National Postcodes to be introduced

Carawaystick

New data protection legislation to be rushed in for eircodes.

http://utv.ie/News/2015/02/03/New-postcode-legislation-will-protect-personal-data-28906

http://www.independent.ie/irish-news/news/government-approves-draft-legislation-on-eircodes-postcode-system-30961926.html

yuloni

This post has been deleted.

antoinolachtnai

BarryM wrote: »

It seems to me that many posters on here are assuming people will use the (seven digit ?) code. Even if they do how many will make errors, if they write it as opposed to having headed paper?? Thinking about it who uses headed paper any more?? Finger problems are probably biggest cause of errors online. Even the "signature" option which might provide some accuracy is not used by many email users.

Eircode decided to ignore the empirical ergonomics literature in the area of memorable codes and decided to adopt some psychobabble mumbo jumbo instead.

The best practice in the area is certainly Schraagen and Van Dongen, Designing a licence plate for memorability, Ergonomics, Vol. 48, No. 7, 10 June2005, 796–806 (see http://www.ncbi.nlm.nih.gov/pubmed/16076738 for abstract)

SPDUB

antoinolachtnai wrote: »

They might be able to do that, but it is going to cost 5 euros or more per delivery. You can't expect to get that degree of attention to detail for 60 cents an envelope.

That makes the whole project pretty expensive.

Well it would be if you just pluck a figure out of the air like 5 euros ( which you have given no basis for )

Getting postmen to do surveying work is a pretty awkward way to go about it.

What are talking about postmen doing surveying work because even if any Geo maps had to be taken out of the office , instead of just at the postman's delivery bench , it wouldn't be surveying work to see that a particular eircode goes to a particular building .

If the postman can't access it, I wouldn't think it's really a delivery point.

The usual problem is that the postman gets there before it opens and while An Post asks for a post box if that is the case some times that can take a while

antoinolachtnai

SPDUB wrote: »

Well it would be if you just pluck a figure out of the air like 5 euros ( which you have given no basis for )

It is my estimate of the minimum reasonable cost for the special handling that would be required. The closest list product that An Post has is registered delivery. Registered delivery costs 6 euros a go retail. Eircode might conceivably get a discount. I think that in reality An Post would have to charge more. It's a lot of work to deliver unusual items like this, without names, with a high level of accuracy.

SPDUB

antoinolachtnai wrote: »

It is my estimate of the minimum reasonable cost for the special handling that would be required. The closest list product that An Post has is registered delivery. Registered delivery costs 6 euros a go retail. Eircode might conceivably get a discount. I think that in reality An Post would have to charge more. It's a lot of work to deliver unusual items like this, without names, with a high level of accuracy.

Well unless they are planning to scan every single eircode letter into the registered letter computer then a large aspect of the cost of a registered letter doesn't apply

a65b2cd

antoinolachtnai wrote: »

You could do this for far less than what An Post are going to get paid for this escapade (about 20 million, 10 million for Geodirectory and another 10 million for deliveries).

This is a good example of the mis-information being put into the public domain. The actual cost details for dissemination were given at the Oireachtas hearing on 19th November:

"A total of €1 million is to be spent on dissemination, advertising, communications and design."

Typically the mis-information is accompanied by emotive language like escapade.

plodder

antoinolachtnai wrote: »

Eircode decided to ignore the empirical ergonomics literature in the area of memorable codes and decided to adopt some psychobabble mumbo jumbo instead.

The best practice in the area is certainly Schraagen and Van Dongen, Designing a licence plate for memorability, Ergonomics, Vol. 48, No. 7, 10 June2005, 796–806 (see http://www.ncbi.nlm.nih.gov/pubmed/16076738 for abstract)

Good memorability of licence plates is important in those cases where licence plates are viewed for a brief period of time and the information is essential for police investigations. The purpose of the current study was to design a new Dutch licence plate that could be remembered well

That's the sensible way to design a system like this. You commission studies to test various aspects of the design. You don't just put it out for tender, choose it on best price and then hope for the best.

The study is very interesting. It seems the number of alternations between numbers and letters is what affects memorability the most. It makes perfect sense as well and doesn't seem to have been considered in the design of Eircode.

One completely separate thing that occurred to me the other day, is for some privacy sensitive applications, like verifying credit card numbers online (or in shops) you could mandate that only 5 characters of an Eircode should be required. The first five characters (including the first two random ones) would be quite useful as an identification token, but useless for any other purpose. That is good from a privacy perspective and it would mean that if shops ask you for your full Eircode, you can rightly question why they need it.

antoinolachtnai

a65b2cd wrote: »

This is a good example of the mis-information being put into the public domain. The actual cost details for dissemination were given at the Oireachtas hearing on 19th November:

"A total of €1 million is to be spent on dissemination, advertising, communications and design."

Typically the mis-information is accompanied by emotive language like escapade.

Firstly there is no final price for this work. I have an email here from Mr Duggan of Eircode which plainly states that there is no final agreement on the issue.

Secondly, the figure you are stating is clearly wrong. One million euros is 45 cents an item. There is just no way to design, print and deliver 2.2 million individually addressed high quality print items with any degree of accuracy for one million euros.

It cannot be done. It is a fantasy. It is going to cost a lot more than that.

You are the one spreading mis-information here.

ukoda

antoinolachtnai wrote: »

Firstly there is no final price for this work. I have an email here from Mr Duggan of Eircode which plainly states that there is no final agreement on the issue.

Secondly, the figure you are stating is clearly wrong. One million euros is 45 cents an item. There is just no way to design, print and deliver 2.2 million individually addressed high quality print items with any degree of accuracy for one million euros.

It cannot be done. It is a fantasy. It is going to cost a lot more than that.

You are the one spreading mis-information here.

It was my understanding the €1 million was purely marketing spend? Print, tv, radio etc

The cost of delivering the code would be seperate I would have thought? And be part of the operationalising budget

my3cents

I actually don't see why people need to know their own Eircode initially. If all the utility companies for gas, water, electric and state bodies use Eircodes on every address then it won't be long before everyone has had free notification of their Eircode which I expect they will be able to check on line anyway. Useful as it will be its not like there's a date by which if you don't know your Eircode the world is going to end.

antoinolachtnai

It is a good idea in principle, but the Department tells me that this is not the way it is being done for non-unique addresses. The Department of Agriculture (for example) will only know a farmer's address (assuming it is non-unique) when it is supplied by the farmer to the Department of Agriculture.

my3cents

antoinolachtnai wrote: »

It is a good idea in principle, but the Department tells me that this is not the way it is being done for non-unique addresses. The Department of Agriculture (for example) will only know a farmer's address (assuming it is non-unique) when it is supplied by the farmer to the Department of Agriculture.

So does that infer unique addresses will get a letter/card with a first class stamp on it and non unique addresses will have a bit more spent on them?

HardenendMan

Good to see the debate has moved on from the code design :-)

The rollout can be simple. Print house will sequence leaflets based on post man's delivery route. This info is part of the address database. He ensures to go his official route without using discretion by taking shortcuts etc.

Carawaystick

How does eircode know there's a new house built between ordering the print run and the delivery run? or does it leave an out-by-one error in the eircodes delivered?
Eircode say their files are updated quarterly.
Eircode say their codes are random so there's no way to detect the eircode going to the house with address
Rock Little,
Arklow,
Co. Wicklow
is different from what it should be
or that the next house, with address
Rock Little,
Arklow,
Co. Wicklow

has the correct eircode.

ld also believe a Post don't prescribe the delivery routes on rural dead ends tightly enough to prevent a postman delivering to all the houses in order of passing them on a bothrín or a different postman serving all the houses o the left before turning around and serving all the houses on the left coming back up the road.

HardenendMan

Carawaystick wrote: »

How does eircode know there's a new house built between ordering the print run and the delivery run? or does it leave an out-by-one error in the eircodes delivered?
Eircode say their files are updated quarterly.

The Eircode database is an extract from An Post's geo directory. So they know of new builds because An Post update their database. There is always going to be a time lag. A new house could potentially have to wait a couple of months before getting their code... but sure moving into a new house leaves for lots of small things like this to be organised! (Bills, internet etc.)

Carawaystick wrote: »

Eircode say their codes are random so there's no way to detect the eircode going to the house with address
Rock Little,
Arklow,
Co. Wicklow
is different from what it should be
or that the next house, with address
Rock Little,
Arklow,
Co. Wicklow

has the correct eircode.

There is a way. It is based on the entries in the aforementioned geo directory database. This database has GPS coordinates for every house. So, in your example above, they can differentiate the houses by the coordinates. Randomisation helps here. You cannot end up at the wrong house down the road because the Eircode will be fairly different.

Carawaystick wrote: »

ld also believe a Post don't prescribe the delivery routes on rural dead ends tightly enough to prevent a postman delivering to all the houses in order of passing them on a bothrín or a different postman serving all the houses o the left before turning around and serving all the houses on the left coming back up the road.

I cannot say for sure, but standard practice in address databases that are used for route optimisation is that EVERY address gets assigned to a route (postman) and gets a sequence along that route. The sequence is obviously assigned based on the results of some route optimisation process. Of course the postman in a rural area might have his own way of walking a route, especially when he doesn't have an item for certain addresses. But he would know well what his "official" route is. Jeez you could even give him a map print out on the day of Eircode leaflet delivery just in case he is forgetful!

Maybe others on here can correct me if I am over simplifying things. Although I think a lot of participants on this thread should stop trying to over complicate things with respect to postcodes!

Carawaystick

Your post made certain assumptions. I questioned them.

If there is a timelag between a home being built and eircode generating an eircode to be delivered on an eircode printed letter/message, the every new house on a route in the country will knock all following houses out.


If the postman has to check the gps location of each delivery, there will be extra costs and hassle involved, as Antoin has pointed out.

HardenendMan

Carawaystick wrote: »

Your post made certain assumptions. I questioned them.

No problem, question away!

Carawaystick wrote: »

If there is a timelag between a home being built and eircode generating an eircode to be delivered on an eircode printed letter/message, the every new house on a route in the country will knock all following houses out.

Yep, you are right - this can be a tricky problem. A solution would be to give the postman an "exclusion" list - a list of addresses that are not contained within the stack of Eircode leaflets. He just has to skip these. This is easy for urban areas (house numbers, street names etc), for rural areas (non-uniques) he will know the route so intimately that there will be little ambiguity as to what house on the lane is a new build in the last 3 months (worst case).

Maybe that won't cover all situations but there will always be exceptions for this type of task. I am sure there will be a central "support office" that they can call with questions/clarifications on the addresses on the exclusion list. But let's bear in mind that this will effect a tiny percentage of Eircoded addresses!

Carawaystick wrote: »

If the postman has to check the gps location of each delivery, there will be extra costs and hassle involved, as Antoin has pointed out.

Nah they won't check a GPS. That is over-complicating things. He will plod along the route, with an exclusion list (I am aware this is just my idea), which will have at most 20 points on it (big developments can be treated as one exclusion), and just have it in his head to check the list when he gets to the new house (which he has seen been build from greenfield site over the past months!).

In saying all that, I really think it is harsh to assume the rollout will be a mess when none of us know what the exact plan is. I am not accusing you of this, just talking generally from what I have seen. I have given simple ideas based on having limited knowledge. I may be wrong, but my argument is that this rollout is not rocket science given the data and knowledge/experience available.

ukoda

plodder wrote: »

One completely separate thing that occurred to me the other day, is for some privacy sensitive applications, like verifying credit card numbers online (or in shops) you could mandate that only 5 characters of an Eircode should be required. The first five characters (including the first two random ones) would be quite useful as an identification token, but useless for any other purpose. That is good from a privacy perspective and it would mean that if shops ask you for your full Eircode, you can rightly question why they need it.

This is not the US. We have 'chip and pin' not 'chip and signature' there is no need for an Irish retailer to ask for your address either in part or full for the purpose of security when using your credit or debit card in store.

If they ask, it's purely for marketing, and they can ask for that now with or without an ericode. And if it's online you have to opt in or out to marketing. If it's for security of transaction online then they would need your full address, altho very very few online retailers use address as security check.

a65b2cd

Carawaystick wrote: »

If there is a timelag between a home being built and eircode generating an eircode to be delivered on an eircode printed letter/message, the every new house on a route in the country will knock all following houses out.

I doubt that a house can be built faster than the eircodes will be disseminated? The postman or postwoman will have had plenty of time to notice a bit of new construction activity. What will be of more interest is whether an eircode can be generated early enough for new builds so as to be useful for delivering construction material.

Liam Duggan said a three weeks delivery period in June: http://www.newstalk.com/reader/47.301/40490/0/

yuloni

This post has been deleted.

SpaceTime

ukoda wrote: »

This is not the US. We have 'chip and pin' not 'chip and signature' there is no need for an Irish retailer to ask for your address either in part or full for the purpose of security when using your credit or debit card in store.

If they ask, it's purely for marketing, and they can ask for that now with or without an ericode. And if it's online you have to opt in or out to marketing. If it's for security of transaction online then they would need your full address, altho very very few online retailers use address as security check.

The US banks dragged their feet for years on chip and PIN. It's only now (slowly) rolling out there. The whole use of zip codes for extra security was a bit of a crude 'hack' onto the system and works because zips could be keyed into a PIN pad.

You might as well verify DOB or a registered phone number. It's also managed to make non-US cards next to impossible to use at some services in the US as it's non-standard thus it's 'broken' the whole concept of universal payment cards.

You'll never see that happening here as Chip & PIN eliminates any need for it and Verified by Visa and MasterCard SecureCode have largely eliminated such messy and insecure systems online. Anyone could be aware of your postcode/zipcode/eircode if they had access to an envelope, your address, were a family member etc. It's totally useless security on a credit card!

plodder

ukoda wrote: »

This is not the US. We have 'chip and pin' not 'chip and signature' there is no need for an Irish retailer to ask for your address either in part or full for the purpose of security when using your credit or debit card in store.

If they ask, it's purely for marketing, and they can ask for that now with or without an ericode. And if it's online you have to opt in or out to marketing. If it's for security of transaction online then they would need your full address, altho very very few online retailers use address as security check.

No, they wouldn't need your full address. The credit card processor would send your Eircode (or part of it) to your bank, who would say, yay or nay it's correct or not.

The 3D Secure system fulfills the same function, but I understand it's not too popular. I've had bad experiences with it myself, and at the end of the day, it's yet another password that people have to remember.

ukoda

plodder wrote: »

No, they wouldn't need your full address. The credit card processor would send your Eircode (or part of it) to your bank, who would say, yay or nay it's correct or not.

The 3D Secure system fulfills the same function, but I understand it's not too popular. I've had bad experiences with it myself, and at the end of the day, it's yet another password that people have to remember.

As I said. In store there's no need for any address info to be taken. We have chip and pin

If it's online, and they decide to use ericode to verify bank details. Then you suggest that you give half of it in the billing address field to verify your bank details?....but then give all of it in the delivery address field?

You've over engineered a solution of giving half an eircode to a problem that doesn't exisit.

ukoda

yuloni wrote: »

Never gonna happen. eircode state they will release updates to the database quarterly. They would then be relying on the likes of Garmin, TomTom, Nokia, Google et al to apply and distribute these same quarterly updates to their services as they are released. This doesn't happen at present and while many of these services have plentiful purchased data contained within, it is years old in many cases

At least we've moved on from "it's never going to be on a sat nav" to "it will be, but will take a while for updates" .....progress

plodder

ukoda wrote: »

As I said. In store there's no need for any address info to be taken. We have chip and pin

If it's online, and they decide to use ericode to verify bank details. Then you suggest that you give half of it in the billing address field to verify your bank details?....but then give all of it in the delivery address field?

You've over engineered a solution of giving half an eircode to a problem that doesn't exisit.

As I said in the post where I raised the issue, it concerns cases where you are buying something on line that doesn't get delivered by post, eg anything downloadable like tickets, software or media, the issue was if Eircodes were to be used for verifying credit cards (because chip & pin can't be used obviously), that disclosing a full Eircode reveals your full address, which might be a privacy concern. However, it then occurred to me that you wouldn't have to disclose the full Eircode. The first five characters would be quite effective for that purpose. It's an advantage of the structure they have designed, not a disadvantage.

ukoda

plodder wrote: »

As I said in the post where I raised the issue, it concerns cases where you are buying something on line that doesn't get delivered by post, eg anything downloadable like tickets, software or media, the issue was if Eircodes were to be used for verifying credit cards (because chip & pin can't be used obviously), that disclosing a full Eircode reveals your full address, which might be a privacy concern. However, it then occurred to me that you wouldn't have to disclose the full Eircode. The first five characters would be quite effective for that purpose. It's an advantage of the structure they have designed, not a disadvantage.

Have you ever bought anything online that only asked for half or part of your billing address with your card details??! Man this just doesn't happen, they will want and should want your full billing address including your postcode even if the item purchased is for electronic delivery or no delivery.

The problem you're trying to solve doesn't exist. They will not use eircodes online to verify card holder details. They will either want your full address. No address. Or they will use verified by visa or mastercard 3D secure.

What your basically saying is: I trust this online company enough to give them my credit card details (card number, expiry date, name on card, CVV No.) but I don't trust them enough so I only want to give them half my eircode because I'm concerned with privacy. There's no emoji for this situation.

Impetus

plodder wrote: »

As I said in the post where I raised the issue, it concerns cases where you are buying something on line that doesn't get delivered by post, eg anything downloadable like tickets, software or media, the issue was if Eircodes were to be used for verifying credit cards (because chip & pin can't be used obviously), that disclosing a full Eircode reveals your full address, which might be a privacy concern. However, it then occurred to me that you wouldn't have to disclose the full Eircode. The first five characters would be quite effective for that purpose. It's an advantage of the structure they have designed, not a disadvantage.

Don't be naive. If Eircode ever becomes a reality, and that is a big IF, big data users will want the full code. Pretend security. In reality marketing analysis and spam. And to sell your information to other companies that don't compete with them, but would like your full info. The Eirccode would present an ideal platform to glue all the bits and pieces of data about you together to build a profile.

As it stands, for payment card verification applications, all processors are happy with a four or five digit code - which is equivalent to the postal district number in antiquated An Post terminology.

In the USA where there is very limited EMV card use, the customer at a gas station with a card issued in the US will be asked to enter their 5 digit Zip code. Chances are someone who stole a card does not know the zip code of the card billing address. In Europe, when you use a card to tank-up, the system requires your card PIN, which is a lot more secure. The same PIN does not infiltrate your data privacy.

If you live in 90210 in the US and get pick pocketed or drop your card, chances are high that 90210 is a thief's key to a tank of fuel. If you live in Europe, and the same issue arises, the thief has 3 chances to guess your 4 to 6 digit PIN.

High resolution postcodes are not a security solution..... they are a security risk by acting as database join points among several tables of stolen data.

ukoda

Impetus wrote: »

High resolution postcodes are not a security solution..... they are a security risk by acting as database join points among several tables of stolen data.

Can you explain exactly what you mean by this "join in a database" thing?

If my data is stolen from a database they will have all my details to commit fraud? Where does this "join tables" come into it? What exactly do you mean? Can you provide an example of this and indicate where eircode has added an additional risk by being part of my file on a system that has been compromised?

Impetus

ukoda wrote: »

Can you explain exactly what you mean by this "join in a database" thing?

If my data is stolen from a database they will have all my details to commit fraud? Where does this "join tables" come into it? What exactly do you mean? Can you provide an example of this and indicate where eircode had added an additional risk by being part of my file on a system that had been compromised?

This is not a database learning site. Suggest you do some googling - and perhaps have a look at this

http://resources.arcgis.com/en/help/main/10.1/index.html#//005s0000002n000000

A join is a common denominator that allows one to combine tables of data, acquired or stolen from different sources.