National Postcodes to be introduced

plodder

Leonard Shelby wrote: »

The first half facilitates manual sortation when the code can't be read by OCR. The second half will be sorted by automation, so no hierarchy required. Hierarchy in the second half of the code would increase complexity and cost, delay roll-out and and affect adoption. Why do that when it isn't required?

How would it increase complexity or cost? An Post could pretend that it's random as proposed currently. The fact that it wouldn't be random makes no difference to their systems.

It can't really delay rollout either, since all (or nearly all) the work has been done in the Geodirectory already. Every address is already assigned to a small area.

What is and isn't 'required' is very much what all this debate is about. The system has been designed to meet the requirements of some stakeholders who were involved, but not everyone with an interest in it.

clewbays

Enduro wrote: »

there is no chance of any DPC anywhere taking that on as a DP issue.

Other countries don't have the problem that Ireland has with a very high level of non-unique addresses. If the DPC does not sort out the current breaches of personal privacy for 1-in-3 Irish households, because their address is not unique, then who will?

recedite

Bray Head wrote: »

Can you cite a source for your claim that new legislation will be needed? I haven't seen this argument made elsewhere.

That seems to be the view of the Data Protection Commissioner, and the matter has not been resolved yet AFAIK;

[FONT=Arial, sans-serif]This Office was consulted by the Department of Communications Energy and Natural Resources, in relation to a unique, seven character postcode to be allocated to every home in the country in 2015. We had previously engaged with the Department on this issue in 2006 and again in 2010 when we expressed a serious concern that a public database linking a code to a single unit residential address could be considered as being personal data of the occupants of that dwelling. [/FONT]
[FONT=Arial, sans-serif]In the Irish context, a person’s home address is an important part of their identity and is the second most important piece of personal information to verify a person’s identity. Furthermore, in the case of a family home, a postcode could link many related individuals in the course of their daily activities. [/FONT]
[FONT=Arial, sans-serif]In essence the unique seven character postcode goes beyond what an “address” is because, through the use of modern technology and “Big Data”, it can be easily assimilated into any sort of electronic device or dataset which could in turn be used for any purpose, ranging from State services to commercial exploitation. [/FONT]


[FONT=Arial, sans-serif]In this regard, we expressed the concern that such datasets which would be verified by this postcode could have the potential for the ready identification of sensitive information about individuals, examples of which would be to identify specific localities that have patterns of crime or illness. This serious concern has since turned into a reality with the Minister’s announcement on the 8th of October 2013 that Cabinet had agreed to the rollout of the unique seven digit character code to every letter box in the State by 2015.[/FONT]
[FONT=Arial, sans-serif]A consortium headed by Capita Ireland has been engaged to develop, implement and operate the new postcode and under the new system, Ireland will be the first country in the world to have a public database of unique identifiers for all properties that will assist citizens, public bodies and business to locate every individual household in the State. This Office is unaware as to how this Consortium in conjunction with the Department will ensure that the Individual citizen’s fundamental right to data protection (and Privacy) will be safeguarded into the future by the use of this postcode. In particular whether these safeguards will be statutorily ringfenced such as proposed to be done in relation to unique seven digit medical identifiers as proposed by the Individual Health Identifier Bill 2013 or how both public and private bodies will be in compliance with the Data Protection Act, in the use of this Postcode. We have made enquires from the Department to this effect and will continue to do so pending complete clarification as to the manner in which the postcode system will operate in compliance with the Data Protection Acts. [/FONT]

http://www.dataprotection.ie/docimages/documents/Annual%20Report%202013.pdf
source

Of course, it wouldn't be the first time that politicians have been given plenty of advance warning about the legal need for new legislation; warnings which went completely unheeded.

clewbays

There are a lot fewer one person households than there are non-unique households and many of them are probably renting such as students and young workers in bedsits. Are you more worried about the possibility that someone (maybe their next-door neighbour) will realise they are living alone or the people who currently don't have any privacy because service delivery companies have to ask the whole locality for directions!

antoinolachtnai

Leonard Shelby wrote: »

The first half facilitates manual sortation when the code can't be read by OCR. The second half will be sorted by automation, so no hierarchy required. Hierarchy in the second half of the code would increase complexity and cost, delay roll-out and and affect adoption. Why do that when it isn't required?

If the second half were hierarchical, that would also facilitate even greater manual sortation when the code can't read by OCR.

What is so complex about using the 4th and 5th character to contain an identifier for the Small Area? How would it delay roll-out? How would it affect adoption other than positively?

The current random system is extremely difficult and complex to roll out. There is no plan on how to do this accurately. It will cost millions of euros and months or years to communicate and check the correctness of the communication to householders of every single code. If there are any mistakes, the quality of the whole thing will jeopardised.

It would have been much better to follow best international practice.

Leonard Shelby

antoinolachtnai wrote: »

If the second half were hierarchical, that would also facilitate even greater manual sortation when the code can't read by OCR.

What is so complex about using the 4th and 5th character to contain an identifier for the Small Area? How would it delay roll-out? How would it affect adoption other than positively?

The current random system is extremely difficult and complex to roll out. There is no plan on how to do this accurately. It will cost millions of euros and months or years to communicate and check the correctness of the communication to householders of every single code. If there are any mistakes, the quality of the whole thing will jeopardised.

It would have been much better to follow best international practice.

If you added hierarchy into the code format it would need to be changed when new builds occur, as is the case in the UK. The Eircode design ensures that postcodes don't need to change.

If you used the 4th or 5th character to identify a small area then you will have identified a small area, which isn't how post is sorted. When small areas change every five years then eircodes would have to change. The small area is stored in the eircode database and is available via a simple lookup, no need to build it into the code.

Any hierarchy affects adoption. There would be a public grouping of houses, which would be used by estate agents, insurance companies, etc. As soon as the public realised this there would be lobbying to "ensure we aren't grouped with the wrong households". Look at the UK for constant examples of this.

An Post will be delivering eircodes to all houses. Just because you don't know how they are doing it doesn't give you free reign to suggest timelines and costs that are entirely fanciful.

If you can point to a best international practice postcode example for another country with 35% non-unique addresses please do.

Sam Russell

Solve the unique addresses first.

tvc15

Sam Russell wrote: »

Solve the unique addresses first.

I think this needs a separate thread, Eircodes solve the non unique issue pretty much as well as any post code will.

Creating a new address structure for the country is a completely separate issue in a lot of ways and this is a discussion on postcodes

plodder

Leonard Shelby wrote: »

If you added hierarchy into the code format it would need to be changed when new builds occur, as is the case in the UK. The Eircode design ensures that postcodes don't need to change.

If you used the 4th or 5th character to identify a small area then you will have identified a small area, which isn't how post is sorted.

The sortation argument is bogus. If An Post can sort mail using randomly assigned codes, then they can just as easily (without making any changes to systems) sort with a hierarchical code.

When small areas change every five years then eircodes would have to change. The small area is stored in the eircode database and is available via a simple lookup, no need to build it into the code.

Who says they'd change every five years? As population increases in certain areas you will get new small areas being created, and in most cases the new areas would be for new properties. There is always a possibility of existing codes to change, but I assume this can happen with the routing keys also. It'd be a brave person who would predict that no postcode would ever change.

Any hierarchy affects adoption. There would be a public grouping of houses, which would be used by estate agents, insurance companies, etc. As soon as the public realised this there would be lobbying to "ensure we aren't grouped with the wrong households". Look at the UK for constant examples of this.

In other words, what you're saying is that a hierarchical code is actually useful and people like estate agents and insurance companies will use it.

yuloni

This post has been deleted.

my3cents

yuloni wrote: »

This post has been deleted.

Reads like another Loc8 pres release.

yuloni

This post has been deleted.

Leonard Shelby

plodder wrote: »

Who says they'd change every five years? As population increases in certain areas you will get new small areas being created, and in most cases the new areas would be for new properties. There is always a possibility of existing codes to change, but I assume this can happen with the routing keys also. It'd be a brave person who would predict that no postcode would ever change.

In other words, what you're saying is that a hierarchical code is actually useful and people like estate agents and insurance companies will use it.

Estate agents would abuse it. Insurance companies don't need it as they have a building level postcode but we've seen from the noise created by certain international FTAI members that some companies would rather perpetuate current inefficient processes rather than invest in Ireland. Best not to give them the option.

Sam Russell

We need a post code, but not a randomised one. Nor one that requires a computer with internet access,nor one that requires fees to access it.

We need one like the Swiss have. If we asked them nicely, do you think they would tell us how to copy it?

ukoda

Sam Russell wrote: »

We need a post code, but not a randomised one. Nor one that requires a computer with internet access,nor one that requires fees to access it.

We need one like the Swiss have. If we asked them nicely, do you think they would tell us how to copy it?

The Swiss postcode is pretty useless for navigation and certainly wouldn't work here with non unique addresses. The Swiss postcode is exactly that...purely a postcode that was designed to deliver mail and follows the postal routes. It's pretty outdated now and is certainly not what I'd call a modern design.

Sam Russell

It is no more useless than Eircode. Eircode has no navigational aspects to it as it is a random code. The navigational elements are bolted on.

ukoda

Sam Russell wrote: »

It is no more useless than Eircode. Eircode has no navigational aspects to it as it is a random code. The navigational elements are bolted on.

It's tied to a precise geo code. Whether that's a database or not doesn't matter.

antoinolachtnai

Leonard Shelby wrote: »

If you added hierarchy into the code format it would need to be changed when new builds occur, as is the case in the UK. The Eircode design ensures that postcodes don't need to change.

If you used the 4th or 5th character to identify a small area then you will have identified a small area, which isn't how post is sorted. When small areas change every five years then eircodes would have to change.

Small areas don't change every five years. Even if a small area were subdivided, this wouldn't necessarily have an impact.

But since you bring it up, what about the much bigger issue of when An Post postal towns are changed, as towns expand and new roads are put in?

The small area is stored in the eircode database and is available via a simple lookup, no need to build it into the code.

For a fee, or under very restrictive conditions. The fee looks set to be far in excess of the fee in a similar country with a similar number of houses and a similar building-level code (which is hierarchical). Singapore charges 122 euros for this database. The plan here is to charge far far more.

Any hierarchy affects adoption.

Yes, it makes it much easier to use and allows the postcode to be opened up. Many more people would use it.

Of course you will say that this is supposition on my part, since there is no evidence, since no one in the world has ever been crazy enough to roll out a non-hierarchical postcode before.

There would be a public grouping of houses, which would be used by estate agents, insurance companies, etc. As soon as the public realised this there would be lobbying to "ensure we aren't grouped with the wrong households". Look at the UK for constant examples of this.

I can't see how they can complain about being in the 'wrong' small area.

I see 2 such cases in the UK in the last year, not 'constant examples'.

If grouping of houses is such an issue, then there is a much bigger problem with the higher level part of the code. According to the documentation, there are significant areas of Dublin where there is a postcode that is simply not known or used. Areas identified are The Ward, Baily, Howth, Portmarnock, Loughlinstown, Kilternan, Newcastle, Saggart and Rathcoole. The population of this area is vast.

There is also a vast area of Dublin that may expect to have a Dublin postcode, but won't. If public opinion about postcodes is really so important, these represent a much more significant scale of problem.

The objection of the public will be much more general, i.e., that the system just makes very little sense in the light of well established administrative boundaries in Ireland. Any logic that applies to the first part of the code is inevitably contradicted in the second part.

An Post will be delivering eircodes to all houses. Just because you don't know how they are doing it doesn't give you free reign to suggest timelines and costs that are entirely fanciful.

An Post are to be contracted to deliver eircode 'items' to all homes. It is a matter for Capita and the Department to ensure that this id done to an adequate standard. I have asked them for their plan for doing this and they have no plan.

It sounds like you have different information. Could you please explain to us how An Post are going to deliver items with no addressee name given to 800,000 non-unique address, and to do this with an accuracy of 99.9999 percent, given that their more usual overall accuracy rate (including non-unique, and with a addressee name given) is more like 98.1 percent, and the regulator's target is only 99.5 percent?

If you can point to a best international practice postcode example for another country with 35% non-unique addresses please do.

Is Ireland really so different from every single other country in the world, that it requires a completely different solution from all the others?

Since you asked, and it is not particularly relevant, London is the obvious example of a postcode originally introduced to deal with non-unique addresses.

spurious

It's a long thread and I don't have the time to read it all, could someone in one sentence say what the problem with Loc8 codes is? They work fine on my GPS thing. Were there some Irish shenanigans involved in the system chosen?

plodder

spurious wrote: »

It's a long thread and I don't have the time to read it all, could someone in one sentence say what the problem with Loc8 codes is? They work fine on my GPS thing. Were there some Irish shenanigans involved in the system chosen?

Loc8 was treated unfairly I would say in that they weren't allowed to tender for pretty spurious reasons. But, the whole idea of using a tender for deciding something technical like this was ridiculous to begin with.

In any case however, I don't think Loc8 (or similar kinds of code) would meet the requirements of providing a unique property identifier. Loc8 isn't able to allocate different codes for multiple apartments in the same building since it is basically a conversion of geo-coordinates into something readable, and all units in the same building have the same geo-location.

SpaceTime

The live access to a database can be fun in remote areas where you can struggle for 3G/4G especially on the two younger networks.

SpaceTime

plodder wrote: »

Loc8 was treated unfairly I would say in that they weren't allowed to tender for pretty spurious reasons. But, the whole idea of using a tender for deciding something technical like this was ridiculous to begin with.

In any case however, I don't think Loc8 (or similar kinds of code) would meet the requirements of providing a unique property identifier. Loc8 isn't able to allocate different codes for multiple apartments in the same building since it is basically a conversion of geo-coordinates into something readable, and all units in the same building have the same geo-location.

Multiple apartments in any other system have the same address and are very easy figure out as each one has a number. About 4 to 6% of the Irish population lives in apartments. Something like 30% live in himts with vague, verbose addresses.

Bray Head

SpaceTime wrote: »

The live access to a database can be fun in remote areas where you can struggle for 3G/4G especially on the two younger networks.

This just isn't a valid criticism of eircode.

There will be a gap in the market for an app that has the database downloaded. The database will be quite small.

For example maps.me is a maps app that is based on openstreetmap. You can download a country in advance of travelling there. It helps you navigate when abroad and saves a large amount in roaming.

Bray Head

SpaceTime wrote: »

Multiple apartments in any other system have the same address and are very easy figure out as each one has a number. About 4 to 6% of the Irish population lives in apartments. Something like 30% live in himts (sic) with vague, verbose addresses.

You would then have a database with some unique and some non-unique postcodes. Not just for apartments but houses split in multiple parts with mysterious addresses like 'Rear of', 'Annex', '36b', etc.

If you are going to design a postcode system that is mainly unique you might as well make it entirely unique. Otherwise it's far, far less useful for a big user like a utility company.

plodder

Bray Head wrote: »

This just isn't a valid criticism of eircode.

There will be a gap in the market for an app that has the database downloaded. The database will be quite small.

For example maps.me is a maps app that is based on openstreetmap. You can download a country in advance of travelling there. It helps you navigate when abroad and saves a large amount in roaming.

Someone made the valid point recently, that it will be highly risky allowing software vendors to bundle "the database" on devices like iphones and android devices. It just isn't possible to keep IP secure on these devices.

If Eircode were structured hierarchically ......(here we go again) then you could possibly allow a free version of the database to be used offline that only contains the geo-locations of the small areas rather than each individual address. Apps might fallback to a dataset like this when they are offline (or have poor signal etc).

SpaceTime

Bray Head wrote: »

You would then have a database with some unique and some non-unique postcodes. Not just for apartments but houses split in multiple parts with mysterious addresses like 'Rear of', 'Annex', '36b', etc.

If you are going to design a postcode system that is mainly unique you might as well make it entirely unique. Otherwise it's far, far less useful for a big user like a utility company.

Most places have addresses like

10 Main Street, Unit 5
Tallaght
Dublin X12 YZZQ

Ireland has

Business Name
Poncyname House,
Main Street
Tallaght
Dublin 24

We actually need to reform addressing here entirely. The local authorities and An Post let the whole thing grow out of control.

Effectively, there is no system. That's a major, major problem.

You can't even fit Irish addresses on a standard shipping lable half the time. They run to 5 lines.

my3cents

SpaceTime wrote: »

Most places have addresses like

10 Main Street, Unit 5
Tallaght
Dublin X12 YZZQ

Ireland has

Business Name
Poncyname House,
Main Street
Tallaght
Dublin 24

We actually need to reform addressing here entirely. The local authorities and An Post let the whole thing grow out of control.

Effectively, there is no system. That's a major, major problem.

You can't even fit Irish addresses on a standard shipping lable half the time. They run to 5 lines.

All the rural addresses around me (north and east) go

Householders Name
Townland
Parish ???
County

We have a poncy name up in 4 inch high letters by the gate so we also add the house name but that wouldn't be the norm. Parish by the way is what it says it is even though they don't exist anymore and corresponds to the area the postman delivers to. Next door would be a different address altogether as they can use

Householders Name
Townland
Town
County

because their postman comes out from the town.

So its all about giving the postal service a hint as to where to route the mail from the address used.

Couriers have a real problem because the Parish is huge and unless you know what it refers to you'd think it was a village 4 miles from us :rolleyes: however if they bother to look up the townland they get within half a mile no problem.

My point is the addressing system is such a mess there is no real way to reform it.

Sheldons Brain

my3cents wrote: »

All the rural addresses around me (north and east) go

Householders Name
Townland
Parish ???
County

We have a poncy name up in 4 inch high letters by the gate so we also add the house name but that wouldn't be the norm. Parish by the way is what it says it is even though they don't exist anymore and corresponds to the area the postman delivers to.

It would be pretty unusual for someone to put a Parish in an address. That line of the address is the delivery office.

my3cents

Sheldons Brain wrote: »

It would be pretty unusual for someone to put a Parish in an address. That line of the address is the delivery office.

Correct it was once the delivery office but that office closed years ago and in the last year the delivery office has moved again. Still coincides with the old Parish boundaries and its the way some people refer to it. In other rural areas I've seen VIA xxxxx where xxxxx is the delivery office to make sure mail is routed correctly.

Sheldons Brain

my3cents wrote: »

Correct it was once the delivery office but that office closed years ago and in the last year the delivery office has moved again. Still coincides with the old Parish boundaries and its the way some people refer to it. In other rural areas I've seen VIA xxxxx where xxxxx is the delivery office to make sure mail is routed correctly.

There is no need to change your address because a delivery office moves!