commission on a significant sample of the voting machines deployed to returning officers confirms that the system can accurately and consistently record voter preferences;

testing of the counting software carried out by experts retained by the Commission using voting information from pilot tests during previous elections in Ireland confirms that it accurately counted the votes recorded at these elections;

parallel testing of the counting software program carried out by experts retained by the Commission using a large number of sample data sets and a similar counting programme developed for the Commission confirms that it can accurately count votes in most situations, including unusual or difficult electoral situations;

miniature end-to-end testing of the system carried out by experts retained by the Commission confirms that it can accurately record and count the votes in the context of multiple simultaneous elections;

That's a lot of confirmation!

........

The Commission wishes to emphasise that its conclusion is not based on any finding that the system will not work, but on the finding that it has not been proven at this time to the satisfaction of the Commission that it will work.

Is this not The Precautionary Principal at work?

What the commission said is that, "within the timeframe", they had not been able to satisfy themselves. What “timeframe” is required to be 100% certain, infinity?

Another important type of sentence is, “it is impossible for anyone to certify its accuracy”. Correct. This will always be the case and this is what is meant by over reliance on the Precautionary Principal.

They said that it was easier to say no than yes and even acknowledged this. I think this point is the most important in the report. It can be interpreted as, “we have studied it in detail and tested it, it looks like it will work, it works elsewhere but our gut feeling is to say no.”

Wakefield made the same point as do those that oppose Fluoridation of water, or the anti-GMO movement, we cannot be sure that it won’t cause a problem so do nothing. As a consequence we have an increase in Measles, less food and rotten teeth. Do nothing means no progress and the decision of the commission means at best delayed progress with the move to eVoting. As a consequence of the commission's decision another 2% of the electorate did not have their votes accurately counted in the recent elections.

I wonder if we set up a similar commission and asked them to confirm for a fact that evolution was how life got on earth, what would they say? By their own logic, they would have to say, “well it looks fairly certain but we have to conclude it’s still just a theory.”

I also detect a sense of annoyance that they had great difficulty in getting access to the source code and obviously, relations between the commission and the suppliers were very poor.

I did say that I would accept the report of the commission but I meant that if they said that they had concluded that it won’t work I would accept it. They didn't say that.

They said it would work but couldn’t be sure. How sure is needed?

This is the central point. It would also affect any similar decision, e.g. whether or not Ireland such switch from the environmentally destructive burning of fossil fuel to Nuclear Power. Even among those of you who support NP, could you say that it is 100% safe? No of course not.

Nothing is, and this is another similarity between the anti-eVoting lobby, Luddism and the general anti-technological movement. In the article I quoted above, the point is made that doing nothing, as the commission recommended, can be worse because you are not 100% sure than doing something. In this case 2% of the votes were not counted.

It is of enormous importance to the future of humanity that progress is not hindered by commissions who by their own words take the easy option and do nothing. It's often referred to as “a safe pair of hands”. Safe pairs of hands never achieved anything. I see this as a major problem and one that is getting worse.

PPS

They spelt computer program incorrectly, they spelt it programme.

PS

Just like the Space Shuttle, Cassini, now about to fire its engines and break into Saturn orbit had no end to end testing.

RainyDay

Hi William - I'm starting to get increasingly skeptical about your real commitment to sketicism. Isn't skepticism all about being based on evidence?

Your quotes from the CEV report are all very interesting, but none of them support your earlier comment of "The commission even said they believed that the system would work". This is a forward-looking statement, i.e. predicting a future outcome. The commission never said any such thing. All the statements you quote from the CEV are present or past tense, not future. This is a subtle but important distinction. For a skeptic, I'm really dissapointed that you choose to jump to conclusions which suit your argument, rather than looking at the actual evidence.

These quotes are selective in the extreme. You pulled the 1 page of positive comments from a 28 page report. The other 27 pages are generally negative, just to put it in context.

Another important type of sentence is, “it is impossible for anyone to certify its accuracy”. Correct. This will always be the case and this is what is meant by over reliance on the Precautionary Principal.

This is just ludicrous - Read the 1st half of the sentance which you conveniently omit;

As the software version proposed for use at the forthcoming elections is as yet unknown, it is impossible for anyone to certify its accuracy

The reason why it is impossible to certify its accuracy is not the Precautionary Principal or any other interesting philospical point. It's far more concrete than that. The reason why the CEV found it impossible to certify the accuracy of the system was because it didn't exist at the time of their report. The version of software to cope with the referendum along with the local/euro elections just didn't exist.

Please stick to the facts here.

williamgrogan

It’s amazing that like so many other people holding such views you spent a large part of your last email trying to insult me. I think that always says a lot. I have to be selective, there are 28 pages. I am quite happy to debate every last sentence in it. You ask me to stick to the facts and then say, “I'm starting to get increasingly skeptical about your real commitment to sketicism”. I brought up a perfectly valid line of argument and linked the Precautionary Principle to the recommendation of the commission and rather than admit this is valid you foam at the mouth.

Back to the point.

All software is amended, most systems continuously. I made 500+ changes to my own system in the last 5 years. If I was to do an end to end complete test of my software for every amendment then an amendment would cost at least 50 times as much to do and would never happen.

This is more or less what the report suggests as necessary and this is because of their acceptance of the Precautionary Principle.

The sentence you quoted, “As the software version proposed for use at the forthcoming elections is as yet unknown”, is meaningless in this context. If it meant anything I can also say that the software that any future client of mine purchases from next week on will be unknown.

When a change is made to a system only that part of the system is ever tested unless it’s a very low level change when random tests would be done throughout the system. Only someone who knows nothing about software would demand a total end to end test of a system after every change.

Testing

I think that many people, and it looks like this includes the commission, literally do not know what testing means or implies. If you test something properly and it passes the test that tells you something concrete about what is being tested. They said they did these tests and the software passed. There are issues related to probability here. The software passed the test. The fact that there will be more changes doesn’t mean they can’t say that it is unreliable just that tests about the area changed are necessary and that the increase in unreliability is very small. The man in the street if told that a particular test meant that there was a 99% probability that something was true would still “feel” that it could still go wrong BUT would incorrectly estimate this with a higher probability than is true. The reverse is where people do the lotto and overestimate their chances of winning, if they understood that doing the lotto every week since the last ice age would still leave them with only a 50/50 chance of winning they might stop.

As to this quote, it is impossible for anyone to certify its accuracy, there are computer scientists that believe this to be the case. That eVoting will never be acceptable and they do base this on a philosophical argument. This is Luddism. To refuse a technology because it cannot be made 100% safe is most definitely taking the Precautionary Principle to extremes and if followed generally, and remember many do want this followed to extremes (e.g. anti-vaccination, anti-GMO, anti-NP etc.), then no future progress can be made.

RainyDay

Hi William - There are no real insults in my last post - just statements of fact and a few of my opinions. I'm not foaming anywhere - just pointing out the lack of consistency in your arguements. Are you still claiming that the CEV stated that " they believed that the system would work"?

I'm amused by your statement that "I am quite happy to debate every last sentence in it" - I presume you're a different William Grogan to the WilliamGrogan that I've been asking to respond in detail to the CEV report in detail since 13th May?

Testing is, of course, a question of degrees and professional judgement. It's not just about testing new versions - the CEV report clearly stated that insufficent end-to-end had not been done, even on the earlier versions.

I don't do complete end-to-end testing for minor changes to my systems either (though mind you, my systems don't decide who is going to run the country for the next five years). The CEV took the professional view (based on professional advice) that the significance of the change required to accomodate the referendum along with the other two/three elections was sufficient to justify complete re-testing. Of course, you are entitled to disagree with this view. Fortunately, the CEV don't agree with you.

And testing isn't the entire solution either - even if the system was tested to the nth degree to prove its accuracy under all possible scenarios, there are still huge security & control issues around the proposed solution, i.e. how do we know that the tested version of the software is the version running on the day? how do we prevent direct control over the MS Access database from outside the application?

williamgrogan

I'm amused by your statement that "I am quite happy to debate every last sentence in it" - I presume you're a different William Grogan to the WilliamGrogan that I've been asking to respond in detail to the CEV report in detail since 13th May?

That’s a question of when not whether. I’m busy.

And testing isn't the entire solution either

What was the point of the commission carrying out the tests they did then?

They presumably cost many tens of thousands of Euros. The tests confirmed that the software worked. Then the commission ignored the tests and shelved eVoting. Odd!

If the commission organised the tests then presumably they were not sloppy, useless and “unprofessional” tests but properly structured and carried out. These types of tests do put a certain lower limit on the unreliability of the software. The commission even emphasised that the tests were carried out on a large body of data and on many machines. If you do properly constructed extensive tests on many different machines and they pass then you can use these machines for the purpose they were constructed with a high degree of confidence that they will work. This is a scientific approach. To do the tests and still “not feel” confident is unscientific, humanlike I agree, but unscientific. Of course the commission are not scientists.

Furthermore the anti-eVoting lobby have claimed that the software was badly written and wasn’t properly tested but the testing helped refute those arguments. Even badly constructed software can function perfectly well. A simple top down program while inelegant and maybe hard to maintain can work with 100% accuracy. (I could go a bit further here and state that there are actually “fashions” in programming with certain techniques becoming popular and then in turn being replaced and not necessarily by something better.)

If the software was extensively tested and found to randomly lose 2% of the votes it would have been rejected. But that is the system we now fall back on because of the technological cowardice and “feelings” of the commission.

Here we are getting close to the nub of the problem. If the commission had enough time to do further end-to-end testing and the software passed that too, as I would now be very confident that it would, would that do? Would any amount of testing do? Probably not to the anti-eVoting lobby as no amount of testing will satisfy them. Might satisfy someone from a probability point of view but not overcome people’s “gut feeling” of insecurity. Even the daft VVAT concept with its random parallel counting is just more testing anyway.

If at the end of the day there is the slightest possibility that an error can occur then certain individuals do not want to take the risk – the serious downside of the Precautionary Principle.

The “HUGE security and control issues” are completely refuted by all election officials in the world whose opinion I have read. I have read dozens of comments by election officials in many countries and not one has expressed that this is a problem never mind a HUGE problem. To the best of my knowledge there is not a SINGLE case of such corruption after billions of votes have been cast. I think most people know that to corrupt an electronic voting system is several orders of magnitude more difficult than a manual system. That’s partly why I think eVoting was introduced in India.

All more security will do is decrease the possibility that corruption will occur further. I have no problem with this and would foresee that after future attempts to interfere or more weaknesses are found further security mechanisms would be introduced. However, 100% security is also impossible and non existent in all walks of life. Why should eVoting have 100% security? Demanding it is Luddism. The attempt to hold up technology with unreasonable demands, like a man walking in front of a car with a flag.

PS

It’s interesting that SF got by far the highest increase in votes. If we had voted via eVoting presumably the internet would now be buzzing with the “news” of a surprise result where a party doubled its vote in the first eVoting election in Ireland, a party that has a history of voting irregularities. No doubt that certain TDs would stand up in the Dail and claim this as evidence that eVoting was unreliable as they have done before.

ecksor

Oritinally posted by williamgrogan
Even the daft VVAT concept with its random parallel counting is just more testing anyway.

Whether it's another test or not isn't the point, the point is that it provides a method of audit after the fact. You have admitted that your problem isn't with providing a method of audit, but with the fact that paper is used for it.

So, to repeat my question again:

And you've even said that you'd have sympathy if the campaigners for VVAT were looking for an electronic VVAT, rather than a paper one. Can you provide an example of a non-paper based VVAT that has stood up to scrutiny?

williamgrogan

Can you provide an example of a non-paper based VVAT that has stood up to scrutiny

I am not going to design an Audit sub system for eVoting, unless someone pays me a few hundred grand. Not am I going to spend my time searching for one.

It’s quite obvious how one could construct a second parallel data collection and counting system supplied by say a different company including a second separate encrypted file with raw data that was separately tested and counted after the election. Actually NASA do this with the Shuttle software. Three separate computer systems fly the shuttle and process in parallel running 3 separate suites of software written by 3 separate companies in 3 different continents. I think there is some argument for this with a shuttle but none for a trivial non critical voting system. I might add that with all this overkill re system reliability it was 2 mechanical faults (which could be interpreted as NASA organisational faults) that brought down the shuttle and not a software fault. The same can be said of the Airbus. There have been several fatal crashes but none caused by a software bug. It seems to me that a software bug is unacceptable but a mechanical failure isn’t. In fact anyone supplying software to business can see this. A minor penny error in a computer program causes consternation but in a manual system errors in the thousands are acceptable. I also see it with computerised stock control versus manual stock control.

The systems do have some level of internal audit before and after the voting. The election officials use this as it is.

The reason that VVAT is called for has nothing to do with its stated purpose, it’s called for because paper is tactile and people are used to it. People simply prefer paper to electrons. The fact that those calling for VVAT have not apparently considered electronic audit is in itself proof of this. The clique that support Ms Mercuri are opposed to eVoting and even with VVAT most of them still oppose it which is why I say that it meets no need anyway. Some people and I agree with this, have said that if eVoting needs VVAT then eVoting cannot be acceptable.

Maybe there is no more elaborate non-paper based audit because no one sees a need for it. The election officials don’t and the eVoting companies don’t.

If a non-paper audit system was designed and implemented all it would achieve is an increase in reliability, not a guarantee of 100% security. From what, 99.9% to 99.91%. We always get back to this point. What is the level of security that is acceptable? The Precautionary Principle.

I think normal system testing and implementation by a professional computer company is sufficient and the anti-eVoting lobby think nothing is acceptable. That’s the difference. No amount of testing and nothing short of a complete parallel paper system will satisfy the Luddites. This has always been the case and always will be when new technology is introduced.

In a nutshell many people want 100% security and without it they do not trust technology. 100% security is impossible. Therefore no matter what is done they will not be satisfied.

RainyDay

Hi William - Let me try a 3rd time, as you seem to be ignoring this question - Are you still claiming that the CEV stated that "they believed that the system would work"?

I can't understand why you are worried about a possible CEV spend of tens of thousands in the context of the overall spend of 55 million by Minister Cullen, including several million on the PR campaign. Are you suggesting that the CEV shouldn't have done any testing? You are again being highly selective in your statement that "The tests confirmed that the software worked. Then the commission ignored the tests and shelved eVoting. Odd!" - Don't forget the other CEV findings, i.e. they were unable to test the actual version of the software that would be running in the live environment as it didn't exist at the time, the end-to-end testing which had taken place was deemed to be insufficient, and the absence of a VVAT significantly raises the standards and quality of other system testing that is required.

Again, I'm surprised that as a card-carrying skeptic, you place such confidence in "dozens of comments by election officials in many countries and not one has expressed that this is a problem never mind a HUGE problem". Surely any such comments are purely anecdotal - Did anyone actually ask the election officials the important question? It is just not true to say that "I think most people know that to corrupt an electronic voting system is several orders of magnitude more difficult than a manual system". With the proposed new system, it is possible for an election official to replace the entire vote database with a 'heres one I prepared earlier' using a single 'copy' command, by copying in a new Access .mdb file. It is just not physically possible to attack the entire set of manual votes in this way. Perhaps you can advise us of your proposed solution to this problem?

Just to confirm my position, I'm not against eVoting in principle (though I do have doubts about the value-for-money issue). I am against the proposed Nedap system, simply because I can't see any way that this eVoting system can be safely implemented without a paper audit trail. But I'm open to arguement - if you can advise how one can protect the Access .mdb file, that would be a great start.

ecksor

Originally posted by williamgrogan
I am not going to design an Audit sub system for eVoting, unless someone pays me a few hundred grand. Not am I going to spend my time searching for one.

Or you could merely take the proposed ones I mentioned earlier in this thread, although they're not around long enough yet to have 'stood up to scrutiny'.

It’s quite obvious how one could construct a second parallel data collection and counting system supplied by say a different company including a second separate encrypted file with raw data that was separately tested and counted after the election.

If you won't supply exact details of how exactly this would work then it's not obvious at all, which is why security protocols take so long to gain acceptance.

The reason that VVAT is called for has nothing to do with its stated purpose, it’s called for because paper is tactile and people are used to it. People simply prefer paper to electrons. The fact that those calling for VVAT have not apparently considered electronic audit is in itself proof of this.

Actually, the e-voting research community who are pro-VVAT does spend time investigating electronic audit, as you should know if you'd been reading the links supplied on this thread. Which is why I ask the question, how can you be so sure that there's a good electronic audit mechanism when you can't supply an example that hasn't been around for longer than a couple of years for the purposes of scrutiny.

If a non-paper audit system was designed and implemented all it would achieve is an increase in reliability, not a guarantee of 100% security. From what, 99.9% to 99.91%. We always get back to this point. What is the level of security that is acceptable? The Precautionary Principle.

You're arguing a quantitative risk based analysis again, despite not having presented one for the irish election system or how to go about doing one (you abandoned the attempt on this thread apparently). What assurance do we have that the accuracy is 99% and not 89% ? That's a genuine question which I'd like to see hard evidence in response to.

For the record I'm inclined to believe that the answer is closer to 99% than 89% and probably closer to 100% than 99% , if you assume smooth running of the system. However, I'm not sure how you go about measuring the risk over the course of several election uses assuming, for arguments sake, one failure per million and whether or not it has the potential to destroy voter confidence.

williamgrogan

Are you still claiming that the CEV stated that "they believed that the system would work"?

For the 3rd time.........

Goto{TestedAndWorks}

RainyDay

The CEV simply never said 'it would work', William. They said 'it can work' and 'it could work' - but not that it would. It's an important distinction.

williamgrogan

The CEV simply never said 'it would work', William. They said 'it can work' and 'it could work' - but not that it would. It's an important distinction

Wow!

So the CEV cannot tell the future. Well that comes as a surprise!

If you run a program and it works, then given the same circumstances it will work again if run in the future. It’s one of the great things about software, it’s predictable. The whole idea of testing is to try and create all possible circumstances or at least to a reasonable degree of certainty.

The same can be said of MMR. If tested on hundreds of thousands of children without any problem then there is a very high probability that it will not cause massive problems and any problems should only occur in a tiny number of cases.

Again this is the difference between our positions. I accept a reasonable degree of certainty knowing that the present system is at best 98% accurate and that the proposed system is more accurate and very important will become more accurate with use and experience. Furthermore the move to eVoting has many other benefits such as allowing voting more often and quicker results and eventually real eVoting via the Internet/ATMs etc. It certainly is one way to encourage people to vote – make it easy. (I’m reminded of one idiot on the radio who said that having to traipse to the polling booth was a good idea as it dissuaded those without a serious interest in politics from voting.)

The CEV also said that they didn’t say it wouldn’t work. So what does that mean?

Think about this in everyday terms; I decide to buy a new car, I know form reading that other people drive this car and that there has been no complaints about reliability, I get a mechanic to thoroughly test it, he says it passes all tests, I drive it on a short journey and it also works, I am then confident to a high degree that if I drive it on a long journey it will get me there. So I buy it. What else should I do never drive on the basis that no amount of testing will not prove 100% that in some set of circumstances that the car will fail? (I even buy the car knowing that it might be interfered with or stolen.)

Even if it does fail the probability is that the failure will be trivial, e.g. it’s actually far more likely that a fault will mean the cigarette lighter will fail then the car will explode or suddenly drive into the ditch. We know this instinctively for cars but even programmers do not have this “feel” for their software. I know because I can understand the fear that even after thorough testing you always think that the software will let you down. It’s irrational.

Software is even more reliable as it doesn’t wear out or break down in the way a car would with repeated usage.

ecksor

Originally posted by williamgrogan
I know because I can understand the fear that even after thorough testing you always think that the software will let you down. It’s irrational.

It's not totally irrational for reasons that I described in my post here

'thorough' is an unsuitable word to use here. You can't test software completely as I'm sure you well know, but 'thoroughly' seems a bit misleading on a thread where you have no problem throwing around figures like 98%/99%/100% with no evidence to back them up.

williamgrogan

With a computer a small change in inputs can cause wildly different outcomes (for whatever reason, boundary conditions, odd workarounds, limitations of datatypes etc.). There is no effective way to judge the 'similarity' or 'nearness' of two different uses of a computing system in the same way that if I plonk my 70 kilos on a chair and it works fine every day, then my 80 kilo friends will probably be ok and my 60 kilo friend will definitely be ok.

“Boundary conditions” and “limitations of datatypes” are both boundary conditions.

Anyone testing software will specifically test these conditions. So we have “odd workarounds”. I don’t really know what you mean by that I never do that.

I do not think your point is valid here. Testing has to be done properly and must take into account what is being tested. I disagree that testing software is more difficult than testing anything else. It’s different but while there are downsides there are also upsides. I would go further and say to test a chair for strength in the real world is far far more difficult then testing a piece of software that records two dimensional data and accumulates it.

ecksor

Originally posted by williamgrogan
“Boundary conditions” and “limitations of datatypes” are both boundary conditions.

A lot of what I mean by "limitations of datatypes" are certainly boundary conditions, but you could also include the likes of rounding problems with float types or the "is a" versus "has a" style design errors that you sometimes find in OO classes. For the purposes of this discussion I don't mind lumping them together though.

Anyone testing software will specifically test these conditions.

Well, my experience is that testing ranges from very poor to reasonably good. I'm certainly not prepared to accept that 'anyone' would test all those conditions. Indeed, I'm not prepared to accept it's practical or necessary or desirable to do so without a solid argument. Is it even possible?

So we have “odd workarounds”. I don’t really know what you mean by that I never do that.

Since you've put a smiley there and you have often mentioned the fact that you've been developing a software system for business use for 20ish years now I'm assuming that's a wee joke ...

I do not think your point is valid here.

I should be clear that I'm not trying to convince you of anything here. I don't think that's possible. I'm merely debating on the off-chance that anyone is still reading and hasn't made up their minds, since this is an important issue.

So, to put the point I tried to make above in another way (and tie it back into the actual point of my participation in this thread), it isn't possible to test all inputs or combinations of inputs into a piece of software any more than it's possible to test all inputs (for want of a better word) or combinations of inputs into something like the stress that a chair can take. However, I think that most people will accept that I can establish the range in which the stress a chair can bear lies using the reasoning I have outlined. That isn't possible with software. There is no way of establishing an empirical "mean time to failure" for a software component in the same way that you would with parts of your car or whatever.

There's a brief description of this at http://www.cigitallabs.com/resources/definitions/software_reliability.html

Even with such information reliably gathered and presented with clear evidence and justification to the irish population, it is still the case that we must be prepared for the eventuality that the system will fail (maybe next week, or maybe not for another 20 years) although we also have the wee issue of having no way of knowing if it does fail if the errors are subtle enough). The system has no method of auditing or checking the result and no way of properly resolving a dispute. Of course we also have the wee issue of having no way of knowing if it does fail.

Testing has to be done properly and must take into account what is being tested. I disagree that testing software is more difficult than testing anything else. It’s different but while there are downsides there are also upsides. I would go further and say to test a chair for strength in the real world is far far more difficult then testing a piece of software that records two dimensional data and accumulates it.

You can say it but it's yet another vague statement with no argument to it. Of course the testing must be done properly, but the testing doesn't solve the problem as I'm trying to illustrate. There must be an audit mechanism that is capable of resolving disputes or problems.

RainyDay

Originally posted by williamgrogan
Wow!

So the CEV cannot tell the future. Well that comes as a surprise!

Well clearly it must come as a bit of a suprise to the guy who told us on Saturday that 'they believed that the system would work'. Who was it that told us this again?

Originally posted by williamgrogan
Again this is the difference between our positions. I accept a reasonable degree of certainty knowing that the present system is at best 98% accurate and that the proposed system is more accurate and very important will become more accurate with use and experience. Furthermore the move to eVoting has many other benefits such as allowing voting more often and quicker results and eventually real eVoting via the Internet/ATMs etc. It certainly is one way to encourage people to vote – make it easy. (I’m reminded of one idiot on the radio who said that having to traipse to the polling booth was a good idea as it dissuaded those without a serious interest in politics from voting.)

William - Please advise exactly how eVoting will allow us to vote more often? It's not cheaper than a manual vote (in fact, it requires MORE manpower to run an eVoting election than a manual election).

Please also advise how eVoting will encourage people to vote. I've seen no data supporting this contention. The best research I saw was from the UK's Electoral Reform Society, and that showed no improvement in turnout levels from a range of eVoting methods (kiosk, Internet, SMS text, Digital TV). The only method that showed an improvement in participation rates was good old-fashioned postal voting.

williamgrogan

Testing chairs v software.

I beg to differ.

Have you taken into account that the failure of a chair is also related to the angle that the fat guy sits on the chair. In other words a 100Kg man could carefully sit on a chair with 4 legs and weak cross pieces and the chair would survive. However a 50Kg weakling might sit on the chair coming from an angle, even a very slight one, and the chair would collapse sideways.

These types of problems occur in the real world. Software to a large extent doesn’t exist in the real world. Therefore I maintain that a software system is easier to test than non software systems. All data input is validated and the “universe” of the program is very restricted.

If you write a simple program to accept and store say a number between 1 and 10 and then repeat this until “S” for stop is pressed, then print the result, it is possible to test this to a very high degree of certainty. As you make the program more complex the testing gets more demanding but for the voting and the counting of votes it never gets so complex as say an accounts package with master files, transaction files and multi-user capabilities etc. Nor as complex as shuttle software or telecommunications software where timing and comms throw a spanner in the works. On a software hardness scale of 1-10, I would rate an eVoting system as 1. I cannot think of any commercial package that’s less complex and less prone to bugs.

The accurate testing of a software package is a matter for the professionals concerned and we must trust them in the same way we thrust other professionals such as those that design and make chairs.

Do you demand evidence that the new chair you buy will support your weight or do you thrust the professionals who made it?

To some extent the CEV don’t trust themselves. They were hired to asses the software and they hired in experts to test it and it passed the tests. They should have accepted that.

Your point re the problems of testing software is academic. Do they try to come up with mathematical statements that guarantee the reliability of hardware components? I think a lot of this is generated because software has the nature it has. Academics discuss the possibility of making software 100% bug free but no one tries to make a car bug free or a house or an airplane except by testing. Then they accept the testing and run with the system, weaknesses and all. We have spent months arguing about certainty in eVoting that goes way beyond what is acceptable in every other system.

Let me ask you a simple question, what is your estimate of the probability that after the commission tested the eVoting system and it was then used and after the local elections even one councillor was elected who shouldn’t be? India ran an election with ~500,000,000 electors with simpler systems and I have yet to hear of any claim that someone was elected who should not have been. In fact I know of not a single claim of this nature from any eVoting system.

The proof…

I can prove what I say is true, that computer systems and software are more accurate than manual systems.

Accounts software is far more accurate than the old pen and paper method. Through the use of computers small companies now maintain a level of accounts detail and accuracy only possible in large corporations 20 years ago. Computer systems have dramatically reduced the accident rate in the airline industry. In fact everywhere computers are used to replace manual systems they dramatically improve accuracy but if we are to believe the anti-eVoting lobby, eVoting will be the exception. If you think about it it’s a ridiculous proposition. Computers = Good, Paper & Pens =Bad, it’s that simple. It’s as ridiculous as the statement by my opponent on the MMR debate who thinks overall doctors do more harm then good.

A very important point to make here is that all the above systems which are more accurate than paper have been developed in a similar way to how the Nedap system was developed, by ordinary programmers, working for private companies, programming right up to the last minute.

Auditing

There are internal audits within the system. Audits that satisfied the Returning Officer who ran the eVoting live test.

To some extent because of the secrecy of the vote one can never 100% audit an election. Because the obvious ways all link back to knowing who voted and how he voted. Therefore we must accept that whether it’s a manual of computer system we cannot trace back.

The recount myth is typical of a misunderstanding here. People who advocate a paper trail want it for recounting. As if recounting was all about verifying the vote. It isn’t. In our manual system the recount is simply there to check that the humans counted correctly and that the votes that were spoiled or partly spoiled are allocated correctly. These situations cannot exist with a computer system so a recount is pointless. One of the benefits of eVoting is that you can eliminate the recounts. The anti-eVoting lobby have put forward the disinformation that a recount is some way to validate that all the votes counted were legitimately cast. It isn’t. You can recount until the cows come home and yet Charlie Haughey’s election agent managed to vote twice. (An electronic voting card with a pin number linked to the DRE might well eliminate this problem and still maintain voter secrecy.)

Next Step

It is going to be very interesting to see what the CEV decides next. To some extent they have painted themselves into a corner. Will they do more testing and if this still shows no problem and the testing is done with the software for say the next general election what then? More testing? Will they define an amount of testing that is enough? Can they? Or will they get stuck in the philosophical unreal academic universe where producing a paper on the number of guardian angles on the head of a pin gets you a doctorship?

Re RainyDay’s comments……

The election officials are the professionals re voting. It’s not anecdotal to comment on what they say. Virtually to a man they are delighted with eVoting. I can’t see they have any axe to grind. They are not biased other than wanting the similar technological fixes to boring counting problems that everyone else wants and has had for years.

I think many of you throw out ideas about corruption without considering them at all. The practical considerations to replacing the database are obvious.

To create such a database to produce meaningful results would be enormously time consuming. If not done with extreme care they would be easily detected. I think we would all get suspicious if there was the same spoiled vote count in every ward or that the vote bore no relationship to the last election count for every individual ward or too much of a relationship. Someone has to want to carry out the corruption and to risk jail if caught. Any party involved in such corruption would almost certainly be finished. You would need to know the database structure. You would need access to the computers and at the appropriate time. Etc……………..

I think too many of the anti-eVoting lobby have seen too many episodes of the X-Files or watched too many bank robbery movies.

What is theoretically possible and likely are miles apart.


Other Comments from the CEV Report now that I have wasted my entire day……….

The majority of the submissions were against implementation of the chosen system. To be expected. It means nothing. You cannot vote on facts. Bad news lobbies are far more prevalent than good news lobbies. There are no pro-supermarket lobbies but lots of opponents yet the vast majority of the population shop in them. There are lots of very active pro-organic food lobbies but it represents only 5% of the food market. There are many lobbies opposed to pornography but it represents the largest use of the Internet. I could go on.

The actual reason eVoting did not get implemented for the last election is very much related to who is currently Taoiseach, Bertie Ahern. I am an admirer of Mr Ahern; I suspect that history will declare him Ireland’s best every leader. I am not a FF supporter and certainly during Charlie’s reign I would never vote for them. However, Bertie is a great compromiser and when the opposition kicked up a fuss Bertie tried a manoeuvre that he assumed would work. Pick a committee and assume they will come up with the correct result and sure even if they don’t then it’s not his fault.

The commission only had 9 weeks and hopefully that was the major reason they felt unable to confirm that they were happy with the system and that with further time they may well support the eVoting system picked.

At the end of the day what did the commission find wrong with the system chosen? That they were still modifying the software? Presumably because they had to because they were asked to. That the commission had not enough time to test it fully. That, rightly so, the company were reluctant to hand over their source code. A tightening of procedures was recommended. Never a bad idea that one. Further security is required. Isn’t it always? No actual criticism was made of the company except maybe that the hardware was not hard enough.

The commissions report is an interim one and obviously we need to see the full report.

I think 4.5 "Recommendations for Action", are reasonable and achievable. I see no reason why the eVoting system that we have purchased cannot be used for the next elections. I am still totally opposed to the Luddite solution of a parallel paper audit.

Victor

Originally posted by williamgrogan
I might add that with all this overkill re system reliability it was 2 mechanical faults (which could be interpreted as NASA organisational faults) that brought down the shuttle and not a software fault. The same can be said of the Airbus. There have been several fatal crashes but none caused by a software bug.

Actually it was software that brought down the aircraft at an airshow in France (crashed in woods). The pilot was doing a touch and go, but the software insisted in "landing" the aircraft. For every pull up manouvre the pilot did the software overcorrected to land and eventually crash the aircraft.

williamgrogan

Actually it was software that brought down the aircraft at an airshow in France (crashed in woods).

It wasn't a bug. The pilot was fired. Furthermore it was an Airshow and he was flying rather carelessly.

Incidentally the majority of crashs are caused by pilot error. The more technology that is used the safer its becoming. A major point here which also applies to eVoting, overall it is likely to be far more accurate no one is claiming that 100% of all votes cast forever will be correctly counted but that it will be far more accurate than a manual system. The era of most risk will be the first few elections but we must get through them to progress.


from Google

about the first crash of the A320 at the airshow:
- from the pilot of the A320 that crashed at the airshow:
. the altimeter was wrong, it said 100 feet when actually it was only 30
. the pilot claims he was pulling back on the stick and increasing the
throttle, but the computers kept the throttle and elevator constant,
holding the plane in a perfect straight-and-level, low speed cruise.
(this can clearly be seen in the incredible film of the plane flying
gracefully into the forest on the shallow hillside)
. the trace from the black box he shows confirms this: the stick is
coming back, but the elevator in fact is going *down* (I really didn't
understand the trace, since the plane flew straight and level into
the trees; perhaps the elevator line was supposed to be a "delta" added
to the stick position by the computer, to arrive at what elevator position
it thought should be "correct")
. "the computer has no eyes; it couldn't see the forest coming up, and so
it assumed I wanted straight and level flight"
. the black box was carried away in a police car without being sealed,
which it is supposed to be by law
. most critical last 4 seconds is missing from the trace. He says there
has been obvious tampering with the results.

- from Airbus Industrie:
. the altimeter problem was documented
. pilot error - he was flying too low, to slow, and thus the crash is
not surprising
. out-of-hand dismissal of tampering accusations

RainyDay

Hi William - I'm still absolutely gob-smacked that a so-called skeptic places such great weight on anecdotal evidence. There has never been any formal survey of returning officers opinions. In relation to the press reports that we have seen, we have no idea what actual questions they were asked by interviewers, so it is very likely we are only seeing one side of the arguement. And even if these two weaknesses were addressed, how much do these people know about technology? Do they understand that the entire vote database (including the audit trail) can be attacked/corrupted with a single command on the vote PC? Do they understand that that it is not possible to verify what version of software is running on the VM? Why would we take the word of a non-technical community regarding a technical solution?

I'm also still awaiting your response to any supporting data to back up your statements that eVoting facilitates more frequent voting and improved voter turnout.

Actually, you're right about replacing the database. This wouldn't be smart, as you wouldn't know how many votes were going to be in the database beforehand. However, it would be amazingly easy to write a bit of VB to pass through the votes in the Access DB and bias them towards one candidate. Not a huge bias of course, just enough to get the seat.

And your comment that "I think we would all get suspicious" demonstrates the critical need for an audit trail better than I could have done so myself. So what if anyone has suspicions - there is absolutely nothing that can be done about it afterwards. Any recounts will simply produce the same (bad) result again & again & again. Without an audit trail, there is nothing that can be done to investigate any suspicious result.

"Any party involved in such corruption would almost certainly be finished" - It won't be a political party that will be doing the corruption. It will be one of the several wealthy individuals with commercial interests that will benefit hugely from a right-wing government, the guys that run privatised hospitals, schools, prisons, TV stations etc etc. We hear in the tribunals every day the corrupt lengths to which these guys are prepared to go to in order to build a little shopping centre or two? Do you really, really think these guys wouldn't be prepared to risk their necks for the potential prize of steering the government to the right in order to line their own pockets over the lifetime of the next government.

Your opposition to a paper audit trail is an entirely valid opinion and point of view, but don't kid yourself that it is a skeptical position based on hard data - nothing could be further from the truth.

williamgrogan

There has never been any formal survey of returning officers opinions

I have had a Google alert running for several months and daily read articles about eVoting. The overwhelming comment by election officials is positive. That in itself is a survey. If a formal one was done and confirmed what I said then you would obviously just repeat what you said, that these are non-technical people so what would they know.

I have said all along that a similarity between the pro-VVAT position and anti-MMR is not just a technical argument but that the anti-eVoting lobby are part of a wider agenda. Many of the activists are pro-Open Source (which Ms Mercuri says is no use) and anti-big business (another similarity with the anti-Big Pharm anti-MMR brigade) and probably anti-Microsoft. I actually think there is a link between the socialist mind set and Open Source. [I mentioned before that I find young people being anti-MS funny because when I was young I was anti-IBM for similar reasons. No one likes a monopoly. Dell incidentally quoted me €600 last week to replace a €20 bulb in my laptop monitor.] One poster even boasted that he had no MS products running on his PC.

In Ireland it became a Government v opposition issue because the opposition chose to make the issue a political football. Political opportunism.

In India for example there is no widespread opposition to eVoting, nor in Europe. This also mirrors the situation with anti-MMR. No opposition anywhere except in the UK and Ireland. It’s similar with Nuclear Power, virtually no opposition in France where 80% of electricity is NP. When this happens you must be very skeptical of the anti-brigades position. I’m sure the Indian’s must be wondering what the fuss is here. Total silence from the anti-eVoting lobby regarding the successful Indian election.

Your comments about corrupt right wing business men plotting to overthrow the government by fixing the election is laughable. I don’t intend turning this into a political debate but I will just repeat a headline from the papers recently, “We are all PDs now”.

Skeptics are and should be dubious about any anti- or pro- movement knowing full well that many exaggerate their position and hold very strong positions based on flimsily evidence. I have already listed the similarities between the anti-MMR movement and anti-eVoting, there are many similarities. I’ll list them again if you like?

Making it easier to vote must increase the vote. It hardly will decrease it will it? England has yet to introduce eVoting so your comments are not valid. The people who are travelling on business cannot vote (2,000?), the students from down the country, (20,000) those that have just moved or are in hospital or on holidays (20,000?) cannot currently vote, with Internet voting they could. My brother and his wife are in Spain on holidays during the election, he would certainly vote if he was here. Proof that making it easier and possible to vote did come from the limited postal vote recently in Britain.

Everything is made easier nowadays so why not voting?

I fully understand the gut feeling that people have that eVoting is unreliable. I had it myself until I switched on my brain.

RainyDay

Hi William - There really is little point in debating with you. Anytime I try to pin you down to specific issues, you fire off with vague generalisations. It's a typical 'tabloid' technique to avoid engaging in rational debate - just fire out a few labels (in your case, open-source/socialist) to distract attention from the real issues. But just let me take a moment to blow away a few of your most recent errors.

.

Many of the activists are pro-Open Source (which Ms Mercuri says is no use) and anti-big business (another similarity with the anti-Big Pharm anti-MMR brigade) and probably anti-Microsoft. I actually think there is a link between the socialist mind set and Open Source.

I'm not pro-Open Source or anti-big business. I'm an employee & stockholder of a large US hi-tech multinational. Do you reckon the background of the leading opponent of the Nedap system here in Ireland, Joe McCarthy (former PD election agent, former IBM'er for many years, now running own consultancy company) is consistent with your socialist/open source labelling? Do you reckon that the Irish Computer Society is now a socialist/open source organisation? This is just rubbish - you'd be far better off engaging in rational debate on the important issues.

In Ireland it became a Government v opposition issue because the opposition chose to make the issue a political football. Political opportunism.

ICTE is not a political organisation (though some of its members are politically active). The Irish Computer Society is not a political organisation. This issue became a political football when Minister Cullen refused to listen to reason. If he had listened to the messages in the Labour Party report on eVoting from October 2003, he would have had 6 months to actually address the core issues (lack of testing, version control, audit trail) which were the blocking issues for the CEV six months later.

Making it easier to vote must increase the vote. It hardly will decrease it will it? England has yet to introduce eVoting so your comments are not valid.

You can repeat this as many times as you look, but it doesn't make it true. You have no data to support this view. [How skeptical is that] The UK ran specific trials in real elections on a range of eVoting methods (kiosk, Internet, digital TV, text message) and postal voting over recent years. The only method that showed an increased turnout was postal voting.

I fully understand the gut feeling that people have that eVoting is unreliable. I had it myself until I switched on my brain.

Funny that - I had the gut feeling that it was reliable, until I really starting looking into the Dept Environment papers. Then my view changed very quickly....

williamgrogan

Many of the activists are pro-Open Source (which Ms Mercuri says is no use) and anti-big business (another similarity with the anti-Big Pharm anti-MMR brigade) and probably anti-Microsoft. I actually think there is a link between the socialist mind set and Open Source

Please note the first word in the paragraph. It does not mean all. That's a different word.

Look up the anti-eVoting web sites or read their posts.

It's not a generalisation to say what I said above. As a skeptic I am very interested in why people think as they do. Most people do not actually *think* rationally about anything. Most of their opinions are hand me downs or as a consequence of cultural or even genetic influences. There is evidence that separated identical twins share political opinions.

I believe that the reasons people adopt positions such as anti-MMR have little or nothing to do with the arguments themselves but more to do with their view of reality. So it is with the anti-eVoting lobby. No doubt there are many that dislike the particular system the government choose but there would be an anti-eVoting lobby no matter what system the government choose. The anti-eVoting lobby in the USA oppose all systems that do not have paper votes and manual counting which of course is not eVoting.

Are you denying that there are people who fall into the category of anti-Business, anti-MS, pro-socialist and pro-Open Source? In fact the anti-Globalisation movement would be all these things. Go to Glastonbury next weekend and you can meet 100,000 of them.

RainyDay

Hi William - This is fairly meaningless stuff. I'm sure that are eVoting activists who have yellow cars. There are eVoting activists who have big noses. There are eVoting activists who practice Bhuddism. This doesn't make the eVoting movement a gang of big-nosed, yellow-car driving Bhuddists. You're not demonstrating any cause & effect here. Of course there are overlaps between the eVoting movement and other movements. But you would really be better off clearing away the fog of labels which you seem to impose on these groups and listening to the actual details of the debate.

And don't expect me to take you seriously as a skeptic given your demonstrated ability to make up 'evidence' to suit your arguement, e.g.

- The CEV said the eVoting system would work
- eVoting will improve turnout
- eVoting will facilitate more frequent votes

williamgrogan

You're not demonstrating any cause & effect here. Of course there are overlaps between the eVoting movement and other movements.

The cause is that some people believe in socialism and oppose business; they are the type of people to whom the word “profit” or “deal” or “shareholder” or “millionaire” or “fat cat” are dirty words. Given a choice between Open Source and a secret code owned by a business whose primary purpose is to make money for its shareholders they go for the Open Source, “the People’s Programs”. The effect is that the merits of the two approaches become secondary. I’m hardly stating anything that’s not obvious here. There is no connection between those with yellow cars and Open Source but I bet there is a connection between those with bicycles and Open Source, not to mention sandals, free love etc.. In fact Linux, which is really a Unix offspring, was written by people who wore sandals, everyone over 40 knows that.

There is no open source solution that the Irish government can roll out, its vapourware. There is a perfectly adequate solution that we purchased. The fact that we are not using it is nothing to do with its merits or otherwise, it is to do with the mind set of its opponents and their success in getting the better of the government.

No self respecting nerdy programmer would consider using paper in a fit as part of a computer system EXCEPT that there were other agendas.

I asked you what this sentence meant, CEV said … "is not based on any finding that the system will not work".

For the 4th time.........

Goto{TestedAndWorks}

If they confirmed as a result of their professional testing that "the system can accurately and consistently record voter preferences" does that not imply that it works and will work again or will the laws of the universe take a different turn next week? You do believe that testing has merit? At worst the CEV said they hadn't enough time to do more testing. If they do more testing and the system passes will you then be happy, because if not there is no point doing the testing? In that regard you will then be in the exact same position as parents who endanger their children's lives by not immunising them.

Ecksor linked to an article where the writer lamented that no evidence of a bug was a problem as the reliability couldn’t be measured as it always came out as infinity. Is this what we are waiting for, some evidence that there is a bug so that the academics can calculate the probability of failure?

Incidentally here’s another link between the anti-MMR movement and anti-eVoting; apparently 1/3 of all parents not immunising their children who have paid for the 3 separate jabs in the UK are medical professionals. Unbelievable!

I asked some simple questions in the last post. Do you think that if people could vote via their ATMs or the Internet, as well as the old fashioned way, that the vote would

a) stay the same
b) go up
c) go down
d) haven't a clue

RainyDay

Hi William - It's becoming clearer and clearer that your objections to eVoting have little to do with eVoting or skepticism. You just don't like sandal-wearing, left-leaning, bike-riding, anti-war, anti-globalisation anti-eVoting people. Of course, you're quite entitled to feel this way, but you're not kidding anyone here with your philosophical guff. It's just a plain old-fashioned personal predjudice.

Paper as part of any IT system is neither inherently good nor bad. It has it's place (like a lotto ticket, or a signed purchase order, or indeed a voter verified audit trail).

You seem to want to take guesses as to the impact of eVoting on turnout. Let's not bother with the guesses - let's look at the real, hard data;

""The average turnout in councils which piloted e-voting methods actually fell by 1.5% and in only one council (Vale Royal) did turnout rise by more than 10%.
..... On the other hand, the use of all-postal ballots raised turnouts consistently to more than 50% and turnout rose by 25% in one council area (Blyth Valley). In those areas which had used all-postal voting before and did so again this time, turnout fell only slightly and this indicates that the rise last time was no flash in the pan."

Your comment that the CEV stated that the eVoting system 'would work' is untrue, regardless of how you twist their words and tell us what they implied. Your comment that eVoting will facilitate 'more voting' is untrue. Your comment that eVoting will improve turnout is untrue. How many more of these untruths are we going to have to wade through?

williamgrogan

Your comment that the CEV stated that the eVoting system 'would work' is untrue, regardless of how you twist their words and tell us what they implied. Your comment that eVoting will facilitate 'more voting' is untrue. {repetition} . How many more of these untruths are we going to have to wade through?

It’s always amusing that when you quote people verbatim that they claim that you are “twisting their words”.

I do believe as a general concept that the opinions, beliefs and religions that people hold have only a little to do with knowledge and logic. You don’t seem to grasp this point at all, e.g. FF voters more often than not come from FF households and the same applies to FG and Labour. Greens presumably come from tree houses

What do you think they mean by the bit I quoted where they clearly say that the tests show that the eVoting system as picked by the government works and as a consequence will work again. You have refused to answer any of the specific questions I have put to you.

I didn’t say that the eVoting system as introduced by the government would increase voter turnout in fact I suspect it might reduce it in the short term as those too old or frightened by the thought of using a computer shy away, what I said was that when we progress to full eVoting via ATMs and the Internet we will see an increase in voter turnout.

Here’s a list of some of the questions that you have not answered ...

1/ What does this sentence mean, CEV said that their advice to shelve eVoting … "is not based on any finding that the system will not work".

2/ If they confirmed as a result of their professional testing that "the system can accurately and consistently record voter preferences" does that not imply that it works and will work again or will the laws of the universe take a different turn next week?

3/ Do you believe that testing has merit? Or to put it another way, what is the meaning or the outcome of all the testing they did? Does it prove anything?

4/ Are you denying that there are people who fall into the category of anti-Business, anti-MS, pro-socialist and pro-Open Source?

5/ Do you think that if people could vote via their ATMs or the Internet, as well as the old fashioned way, that the vote would

a) stay the same
b) go up
c) go down
d) haven't a clue

PS

You said regarding my attitude to sandal wearing lefties, “It's just a plain old-fashioned personal prejudice.”

Wrong again, I’ve been to Glastonbury about 15 times. I like sandal wearing lefties, just not their opinions on certain matters such as being anti-technology.

RainyDay

Hi William - I was really hoping we could have a rational debate without it breaking down into a blow-by-blow analysis of your comments, but unfortunately, that is not the case, so here goes.

Originally posted by williamgrogan
It’s always amusing that when you quote people verbatim that they claim that you are “twisting their words”.

Pure fiction. No where in the CEV report does it say verbatim that "they believed that the system would work?". They did say that it can work, it could work and that their reasons for shelving the system were not because of any finding that it would not work. But they did not say that it would work. You have twisted their words to come to this conclusion, but it is your conclusion, not theirs.

Originally posted by williamgrogan
What do you think they mean by the bit I quoted where they clearly say that the tests show that the eVoting system as picked by the government works and as a consequence will work again.

The consequence/conclusion/implication that it 'will work again' is yours, not the CEVs. The CEV did not say that it will work again.

Originally posted by williamgrogan
I didn’t say that the eVoting system as introduced by the government would increase voter turnout in fact I suspect it might reduce it in the short term as those too old or frightened by the thought of using a computer shy away, what I said was that when we progress to full eVoting via ATMs and the Internet we will see an increase in voter turnout.

Poor memory, William. What you actually said was

Furthermore the move to eVoting has many other benefits such as allowing voting more often and quicker results and eventually real eVoting via the Internet/ATMs etc. It certainly is one way to encourage people to vote – make it easy.

You did not say that Internet/ATM voting would encourage people to vote. You said that the move to eVoting would
- allow voting more often
- [provide] quicker results
- [facilitate[ real eVoting via Internet ATM
- one way to encourage people to vote

You did not qualify your last point (encourage people to vote) as specifically relating to Internet/ATM voting. But if this is what you really meant, than so be it. You have still presented no data to support this view and you have ignored the evidence from the 2003 UK trails showing that it is incorrect.

Originally posted by williamgrogan

1/ What does this sentence mean, CEV said that their advice to shelve eVoting … "is not based on any finding that the system will not work".

It means just what it says - no more, no less. It does not mean that the CEV believe that the system will work. That is the conclusion that you have jumped to, but it is not what the CEV said.

Originally posted by williamgrogan

2/ If they confirmed as a result of their professional testing that "the system can accurately and consistently record voter preferences" does that not imply that it works and will work again or will the laws of the universe take a different turn next week?

You are conveniently ignoring the rest of the CEV report. In a way, the laws of the universe will change. Or to be more specific, the requirement changes. From running 2-3 simultaneous elections, to running 3-4 simultaenous elections including the referendum. As the software to do this didn't exist, the CEV (& the Dept Environment) could not test the software, therefore they could have no confidence about how it would work in the live environment.

Originally posted by williamgrogan

3/ Do you believe that testing has merit? Or to put it another way, what is the meaning or the outcome of all the testing they did? Does it prove anything?

I answered this earlier in the thread. Of course, testing has merit. It's all a question of judgement, quality & quantity. The CEV and most other commentators believe that insufficent testing was done. And given that they weren't testing the version of software going to run on the day, it actually proves very little.

Originally posted by williamgrogan

4/ Are you denying that there are people who fall into the category of anti-Business, anti-MS, pro-socialist and pro-Open Source?

Yes, I'm denying this. And I'm denying that I shot JFK & Abe Lincoln & Mr Burns too. Jeez man, what are you smoking?

Originally posted by williamgrogan

5/ Do you think that if people could vote via their ATMs or the Internet, as well as the old fashioned way, that the vote would

a) stay the same
b) go up
c) go down
d) haven't a clue

I believe that it would go down. But it really doesn't matter what I believe or what you believe. The best independent evidence available on this issue that I've found is from the UK 2003 trials, and this shows that the voter turnout decreases slightly.

williamgrogan

They did say that it can work, it could work and that their reasons for shelving the system were not because of any finding that it would not work. But they did not say that it would work.

They also said that they had it tested professionally. When companies write software and test it properly it goes live. It doesn’t get shelved. Otherwise what’s the point in testing it or for that matter writing it?

I said that real eVoting via the internet and ATMs would encourage people to vote as it would make it easier. The proposed system is not easier so I was not referring to it. But like all systems we cannot progress unless we make a start.

I “haven’t presented any data” to prove online voting will not mean a higher turnout? You say that you think that being able to vote online as well as in polling booths will mean voting turnout will decline. You can’t be serious? What about the tens of thousands of people who currently cannot vote because they are not in their constituencies? More opportunities to vote must mean a higher turnout. I think I can safely leave the readers decide whether you are likely to be correct or I am.

And given that they weren't testing the version of software going to run on the day, it actually proves very little.

This is incorrect. When changes are made to software more testing is required but not a complete system test of everything. Otherwise software could never me amended for a reasonable cost. Nothing that we make is tested exhaustively after every change. Any change has a probability of introducing a bug but an adequate level of testing is all that is required. Demanding too high a standard is back to the Precautionary Principle problem where you demand standards way beyond what is adequate. The present voting is not 100% reliable. Nothing is. We will always get back to this point. Red flag man in front of the car.

If at the end of the day you are saying that the CEV needs to do more testing are you then prepared to allow the eVoting system the government picked to be used if it passes these further tests or are you saying no amount of testing will do? You see this is exactly the same argument that the anti-MMR people put forward. No amount of testing or studies will make them happy but that’s because they don’t understand the maths of testing and doing studies.

RainyDay

When companies write software and test it properly it goes live. It doesn’t get shelved.

That is a very general statement. I'd agree with a qualified version of it - something like;

When companies write software which offers some significant benefits to the client and test it properly in a manner which closely matches the live environment it goes live. It doesn’t get shelved provided the price is sensible.

I said that real eVoting via the internet and ATMs would encourage people to vote as it would make it easier. The proposed system is not easier so I was not referring to it. But like all systems we cannot progress unless we make a start.

Let me understand exactly what you are saying here. So you're accepting that the Nedap/Powervote won't make voting easier. But your saying that we should spend €55m of taxpayers money on installing this very imperfect system to 'make a start', i.e. so the next generation of voting systems which we splash out taxpayers money on might actually make things easier?

This makes no sense. Here's an alternative approach - we keep our money in our pocket, we let other countries 'guinea pig' the systems, and when a decent system is available, we spend our money on it. Unfortunately, our hard-earned cash was burning a hole in Martin Cullen's pocket and he couldn't wait to hand it over to the Dutchmen.

I “haven’t presented any data” to prove online voting will not mean a higher turnout? You say that you think that being able to vote online as well as in polling booths will mean voting turnout will decline. You can’t be serious? What about the tens of thousands of people who currently cannot vote because they are not in their constituencies? More opportunities to vote must mean a higher turnout. I think I can safely leave the readers decide whether you are likely to be correct or I am.

At the risk of repeating myself, you can keep repeating this message until you believe it, William, but it doesn't make it right. It's not a question of your opinion vs my opinion. It's a question of your opinion vs one year of research in the UK. Oh & by the way, you might want to check out the results of this UK Govt survey on voter attitudes to ATM voting;

- The concept of voting via an ATM was overall disliked and generally rejected.
- A number of barriers were discussed in terms of this potential method of voting. The key concern was the perceived lack of privacy whilst voting. Respondents disliked using ATMs for financial services as they were not felt to offer enough privacy. It was often commented that people queued close to the machine when in use and could view the screen. The location of ATMs on the street also gave rise to fears of intimidation.
- Respondents queried how they would access the system. The use of cash point cards was rejected as an infringement of confidentiality. It was also questioned how those without a bank account and cash card would use the system.
- It was suggested the use of ATMs would be slow as queues would develop at machines.
- Concerns were also raised in terms of the reliability of ATMs, they were reported as often out of use.
- The use of ATMs was generally felt to be inconvenient as voters would not be able to vote from home.
- In a small number of cases respondents suggested that the use of ATMs would be unethical. It was thought that this method would lead to the commercialisation of voting.

But hey, let's not get bogged down in mere facts & figures. Let's just go with William's gut instinct - that's got to be far more reliable.

I think I can safely leave the readers decide whether you are likely to be correct or I am

I'm delighted that we can agree on one thing - lets leave the other readers to compare the two options;
a) William's gut instinct - with no detailed analysis or supporting data
b) One year of live trials of a whole range of eVoting systems in the UK and a detailed UK Govt survey of voter attitudes.

This is incorrect. When changes are made to software more testing is required but not a complete system test of everything. Otherwise software could never me amended for a reasonable cost. Nothing that we make is tested exhaustively after every change. Any change has a probability of introducing a bug but an adequate level of testing is all that is required.

You're missing the point. Yes - testing is an issue and you & I will never agree on this. But the other aspect of the 'vapourware'/missing software issue is that it clearly demonstrates the gross mismanagement of this entire project by the Dept of Environment. They spent €55 million of yours & my hard-earned taxes and they STILL didn't have a working version of the software to meet the requirements (referendum with local/euro elections) just two months before the election! I would guess that the CEV were seriously concerned about the abilities of all involved in delivering this system to get a working, tested system in place by 11th June, given that they had spent five years farting around on this and still didn't have the right software in place. It's not a great idea to base the next five years of local authorities on a rushed software change.

If at the end of the day you are saying that the CEV needs to do more testing are you then prepared to allow the eVoting system the government picked to be used if it passes these further tests or are you saying no amount of testing will do?

There is no way that I could support the Nedap/Powervote system unless it is retro-fitted with a voter-verifiable audit trail. However, I'm open to persuasion. Tell me exactly how you can be certain that the correct version of software will be running on each voting machine and count centre PC? Tell me how you will protect the integrity of each vote at each stage of its lifetime - from the time the voter presses the buttons until the time the result is announced? Unfortunately, your answers on these issues to-date have only stretched as far as 'I trust the returning officers' and that just doesn't hit the mark for me.

And finally, are you going to be honest enough to withdraw your statement that the CEV said that the Nedap/Powervote system 'would work'?