Mary Fisher wrote:
If a transaction is made with your PIN then you are deemed to have
authorised the transaction unless *you* can prove otherwise. And even if
you and your card were in a different country at the time how do you prove
you did not permit someone to clone your card and use your PIN?
Are you allowed topermit someone to do that?

Probably not allowed to permit someone to clone your card, but how do
*you* prove you didn't?

Owain

On Wed, 25 May 2005 16:25:51 +0100, Mark
wrote:

On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
. ..

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form. I'm
sure this is not a sign of wrongdoing, so why assume that anyone else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on most
forms that information is irrelevant (i.e. I'm making a political objection)
and what's more the options given are not ethnic origins. I don't hide the
information, I clearly state that the information is irrelevant. It's never
been followed up so presumably it IS irrelevant. If I didn't fill in the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant - but we will _have_ to
give it all the same or face fines or imprisonment. I'm very
uncomfortable about that situation which is one (of many) reasons that
I oppose ID cards.

As an IT consultant myself I can see the whole project going massively
over budget and probably be abandoned eventually because it never
works.

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

MM

In message , MM
writes
On Wed, 25 May 2005 16:25:51 +0100, Mark
wrote:

On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
...

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form. I'm
sure this is not a sign of wrongdoing, so why assume that anyone else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on most
forms that information is irrelevant (i.e. I'm making a political objection)
and what's more the options given are not ethnic origins. I don't hide the
information, I clearly state that the information is irrelevant. It's never
been followed up so presumably it IS irrelevant. If I didn't fill in the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant - but we will _have_ to
give it all the same or face fines or imprisonment. I'm very
uncomfortable about that situation which is one (of many) reasons that
I oppose ID cards.

As an IT consultant myself I can see the whole project going massively
over budget and probably be abandoned eventually because it never
works.

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

Well, apart from that, the finger in the air calculation that it would
take 50 years to get everyone's information made me chuckle

--
geoff

"raden" wrote in message
...
In message , MM

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

Well, apart from that, the finger in the air calculation that it would
take 50 years to get everyone's information made me chuckle

And that's assuming everybody was helpful and co-operated.

On Thu, 26 May 2005 00:11:50 +0100, "Mike" babbled
like a waterfall and said:


"raden" wrote in message
...
In message , MM

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

Well, apart from that, the finger in the air calculation that it would
take 50 years to get everyone's information made me chuckle

And that's assuming everybody was helpful and co-operated.

The retinal scan alone is expected to take 5 minutes if perfect
working is followed.

So say 35.000.000 at 5 minutes a shot.

Shame is that the bill for this will be a state secret until some
years hence, when it is all abandoned, and the buck comes home, and we
realise how much we have paid these lunatics.

On 25 May 2005 16:03:48 GMT, (Andrew
Gabriel) wrote:

In article ,
Mark writes:
As an IT consultant myself I can see the whole project going massively
over budget and probably be abandoned eventually because it never
works.

It's rather rare for the government to abandon a project which
doesn't work. That requires both:
a) An admission it didn't work, and
b) A realisation that it's best abandoned.
It's rare to see either one of these in practice, leave alone both.

Governments are very experienced in sidestepping these issues. After
all the current regime has managed to survive much worse (e.g. WMD -
or the lack of).

Maybe it will take a change of government but I'm sure even nulabor
will not be here forever.

Mark

On Wed, 25 May 2005 16:58:25 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
.. .
On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
...

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form. I'm
sure this is not a sign of wrongdoing, so why assume that anyone else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on most
forms that information is irrelevant (i.e. I'm making a political
objection)
and what's more the options given are not ethnic origins. I don't hide the
information, I clearly state that the information is irrelevant. It's
never
been followed up so presumably it IS irrelevant. If I didn't fill in the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant

Such as?

All of it? ;-)

Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

Mark

On Wed, 25 May 2005 21:13:40 +0100, MM wrote:

On Wed, 25 May 2005 16:25:51 +0100, Mark
wrote:

On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
...

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form. I'm
sure this is not a sign of wrongdoing, so why assume that anyone else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on most
forms that information is irrelevant (i.e. I'm making a political objection)
and what's more the options given are not ethnic origins. I don't hide the
information, I clearly state that the information is irrelevant. It's never
been followed up so presumably it IS irrelevant. If I didn't fill in the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant - but we will _have_ to
give it all the same or face fines or imprisonment. I'm very
uncomfortable about that situation which is one (of many) reasons that
I oppose ID cards.

As an IT consultant myself I can see the whole project going massively
over budget and probably be abandoned eventually because it never
works.

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

It would be laughable if it weren't true ;-)

Mark.

On Thu, 26 May 2005 09:09:45 +0100, Mark
wrote:


On Wed, 25 May 2005 16:58:25 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
. ..
On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
m...

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form. I'm
sure this is not a sign of wrongdoing, so why assume that anyone else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on most
forms that information is irrelevant (i.e. I'm making a political
objection)
and what's more the options given are not ethnic origins. I don't hide the
information, I clearly state that the information is irrelevant. It's
never
been followed up so presumably it IS irrelevant. If I didn't fill in the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant

Such as?

All of it? ;-)

Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

I might to 'hide' the fact, i.e. not broadcast it, that I worked in
Germany for many years, if I were applying for a job with someone who
was known to be anti-German, for example (perhaps his parents' house
was bombed in the war and everyone except him was killed). Now, one
might argue that I shouldn't be considering such a post working with
people who still hold a grudge after 60 years, but the point is, I
have something to hide that might threaten my chances of employment if
it came out. Conversely, if I were to apply to an educational
establishment as mentor or classroom assistant in modern languages,
that fact might be all to my benefit and I should publicise it as much
as possible. So, as in much of life, it's horses for courses.

MM

On Thu, 26 May 2005 00:11:50 +0100, "Mike" wrote:


"raden" wrote in message
...
In message , MM

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

Well, apart from that, the finger in the air calculation that it would
take 50 years to get everyone's information made me chuckle

And that's assuming everybody was helpful and co-operated.

The Government has now released the time per person to complete the
biometric scans, which is around 8 minutes, I believe. I can just
imagine the civil servants right now in Whitehall beetling away with
pocket calculators to whittle that time down to 7 minutes. "If we had
a kind of conveyor belt process, those in the queue could be
pre-scanned with handhelds, maybe?"

All kinds of corners will be cut to make this ridiculously daft scheme
appear feasible. I am especially worried about the eye scan. The
Government assures us that the iris scan is non-intrusive, but I have
read too many horror stories about maladjusted radiotherapy equipment
killing the patient that I feel less than sanguine about having
complete strangers, who may just be on minimum wage, shine lights in
my eyes.

MM

On Wed, 25 May 2005 23:19:28 GMT, EricP
wrote:

On Thu, 26 May 2005 00:11:50 +0100, "Mike" babbled
like a waterfall and said:


"raden" wrote in message
...
In message , MM

See my latest post in the uk.politics.misc thread, referring to
tonight's Channel 4 News. I think if the Govt continues like this, the
ID card will be laughed out of the chamber.

Well, apart from that, the finger in the air calculation that it would
take 50 years to get everyone's information made me chuckle

And that's assuming everybody was helpful and co-operated.

The retinal scan alone is expected to take 5 minutes if perfect
working is followed.

One thing we DO NOT want is a retinal scan! What is proposed is an
iris scan. A retinal scan is much more intrusive and far more
potentially dangerous if not carried out by true professionals (as
opposed to 'professionals' who have been given two days' training by a
duplicitous government in order to meet its targets). Retinal scans
shoot a low-intensity beam of light in to the eyeball and record the
pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

MM

"MM" wrote in message
...
On Thu, 26 May 2005 09:09:45 +0100, Mark
wrote:


On Wed, 25 May 2005 16:58:25 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
. ..
On Wed, 25 May 2005 12:52:12 +0100, "Mary Fisher"
wrote:

"Mark" wrote in message
m...

I seem to remember from an earier thread (correct me if I am wrong)
that you declined to specify your ethnic origin on a census form.
I'm
sure this is not a sign of wrongdoing, so why assume that anyone
else
doing the same does have something criminal to hide?

I'm not worried about disclosing that I'm a white Caucasian but on
most
forms that information is irrelevant (i.e. I'm making a political
objection)
and what's more the options given are not ethnic origins. I don't hide
the
information, I clearly state that the information is irrelevant. It's
never
been followed up so presumably it IS irrelevant. If I didn't fill in
the
'gender' box they'd be back.

I would argue that much of the information that we will be forced to
give for our ID Card will also be irrelevant

Such as?

All of it? ;-)

Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.


But the fact that discrimination is now illegal would help you win a case in
this situation, Mark.


I might to 'hide' the fact, i.e. not broadcast it, that I worked in
Germany for many years, if I were applying for a job with someone who
was known to be anti-German, for example (perhaps his parents' house
was bombed in the war and everyone except him was killed). Now, one
might argue that I shouldn't be considering such a post working with
people who still hold a grudge after 60 years, but the point is, I
have something to hide that might threaten my chances of employment if
it came out. Conversely, if I were to apply to an educational
establishment as mentor or classroom assistant in modern languages,
that fact might be all to my benefit and I should publicise it as much
as possible. So, as in much of life, it's horses for courses.

MM


But discrimination is illegal, so this wouldn't / shouldn't go toward you
not getting a job. The courts and lawyers would love this one.

MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

"Mark" wrote in message
...


Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

How do you know that I have.I keep sayng I haven't, do you think I'm
untruthful?

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

Well, haven't but do you really think that's going to be on an ID card?

Very many people seem to be anticipating the worst. It's just as easy to
anticipate the best :-)

Mary

Mark

On Thu, 26 May 2005 12:14:32 GMT, "BigWallop"
wrote:

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

But the fact that discrimination is now illegal would help you win a case in
this situation, Mark.

Maybe, if you have the time, money and knowledge to prove a case in
court. However I guess this would be difficult to do in practise.

But discrimination is illegal, so this wouldn't / shouldn't go toward you
not getting a job. The courts and lawyers would love this one.

There's more to discrimination that not getting a job.

Mark.

On Thu, 26 May 2005 14:07:09 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
.. .


Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

How do you know that I have.I keep sayng I haven't, do you think I'm
untruthful?

You did not enter your ethnic origin in your census form. Is that not
hiding information?

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

Well, haven't but do you really think that's going to be on an ID card?

Many of the items of information that will be stored by the ID card
system have been, and will continue to be, used as discrimination.

Very many people seem to be anticipating the worst. It's just as easy to
anticipate the best :-)

Please can I borrow your rose coloured spectacles? ;-)

Mark

On Thu, 26 May 2005 13:44:11 +0100, John Rumm
wrote:

MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

Yeah, and they've only found out about this NOW?!! Yet again I can
perceive caterwaulings from the bowels of Whitehall, screaming to EDS
and others: "But you didn't SAY it wasn't any good for some people!
And now we've paid you lots and lots of taxpayers' money!"

MM

"Mark" wrote in message
...
On Thu, 26 May 2005 14:07:09 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
. ..


Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

How do you know that I have.I keep sayng I haven't, do you think I'm
untruthful?

You did not enter your ethnic origin in your census form. Is that not
hiding information?

I've addressed this precise point in another reply. Others might tell you to
keep up, I'm far too polite.

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

Well, haven't but do you really think that's going to be on an ID card?

Many of the items of information that will be stored by the ID card
system have been, and will continue to be, used as discrimination.

Such as?

Very many people seem to be anticipating the worst. It's just as easy to
anticipate the best :-)

Please can I borrow your rose coloured spectacles? ;-)

I doubt that anything would work for your jaundiced view. But you have
solace in that you're not alone, rather, apparently, one of a majority.

There, does that make you feel better?

Mary

Mark

In message , John
Rumm writes
MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

That would be ... err ... terrorist types ?

pesky foreigners

The CH4 news poll of yesterday shows a significant swing away from those
wanting ID cards now

20% for

80% against

So people are slowly waking up

--
geoff

In message , Mary
Fisher writes

Please can I borrow your rose coloured spectacles? ;-)

I doubt that anything would work for your jaundiced view. But you have
solace in that you're not alone, rather, apparently, one of a majority.

There, does that make you feel better?

There is an RFD to set up an ID card news group on uk.net.news.announce
.. Sorry if this has been covered but I have not been following this
thread.

regards
--
Tim Lamb

"raden" wrote in message
...
In message , John
Rumm writes
MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

That would be ... err ... terrorist types ?

pesky foreigners

The CH4 news poll of yesterday shows a significant swing away from those
wanting ID cards now

20% for

80% against


35/65 on Radio 2 today. But still most against.

On Thu, 26 May 2005 19:47:31 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
.. .
On Thu, 26 May 2005 14:07:09 +0100, "Mary Fisher"
wrote:


"Mark" wrote in message
...


Maybe I didn't make my point very clear. I was just trying to counter
the argument "having something to hide is always indicative of
wrongdoing.". You (and many other people) have legitimate reasons for
'hiding' information.

How do you know that I have.I keep sayng I haven't, do you think I'm
untruthful?

You did not enter your ethnic origin in your census form. Is that not
hiding information?

I've addressed this precise point in another reply. Others might tell you to
keep up, I'm far too polite.

I've read your previous posts. You chose not to reveal some
information on your census form because, you say, it was irrevelant.
That's OK in your view. On the other hand, you say, that not wanting
to reveal personal information for an ID card is a sign of wrongdoing.
I don't see the difference myself.

Please don't misinterpret me. I am not saying that you are untruthful
or have done anything wrong or have anything criminal to conceal. I
am saying that other people who do not wish to divulge personal may
not either.

If you've ever been victim of arbitary discrimination, for example,
you may wish to 'hide' the feature of this discrimination for fear it
may happen again.

Well, haven't but do you really think that's going to be on an ID card?

Many of the items of information that will be stored by the ID card
system have been, and will continue to be, used as discrimination.

Such as?

Name, Nationality, Photograph to name but three.

Very many people seem to be anticipating the worst. It's just as easy to
anticipate the best :-)

Please can I borrow your rose coloured spectacles? ;-)

I doubt that anything would work for your jaundiced view. But you have
solace in that you're not alone, rather, apparently, one of a majority.

I would consider my view realistic rather than jaundiced. With the
case of ID cards we have to look at all aspects and effects of it,
not just its possible benefits before deciding whether we support it
or not.

There, does that make you feel better?

Should it?

Mark.

On Thu, 26 May 2005 23:16:15 +0100, "Mike" wrote:


"raden" wrote in message
...
In message , John
Rumm writes
MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

That would be ... err ... terrorist types ?

pesky foreigners

The CH4 news poll of yesterday shows a significant swing away from those
wanting ID cards now

20% for

80% against


35/65 on Radio 2 today. But still most against.

Overwhelmingly against with Any Questions audience last night
(Friday).

MM

Huge wrote:

And although I'm pleased that public opinion is coming around, I'm
saddened that it's apparently on cost rather than civil liberties
grounds.

I get the feeling more and more will be against as they begin to
understand more of what is involved on a practical level.

When they start sending out demands that you appear at a place and time
of their choosing to be scanned, and interviewed, all under threat of
prosecution if you fail to show. It will seem far more oppressive.

You can also see the fiasco (and another of the critical weaknesses of
the whole scheme) that will begin when you arrive at the interview and
need to prove who you are!

--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

John Rumm wrote:

You can also see the fiasco (and another of the critical weaknesses of
the whole scheme) that will begin when you arrive at the interview and
need to prove who you are!

You certainly can. I've only skimmed the full report of the enrolment
trial as yet, over at

http://www.homeoffice.gov.uk/docs4/U...ial_Report.pdf

But one little nugget leapt off the page at me: the time to *verify* the
biometrics is quite *amazingly* long. Verification is what we're all
going to be doing, multiple times a day - transport, payment, other
interactions with either officialdom or large-providerdom (as the Govt
are hoping that private companies will flock to use the Govt-backed ID
and pay a little transaction fee each time). We've all seen biometric
verification done in the movies; some of us have experienced them, at
trade shows and the like. The notion I think we all share is a glance up
at the camera (for facial or iris recognition), or popping a finger or
two onto a reader (for fingerprint), with a go/no-go signal inside a few
seconds.

This trial gave the typical time to verify as between 30 and 60 seconds!
More precisely, the median times for verification, of a
population-representative thousand-or-so volunteers, were (p.234,
Appendix B):
facial: 33s
iris: 49s
fingerprint: 56s

These are median figures, i.e. the delay expected by half the sample is
up to this figure, by the other half more than this figure. These
medians are shorter, i.e. more flattering to the Govt's case, than the
mean (what we usually call the 'average'), as they're not skewed by the
very long outer tail of the distribution. (In computing the averages,
they sliced off the longest time-to-verify for the 'top' 1% of all three
technologies, which were all starting at OVER 5 MINUTES. It's gonna be
great getting stuck behind one of those one-in-a-hundred people at the
Severn Bridge Corssing, innit... Maybe these figures give a more useful,
'whole-process' time. After all, you have to take off gloves/hat, look
straight at the camera for the face/iris ones, let it look again if you
moved, or a passing car cast a shadow or reflection (no, I'm NOT making
that up, it's in the report as the cause of many misses in the London
Passport Office setup).

The accuracy figures were also *way* below the manufacturer's sales
claims (well, dip *me* in chocolate and feed me to the sexual
(p)reference censoreds). These have been more widely reported - facial
recognision only 70%, iris much better at 96% - still *way* below Prof
Daugman's figures (illustrating, I guess, the gulf between lab
conditions and the world of more normal equipment operators!), and 86%
for fingerprints. Lower figures apply for participants with a variety of
disabilities, which this trial specifically recruited many of, as a
separate experimental group.

Nevertheless, this is the technlogy which David Blunkett assured us
would make identity theft 'impossible. Not just practically impossible:
impossible'. Snort.

More snorting in some left-leaning pinko-tinged rag (this week's New
Statesman) by some Zaba woman, btw. I *think* the web edition lets you
take a one-time peek at one article - it's
http://www.newstatesman.co.uk/200505300020. But what the site really
wants you to do is to pay for reading, a bargain quid for one whole issue.

Stefek

On 28 May 2005 11:28:59 GMT, (Huge) wrote:

MM writes:
On Thu, 26 May 2005 23:16:15 +0100, "Mike" wrote:


"raden" wrote in message
...
In message , John
Rumm writes
MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

That would be ... err ... terrorist types ?

pesky foreigners

The CH4 news poll of yesterday shows a significant swing away from those
wanting ID cards now

20% for

80% against


35/65 on Radio 2 today. But still most against.

Overwhelmingly against with Any Questions audience last night
(Friday).

You think this will make any difference? There were large numbers of
people against the Dunblane Gun Grab, the Iraqi War, the ban on hunting
and I'm sure there are other things I forgot, but it made not the
slightest difference to B'Liar.

And although I'm pleased that public opinion is coming around, I'm
saddened that it's apparently on cost rather than civil liberties
grounds.

"It's the economy, stupid!"

Seems like this saying holds true in so many ways.

However, the Government's spokespersons, in particular Hain and
Clarke, *still* claim overwhelming support! I heard this trotted out
at least half-a-dozen times in the last week. They must think if they
repeat the lie often enough it will become true. Their claim, by the
way, is based on a poll done ages ago, when the full ramifications of
the ID card were unknown by most ordinary members of the public,
which, I posit, is still the case today.

MM

On Sat, 28 May 2005 20:21:23 +0100, Stefek Zaba
wrote:

John Rumm wrote:

You can also see the fiasco (and another of the critical weaknesses of
the whole scheme) that will begin when you arrive at the interview and
need to prove who you are!

You certainly can. I've only skimmed the full report of the enrolment
trial as yet, over at

http://www.homeoffice.gov.uk/docs4/U...ial_Report.pdf

But one little nugget leapt off the page at me: the time to *verify* the
biometrics is quite *amazingly* long. Verification is what we're all
going to be doing, multiple times a day - transport, payment, other
interactions with either officialdom or large-providerdom (as the Govt
are hoping that private companies will flock to use the Govt-backed ID
and pay a little transaction fee each time). We've all seen biometric
verification done in the movies; some of us have experienced them, at
trade shows and the like. The notion I think we all share is a glance up
at the camera (for facial or iris recognition), or popping a finger or
two onto a reader (for fingerprint), with a go/no-go signal inside a few
seconds.

This trial gave the typical time to verify as between 30 and 60 seconds!
More precisely, the median times for verification, of a
population-representative thousand-or-so volunteers, were (p.234,
Appendix B):
facial: 33s
iris: 49s
fingerprint: 56s

These are median figures, i.e. the delay expected by half the sample is
up to this figure, by the other half more than this figure. These
medians are shorter, i.e. more flattering to the Govt's case, than the
mean (what we usually call the 'average'), as they're not skewed by the
very long outer tail of the distribution. (In computing the averages,
they sliced off the longest time-to-verify for the 'top' 1% of all three
technologies, which were all starting at OVER 5 MINUTES. It's gonna be
great getting stuck behind one of those one-in-a-hundred people at the
Severn Bridge Corssing, innit... Maybe these figures give a more useful,
'whole-process' time. After all, you have to take off gloves/hat, look
straight at the camera for the face/iris ones, let it look again if you
moved, or a passing car cast a shadow or reflection (no, I'm NOT making
that up, it's in the report as the cause of many misses in the London
Passport Office setup).

The accuracy figures were also *way* below the manufacturer's sales
claims (well, dip *me* in chocolate and feed me to the sexual
(p)reference censoreds). These have been more widely reported - facial
recognision only 70%, iris much better at 96% - still *way* below Prof
Daugman's figures (illustrating, I guess, the gulf between lab
conditions and the world of more normal equipment operators!), and 86%
for fingerprints. Lower figures apply for participants with a variety of
disabilities, which this trial specifically recruited many of, as a
separate experimental group.

Nevertheless, this is the technlogy which David Blunkett assured us
would make identity theft 'impossible. Not just practically impossible:
impossible'. Snort.

More snorting in some left-leaning pinko-tinged rag (this week's New
Statesman) by some Zaba woman, btw. I *think* the web edition lets you
take a one-time peek at one article - it's
http://www.newstatesman.co.uk/200505300020. But what the site really
wants you to do is to pay for reading, a bargain quid for one whole issue.

She a relative of yours?

By the way, I see the whole article, not just a 'tempter'. And I
haven't subscribed to the site. (May have *registered* at some point,
however.)

MM

Stefek Zaba wrote:

You certainly can. I've only skimmed the full report of the enrolment
trial as yet, over at

http://www.homeoffice.gov.uk/docs4/U...ial_Report.pdf

I will digest later.... ;-)

But one little nugget leapt off the page at me: the time to *verify* the
biometrics is quite *amazingly* long. Verification is what we're all
going to be doing, multiple times a day - transport, payment, other
interactions with either officialdom or large-providerdom (as the Govt
are hoping that private companies will flock to use the Govt-backed ID
and pay a little transaction fee each time). We've all seen biometric
verification done in the movies; some of us have experienced them, at
trade shows and the like. The notion I think we all share is a glance up
at the camera (for facial or iris recognition), or popping a finger or
two onto a reader (for fingerprint), with a go/no-go signal inside a few
seconds.

All this assumes that you have a willing subject at well, not
deliberately moving their finger, glancing away, blinking at just the
worng moment etc.

What about all the people who are asked to remove their contact lenses
so they can be scanned?

This trial gave the typical time to verify as between 30 and 60 seconds!
More precisely, the median times for verification, of a
population-representative thousand-or-so volunteers, were (p.234,
Appendix B):
facial: 33s
iris: 49s
fingerprint: 56s

That with a sample of 1000 subjects...?

Logic would suggest the best case scenario is that you can in some way
sort or order the biomentric data. In which case your search time will
increase in proportion to log(base2) #subjects.

So scaling from 50 secs for 1000 subjects, to 45m subjects we are
talking median recognition times of getting on for 25 mins! That is best
case and median. Get the feeling you might be late for work?

(p)reference censoreds). These have been more widely reported - facial
recognision only 70%, iris much better at 96% - still *way* below Prof
Daugman's figures (illustrating, I guess, the gulf between lab
conditions and the world of more normal equipment operators!), and 86%
for fingerprints. Lower figures apply for participants with a variety of
disabilities, which this trial specifically recruited many of, as a
separate experimental group.

The more worrying figure in a way is the 30, 4, and 14% false positive
rate those figures imply!

More snorting in some left-leaning pinko-tinged rag (this week's New
Statesman) by some Zaba woman, btw. I *think* the web edition lets you
take a one-time peek at one article - it's
http://www.newstatesman.co.uk/200505300020. But what the site really
wants you to do is to pay for reading, a bargain quid for one whole issue.

Seemed to be OK without whinging.... I am sure the Mozilla cookie
manager could keep it that way ;-)

--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

On Sat, 28 May 2005 19:23:19 +0100, John Rumm
wrote:


When they start sending out demands that you appear at a place and time
of their choosing to be scanned, and interviewed, all under threat of
prosecution if you fail to show. It will seem far more oppressive.


--
Peter Parry.
http://www.wpp.ltd.uk/

On Sat, 28 May 2005 19:23:19 +0100, John Rumm
wrote:

When they start sending out demands that you appear at a place and time
of their choosing to be scanned, and interviewed, all under threat of
prosecution if you fail to show. It will seem far more oppressive.

Apparently you are exempt if you are one or more of :-

Irish (including Irish "Traveller")
Scots
A visiting terrorist



--
Peter Parry.
http://www.wpp.ltd.uk/

MM wrote:

.... some Zaba woman, btw. I *think* the web edition lets you
take a one-time peek at one article - it's
http://www.newstatesman.co.uk/200505300020. ....

She a relative of yours?

Not sure whether spice typically get called 'relatives' ;-)

John Rumm wrote:

That with a sample of 1000 subjects...?

Logic would suggest the best case scenario is that you can in some way
sort or order the biomentric data. In which case your search time will
increase in proportion to log(base2) #subjects.

So scaling from 50 secs for 1000 subjects, to 45m subjects we are
talking median recognition times of getting on for 25 mins! That is best
case and median. Get the feeling you might be late for work?

There's little scaling suggested in the test results, which are in 3
chunks: about 2000 (I was wrong in the above saying it was 1000) people
they call the 'quota' group - this being demographically similar to the
whole UK population, i.e. similar proportions of age, gender, ethnic
background, and the like; around 7000 in the 'opportunistic' group -
non-preselected Members Of The Public, maybe including some civil
cervants at the trial sites, who were just asked if they'd like to
volunteer, and not quite 1000 specifically chosen people with
disabilities. They give the results of almost all the many measurements
separately for each of these 3 categories; median times for the quota
group (size 2000) differ very little from those for the opportunistic
group (size 7000), suggesting to me that the actual IT-system processing
time may be a small part of the overall time they're measuring. [But see
below for a more devastating 'explanation' of this time invariance!]

In any case, for most of the multi-dimensional 'best-match' pattern
recognition algorithms, search time is sublinear in the size of the
known datapoints - you do a nearest-neighbour calculation with a radius
related to your target false-accept/false-reject rates and the
population size; if you find 0 matches in that radius you say 'new
point', if you find 1 you assert it's that one, if you find 1 you say
'ambiguous' (and maybe rescan for differentiating features). For
Daugman's iris algorithm, a different approach is taken: they calculate
binary values for around 250? distinct features (each bit is just 1 or
0), the features being chosen to give a high degree of statistical
independence between different subjects. Do this right, and you should
get very high discrimination even for seriously big population sizes,
easy feature comparison, and essentially constant-time lookup. (John
Daugman's got plenty of writeups on the Web). At an hour-long
presentation of his I attended in December last year, he did stress the
need for decent cameras, reasonable lighting (among other things, you
need the pupil not to be too dilated - or you lose resolution of some of
the features; dunno how well iris recognition will work in the clubbing
scene, mind!), and working hard at enrolment to get good, repeatable
measurements to be the 'reference' point. Don't remember him mentioning
that spectacles were a problem, though they certainly seem to have been
in this trial - reflections + crappy camera again? He uses his control
of the patent on this technique to try to make sure manufacturers
implement the *whole* algorithm, properly, allowing for population size
properly for example; but the failure (non-recognition) rate on these
small populations is quite a lot higher than the figures I'd expect from
his presentation.

The more worrying figure in a way is the 30, 4, and 14% false positive
rate those figures imply!

Indeed, they're much more worrying (they're not 'false positive' but
'false negative' rates, mind: these are people who've just been enrolled
attempting to be recognised post-enrollment, so non-recognition means
the recognition sample was too different from the enrollment one to be
recognised as - but I'd heard about those already, while the extended
recognition times were the first surprise I fell across in a tired,
initial skim.

**BOGOSITY ALERT** **BOGOSITY ALERT** **BOGOSITY ALERT**
I've just tried to discover how close in time the 'verifications' were
to the 'enrollments'. I can't be entirely sure, but the description on
p.7 looks like the 'verifications' were done in the same half hour?
session as the enrollments. If so, these would ASTONISHINGLY FAVOURABLE
and UNREPRESENTATIVE circumstances! In real use, the biometrics are
captured during an enrollment session, and are then expected to be
verified months/years later. The writeup I refer to on p.7 says the
whole process went like this for each subject: 'Registration; Photograph
participant (head and shoulders); Record facial biometric; Record iris
biometric; Record fingerprint biometric; Record electronic signature;
Print card; Post-enrolment questionnaire; Verification;
Post-verification questionnaire.' - so it really looks like all happened
in one signature. To be fair to this trial's objectives, it says quite
clearly (p.5/section 1.1.2, Objectives): 'The goal of the UKPS iometrics
Enrolment Trial was to test the processes and record customer experience
and attitude during the recording and verification of facial, iris and
fingerprint biometrics, rather than test or develop the biometric
technology itself – it was not a technology trial.' What's even more
astonishing is to read a few sentences later that they were 'testing
fingerprint and iris biometrics for one-to-many identification and
testing facial, iris and fingerprint biometrics for one-to-one
verification' - so in the case of facial recognition certainly, and iris
and fingerprint probably - these 'verification rates' are NOT for 'which
recorded biometric is the one I'm being presented with most like' - the
tougher, identification problem; but the much simpler 'to within
appropriate likelihood of error, is the one I'm being presented with
sufficiently similar to THIS one'.

I'm truly gobsmacked, and hope I'm misreading due to the lateness of the
hour. If this trial is really saying that *verification* rates with
'meaningful' error thresholds under 'semi-careful' conditions (operators
being careful, participants being willing volunteers, ...) are this low
- then whole-population *identification*, on the basis of this study,
looks like a COMPLETE NON-STARTER.

This is so surprising a conclusion that I've tried to falsify my
assumptions by reading Appendix C, the 'Technology' section (just 5
pages). But they seem to confirm them: whole-DB-match was done only for
fingerprint *enrolment*, to check that similar prints hadn't already
been enrolled, while verification was limited to 1-to-1! For iris,
Daugman's encapsulation seems to mean that the trial was 'forced' to do
1-to-many - but note that there weren't 'many' even at the end of the
trial, and only the last enrollee was matched against all the others!

So it really looks as if this 'trial' was genuinely focused on 'how do
people feel about this stuff', as its Objectives and the vast bulk of
its analysis make clear. Therefore, ANY extrapolations of performance of
the biometric technology itself will be LUDICROUSLY OVER-OPTIMISTIC; and
any Minister doing so is at best extremely shoddily advised.

Gak. I feel a posting to No2Id coming on...

Stefek

In message , MM
writes
On Thu, 26 May 2005 23:16:15 +0100, "Mike" wrote:


"raden" wrote in message
...
In message , John
Rumm writes
MM wrote:


pattern of veins in the eye, whereas an iris scan 'merely' records the
unique patterns in the iris.

Unique and distinguishable if you are caucasian, European / north
American that is.... It seems they are ignoring the slight problem with
many other races where the scan is too indistinct to actually be any use.

That would be ... err ... terrorist types ?

pesky foreigners

The CH4 news poll of yesterday shows a significant swing away from those
wanting ID cards now

20% for

80% against


35/65 on Radio 2 today. But still most against.

Overwhelmingly against with Any Questions audience last night
(Friday).

And now this

http://observer.guardian.co.uk/polit...4,00.html?gusr
c=rss

....The LSE believes the government has grossly underestimated the cost
of the technology involved in making the system work. Last week the
government estimated the biometric card readers needed to scan the cards
would cost £250-£750. 'A more likely figure ... would be in the range of
£3,000 to £4,000 per unit,' the report suggests

The report also raises doubts about whether the government is right to
assume a 10-year life span for each card. 'All technical and scientific
literature indicates that biometric certainty diminishes over time, and
it is therefore likely that a biometric - particularly fingerprints and
facial features - will have to be re-scanned at least every five years.
This cost must be taken into account.' A further problem, which the
government appears not to have factored in, is 'refuseniks' - people who
will not co-operate. 'There is evidence that this population could
create a substantial additional cost burden. The administrative costs of
handling this group will be substantial,' the report states.

The LSE also questions the strain placed on the system by individuals
notifying a change in their personal circumstances, as they will be
required to do so by law...


--
geoff

In article ,
raden writes:

The report also raises doubts about whether the government is right to
assume a 10-year life span for each card. 'All technical and scientific
literature indicates that biometric certainty diminishes over time, and
it is therefore likely that a biometric - particularly fingerprints and
facial features - will have to be re-scanned at least every five years.

This sort of card, when carried around all the time, doesn't have
a 5 year life anyway. For starters, a significant number will have
been lost or stolen in that time, and for those that won't, a
significant number will have worn out through normal flexing,
abrashion, etc. Ask any company which issues such cards to their
employees (who only carry them 8 hours a day, 5 days a week).

--
Andrew Gabriel

Huge wrote:

That and the application of 25kV to the SmartMedia contacts.


I hope you're not suggesting that citizens should interfere with this
valuable piece of Crown property. Besides the moral horror that such an
action engenders in all upstanding citizens, you'd be commiting a
specific offence under the proposed Act, as far as I can see: under
Subsection 1 of Section 13, one is required to noitify the Authorities
if one knows, or has reason to suspect, that the card has been (among
other things) tampered with, damaged, or destroyed; failure to so notify
shall, under Subsection 6 of the same Section 13, render one liable to a
civil penalty not exceeding 1,000 notes.

So make sure you keep your card well away from that van der Graaf
generator, Huge!

Huge wrote:
That and the application of 25kV to the SmartMedia contacts.

It might not take that. Mobile phones will wipe a lot of magnetic
stripe cards, just by being carried next to them. I'd be interested to
see if the chips can withstand being 5mm from a transmitting mobile
phone for long periods.

Regards
Capitol

"Capitol" wrote in message
...


Huge wrote:
That and the application of 25kV to the SmartMedia contacts.

It might not take that. Mobile phones will wipe a lot of magnetic
stripe cards, just by being carried next to them. I'd be interested to
see if the chips can withstand being 5mm from a transmitting mobile
phone for long periods.

The prototypes survived that sort of thing just fine. After all the SIMM
card in your phone is about this far from the aerial all it's life.

But 25kV is pushing it. They can survive static at this sort of potential
but deliberate maltreatment will zap them.

Stefek Zaba wrote:


specific offence under the proposed Act, as far as I can see: under
Subsection 1 of Section 13, one is required to noitify the Authorities
if one knows, or has reason to suspect, that the card has been (among
other things) tampered with, damaged, or destroyed; failure to so notify
shall, under Subsection 6 of the same Section 13, render one liable to a
civil penalty not exceeding 1,000 notes.

"Hello, is that ID card central?"

"Yes Sir"

"Oh, right, I understand I am supposed to tell you when I think my card
may have been damaged or tampered with"

"Yes sir that is correct"

"Ah, OK, I think that my one is knackered then"

"What makes you think that Sir?"

"Well I hooked up those little gold bits to the mains again, and it
started smoking. Then the card sort of melted. Will I get into trouble?"

"No Sir, you only get into trouble if you fail to tell us"

;-)


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

I see today's estimate is that the system will cost £300 per user.

On Sun, 29 May 2005 22:48:36 +0100, Stefek Zaba
wrote:

Huge wrote:

That and the application of 25kV to the SmartMedia contacts.


I hope you're not suggesting that citizens should interfere with this
valuable piece of Crown property. Besides the moral horror that such an
action engenders in all upstanding citizens, you'd be commiting a
specific offence under the proposed Act, as far as I can see: under
Subsection 1 of Section 13, one is required to noitify the Authorities
if one knows, or has reason to suspect, that the card has been (among
other things) tampered with, damaged, or destroyed; failure to so notify
shall, under Subsection 6 of the same Section 13, render one liable to a
civil penalty not exceeding 1,000 notes.

So make sure you keep your card well away from that van der Graaf
generator, Huge!

So many people will be hit with unexpected fines that this alone will
make the system unworkable. Crikey, we've seen what a song and dance
the media (rightly) can make about some old dear who's been refused
operations four times. Imagine hundreds of thousands of voters
suddenly with a demand for a thousand quid on their doormats because
they dropped their card or placed it too close to static or something.

And will fines be levied if cards are stolen? If not, then no card
will be lost; all that are will be declared stolen. "I was holding it
in my hand and the mugger just took it, Herr Gauleiter!"

But again, none of the public knew about any of these little arcane
details when they gave their 'overwhelming' support that the
Government claims.

MM