Eircode - its implemetation (merged)

The disastrous dimension to Eircodes is that An Post does not use them even though they received substantial money to amend their address reading software. A consequence of not having to use them on letter post is that they are not being used by probably 99% plus of persons and are not embedded in all public sector databases. Hence they are evolving into being used mainly as a coded form of X/Y coordinates and as a kind of index for looking up copies of the Geodirectory and ECAD databases. This only realises a small part of their potential. If they were used and captured by default then national databases would be hugely more valuable e.g. researchers could track the geographical pattern in the outbreak of diseases or duplication in registers could be more easily identified.

If the upfront decision included a knowledge that An Post would not use them then I think LOC8 codes would have been a better choice.

A job half done.

Skedaddle wrote: »

The CSO's statistics gathering needs are irrelevant to Eircode.

This isn't about the CSO at all. Their needs are catered for.

It's meant to be a stand alone system that uniquely identifies addresses. You couldn't keep changing people's postal codes to link up with whatever the CSO decides to do year to year. The codes are still useful to it and you can still break them down into smaller area and use them in lots of different ways.

With respect, the census is held no more than once every five years. They don't change things every year. The data that has been presented so far, over estimates the number of postcodes that would have changed. We need accurate information to discuss this point.

My view is it is that there was no perfect solution to this and Eircode is doing a reasonable job.

Agreed. There is no perfect solution and it's doing a reasonable job in some respects. I'm still concerned about the privacy aspect. For what it's worth I think a postcode should be anonymous, and what eircode is should be called a location code. Eircodes are basically geographic cookies, and it's going to be interesting to see where they turn up over time. If you want both the anonymous postcode and the specific location code to be integrated, then a hierarchical code is the only way to do it.

Had they attempted to rollout totally hierarchical postal codes that defined smaller and smaller areas viably in the code, you would have immediately turned it political and people would have been demanding code changes to suit address snobbery or refusing to use them at all.

The amount of fuss that went on about Dublin 6W years ago shows just how ridiculous these things can get.

Dublin 6W happened at the end of the Dept. of P&T era. So, it was inherently political. The process of even getting a phone line had politicians all over it. Different era.

I accept there would be some controversy whipped up by the media no doubt, but not the amount suggested.

You also get huge arguments over telephone area codes in the UK. https://www.mirror.co.uk/news/uk-news/furious-900000-housebuyers-outraged-after-10162798

Phone numbers are a slightly different issue with modern technology, but anyone who asks to get their UK postcode changed for the same snobbery reason gets short shrift.

( Due to the technology historically used in the British telephone network, area codes were often very small and only covered a single town. That resulted in all sorts of arguments like the above when towns grew into neighbouring codes. It's different in big British cities, and also here due to a history of different technologies being used. I won't bore you with the details, but it meant more you could have a lot more exchanges in a given area code than in the original uk systems).

Overall, I think Eircode works and you'll see the volume of post with the codes on them ramp up significantly as utilities and banks start integrating it. That's what will ultimately drive wider awareness.

Also it's already being used for utilities, like checking for OpenEir and Siro rollout of fibre services. I've even used it with Dominos pizza instead of entering my address.

There's no doubt we can argue until the cows come home, the relative merits/disadvantages of the two ways of doing it.

But, one thing is clear to me. Having decided to go the route of using an opaque, unstructured code, the decision should have been to make a significant, useful chunk of the data free for use. By all means, charge for the address file, like in the UK, but why should people be nickel-and-dimed in Ireland to do things you can do with other countries postcodes for free? The logical answer to that is to provide a database for free of all Eircodes linked to their small area code at the very least. That would be roughly equivalent to what you can do with the Uk postcode for free.

plodder wrote: »

So long as anyone you give your eircode to, and anyone they give it to, understands that it identifies your address, then there won't be a problem. The trouble is that most (countries') postcodes don't identify addresses. They are anonymous and generally not handled as carefully as addresses. So, organisations/companies with reasonable privacy policies might still share data based on postcodes without (to their knowledge) compromising anyone's privacy. They'll have to make sure that Irish postcodes are either removed or truncated to the routing key before sharing any such data.

Why?

Addresses are already out there.

Google Maps, electoral register, Geodirectory, using your eyes on a street...

echat wrote: »

1. I haven't seen any public data on Eircode take-up rates, can you please post a reference to where you saw them.

2. The main public sector databases were encoded in the first half of 2015 before the launch of Eircodes. A unique address is required to assign an Eircode. So around 60% of Revenue and Social Protection clients probably got Eircodes and the rate for Department of Agriculture clients was probably less than 15%. So in the last 30 months have those rates increased or decreased?

3. The routing key is useless given that An Post are not requiring it. The only saving grace is that they used the Dublin postal districts.

4. Is this what was expected 30 months on for probably around €30 million.

Eircode is used over 1 million times a week. You can watch them being entered on websites in real time here. Government departments and Local Authorities are integrating Eircode into systems and processes as IT systems are changed. This saves on implementation costs.

echat wrote: »

1. I haven't seen any public data on Eircode take-up rates, can you please post a reference to where you saw them.

2. The main public sector databases were encoded in the first half of 2015 before the launch of Eircodes. A unique address is required to assign an Eircode. So around 60% of Revenue and Social Protection clients probably got Eircodes and the rate for Department of Agriculture clients was probably less than 15%. So in the last 30 months have those rates increased or decreased?

3. The routing key is useless given that An Post are not requiring it. The only saving grace is that they used the Dublin postal districts.

4. Is this what was expected 30 months on for probably around €30 million.

I didn't say anything about take up rates, what you explicitly said was that An Post doesn't use them. We've had posters who are heavily involved in the industry state that they do and provide information to back them up. You then came up with a figure of 'probably 99%' of people don't use them. On what basis are you presenting this estimate? And AFAIK the €30M is the cost over (IIRC) 10 years not 30 months as stated by PDVerse and others in the past.

TheChizler wrote: »

I didn't say anything about take up rates, what you explicitly said was that An Post doesn't use them. We've had posters who are heavily involved in the industry state that they do and provide information to back them up. You then came up with a figure of 'probably 99%' of people don't use them. On what basis are you presenting this estimate? And AFAIK the €30M is the cost over (IIRC) 10 years not 30 months as stated by PDVerse and others in the past.

The basic idea of a postcode is for people to write them on an envelope so the postal body can more efficiently deliver post. You seem to be suggesting that An Post is using them internally in some way? They stated many times that they don't need them because they had already developed a system that managed without them.

I am basing the 99% as my estimate of what % of persons sending post are writing an Eircode as part of the address and I have asked a number of times if anyone has any actual usage rates.

The 30 months is the period between 13th July 2015 when they were launched and 13th January 2018. Yes all of 30 months ago.

The €30 million has already been spent much of it before the launch.

Sam Russell wrote: »

The point about Eircodes are they are not designed to be useful for people to use.

If the Bank gets an Eircode from a bill, that is fine but I was pointing out that the utility knows the location of their meter so do not need you to tell them. For non-unique addresses, it is nearly impossible to find out the Eircode without asking the addressee for it, which is not always possible. It is not something that people make public in any way, like nailing it to their gate post or including it in a return address, but that might change.

If An Post differentiated on postal rates favouring those that use it, it would become very common to see it on envelopes.

How do you mean they aren't designed to be useful for people to use? How are you defining useful? I'm not trying to critise - I'm genuinely wondering.

I think they are extremely useful. Over Christmas I had a few places to go where I'd never been before. Everytime I asked for an Eircode and was given it straight away. It never sounded like anyone had to check, or look i up. Also, made navigating an absolute breeze. Also had a handyman come over a few months ago. Just gave him my Eircode (nothing else) and he found me without any trouble. For years I've had to guide people on the phone, which is a right pain.

Also, regarding the seeming randomness of them - this prevents a 'desirable postcode' effect, which I guarantee it would have affected urban centers like Dublin.

Sam Russell wrote: »

If the Bank gets an Eircode from a bill, that is fine but I was pointing out that the utility knows the location of their meter so do not need you to tell them. For non-unique addresses, it is nearly impossible to find out the Eircode without asking the addressee for it, which is not always possible. It is not something that people make public in any way, like nailing it to their gate post or including it in a return address, but that might change.

This is a common belief, but at the time of Eircode launch it certainly wasn't true. People believe that ESB Networks have the exact location of their meter, and thus can easily identify the Eircode. They don't. Usually they have a sub-station location that feeds a number of houses, this is especially the case in rural areas for non-unique addresses. I'm not an expert in this area, and I have no knowledge of what has changed since July 2015 or whether the roll out of smart metering will change this (hopefully it will).

Eircode makes non-unique addresses unique, in the same manner that a house number on a street makes an address unique. If someone only gives you their street address there isn't any way to find out the house number by looking at the list of addresses or looking at the locations on a map (assuming you've never visited) unless they tell you their house number. Similarly for rural non-unique addresses, the person needs to provide their Eircode. This isn't a design flaw, its the solution to a problem.

Sam Russell wrote: »

Useful for people means useful unaided. If I give someone my address, they find my house no bother.

This isn't correct. No one holds the countries addresses in their head. You either reference a map or ask someone directions. No one could find my house in my estate unaided. Most people reference online maps, e.g. Google Maps for this task. In this way providing an Eircode or a unique address are equivalent. Eircode solves the non-unique address problem without having to change addresses. This wasn't a missed opportunity, it was a design goal.

Just on a piece of actual implementation, Costa now can locate your nearest Costa coffee shop if you enter your Eircode on their website. How useful that is is debatable but shows companies are becoming more aware of it and happy to integrate it.

Eircodes will be like every initiative in this country, like leap cards, uproar at first and within a couple of years we will be using them without thinking with no thanks for the people that introduced it