Eircode discussion

clewbays

ManAboutCouch wrote: »

I think you're missing my point. You stated that you don't think anything other than a very vague age and sex breakdown is available at Small Area level. The link I posted shows that that's not true, an extensive set of census variables is available at that level.

The point I made is that complex and detailed analyses (your own words) cannot be made available at small area level because you quickly reach confidential data e.g. the one Lithuanian in the small area you gave a link to cannot be cross-classified with other socio-economic questions.

Small areas are excellent for mapping data for publishing as pdf files but not for making cross-tabulations. They are small enough to get a subtle palette of variation and they are consistent in size in terms of number of households.

plodder

ManAboutCouch wrote: »

Yes, I think so, it looks like it's to be filled in by the Enumerator rather than the resident.

I see the census form is online now. Here is the top of the front page



The space over on the right is for the eircode presumably. So, my guess is that the enumerators will have separate lists of small area code, address, eircode taken from the database and they will write in the details on each form.

The structure of the small area code is interesting.

2 digits for county
3 digits for EA (electoral area?)
4 digits for district
1 digit (SA small area?)

maybe the last digit is a check digit.

plodder

From other thread:

Boatmad wrote:

its a silly argument , because you can not solve the issue of large numbers of non unique addresses , without creating a pointer to a unique one.
I would argue that my address is not private, though my identify is. ( up to a point) , but its up to me to be careful in how I distribute both my address and my identify

the fact is as Mark Zuckerberg said, privacy is over , as he built a large wall around his house !!!

These discussions only become silly, when someone nitpicks over observations that are really quite obvious.

The discussion is useful even now, because there are still things that can be done to mitigate these problems, simply by being more open about them eg on the Eircode website pointing out the privacy implications of the finder.

BoatMad

plodder wrote: »

From other thread:


These discussions only become silly, when someone nitpicks over observations that are really quite obvious.

The discussion is useful even now, because there are still things that can be done to mitigate these problems, simply by being more open about them eg on the Eircode website pointing out the privacy implications of the finder.

The privacy argument is not a silly argument, the argument over unique Eircodes is .

The privacy argument of online data is however far bigger then Eircode

stoneill

I just had a courier phoning me looking for directions to my address - even though eircode was used.

marmurr1916

stoneill wrote: »

I just had a courier phoning me looking for directions to my address - even though eircode was used.

When they get to you, show them how they can find your house using the mobile version of the Eircode Finder website.

plodder

Posted by "you know who" on Twitter, but I thought I'd repeat it here. It's an example of exactly the kind of privacy glitch I wrote about before, which Eircode creates. It's a Europe wide survey with a standard form that includes a postcode. The authors of the survey obviously don't know that Eircodes are potentially personal information. So, RTE had to jump through hoops:

a) to warn that providing your postcode is the same as providing your full address

b) undertake to remove any postcodes that were provided as clearly they do not want the hassle of "minding" people's personal data, when it is supposed to be an anonymous survey.

Click on the image to see the letter RTE wrote about it.

https://twitter.com/loc8code/status/729665930456924160

BoatMad

Im not sure what the issue is , any postcode that is uniquely identifiable by way of address lookup, gives away your location ( I mean thats the point )

Loc8 is exactly the same

plodder

BoatMad wrote: »

Im not sure what the issue is , any postcode that is uniquely identifiable by way of address lookup, gives away your location ( I mean thats the point )

Loc8 is exactly the same

The issue would be the same for any postcode that identifies your exact location. It's a bigger problem for a non-hierarchical code (or not very hierarchical code). In any case, Eircode is the code we have, and the issue can't be dismissed because another hypothetical code also has the same issue.

BoatMad

plodder wrote: »

The issue would be the same for any postcode that identifies your exact location. It's a bigger problem for a non-hierarchical code (or not very hierarchical code). In any case, Eircode is the code we have, and the issue can't be dismissed because another hypothetical code also has the same issue.

its more an issue of awareness then anything else

GJG

plodder wrote: »

Posted by "you know who" on Twitter, but I thought I'd repeat it here. It's an example of exactly the kind of privacy glitch I wrote about before, which Eircode creates. It's a Europe wide survey with a standard form that includes a postcode. The authors of the survey obviously don't know that Eircodes are potentially personal information. So, RTE had to jump through hoops:

a) to warn that providing your postcode is the same as providing your full address

b) undertake to remove any postcodes that were provided as clearly they do not want the hassle of "minding" people's personal data, when it is supposed to be an anonymous survey.

Click on the image to see the letter RTE wrote about it.

https://twitter.com/loc8code/status/729665930456924160

This is certainly an issue. Not going to go into the whole Loc8 thing, obviously the issue would be identical if it was used.

The real point is the fact that these decisions shouldn't be left to be made on the hoof by different suppliers, the Data Protection Commissioner should be all over this issue and should have had codes of practice in place before Eircode - or any other postcode - was launched. There should be thought-out standard policies that smaller organisations can adopt, and an information campaign to inform data holders of best practices.

Instead we got vague hand-waving about there being data protection 'issues' without the DPC even being able to articulate what the issues are. Of course there are issues. Every information technology has data protection issues. It's the DPC's job to work out what they are, and advise solutions.

plodder

BoatMad wrote: »

its more an issue of awareness then anything else

Awareness is only one aspect of it. As you can see from this case, RTE had to go to some trouble to remove the postcodes from the returned surveys, (presumably) because they never wanted to be in possession of personal information.

ukoda

is everyone assuming here that the people who gave their eircode to RTE didn't want them to know their location,

i wonder if RTE had an "address" field, would they be required to put the warning:

IF YOU LIVE SOMEWHERE THAT HAS A UNIQUE ADDRESS PLEASE BE AWARE THAT GIVING US YOUR UNIQUE ADDRESS MEANS THAT WE WILL HAVE YOUR UNIQUE ADDRESS

:rolleyes:

plodder

ukoda wrote: »

is everyone assuming here that the people who gave their eircode to RTE didn't want them to know their location,

i wonder if RTE had an "address" field, would they be required to put the warning:

IF YOU LIVE SOMEWHERE THAT HAS A UNIQUE ADDRESS PLEASE BE AWARE THAT GIVING US YOUR UNIQUE ADDRESS MEANS THAT WE WILL HAVE YOUR UNIQUE ADDRESS

:rolleyes:

No, because everyone knows their address is unique. Not everyone knows that their postcode uniquely identifies their address. That might sound strange to someone who has been looking at (and defending) it for some time. But, genuinely, most people don't know or want to know any of these details.

ukoda

plodder wrote: »

No, because everyone knows their address is unique. Not everyone knows that their postcode uniquely identifies their address. That might sound strange to someone who has been looking at (and defending) it for some time. But, genuinely, most people don't know or want to know any of these details.

I absolutely defend the need for unique addressing and the solution for Ireland is a unique postcode because that works best for our unique mess of addressing

I just don't buy in to the hysteria of "there's huge data protection concerns with it"

Even in the example above, no harm was done to anyone by RTE. it made zero impact to their privacy, and yes some will argue a dodgy company might misuse the data in this scenario. But of course the counter arguement to that has always been, it's a small minority who might do this and if their intent is to mine data dubiously then they will always find a way even without eircode. And also, if the company is dodgy the public have an onus to be vigilant in not giving them any data in the first place.

BoatMad

Not everyone knows that their postcode uniquely identifies their address.

clearly , as I said , an awareness issue

plodder

So, here we are almost a year after Eircode was launched, and 18 weeks after former Minister for Communications Alex White said that Eircode support on google maps would be signed off 'in the next few weeks'.

As of today, there is no Eircode on satnavs, or google maps. A survey done by CBRE finds that 96% of the transport/logistics industry don't support it. The head of FTAI says 'Eircode is going nowhere'. As for google, Eircode might not be a priority internally, but there is no indication that any agreement has been reached between themselves and Eircode for support at any time in the future.

What does exist is an app from Autoaddress, which it has to be said, works quite well, but it is not at all clear how this app complies with Eircode's licensing terms, primarily because it is offered free of charge, yet there is per transaction cost associated with Eircode lookups. This calls into question the sustainability of the app as a free product which users can rely upon going forward.

What also exists is a proposal (which possibly originated from discussions here) of a free data set, in addition to the ECAD and ECAF that would be equivalent to the code point open dataset that exists for UK Postcodes (and is also free of charge). The transport industry seems to be favourably disposed to this idea. It puts Irish postcodes on the same kind of commercial footing as the UK postcode, with a basic free product, and a premium paid for product. Yet, some people here are still arguing against it, on the basis that the state should never invest in any infrastructure that the private sector can profit from unless users pay directly for it. This is odd because the state invests in roads and all manner of infrastructure that private enterprise profits from. In a lot of cases (as with roads), the cost of charging directly for usage can outweigh the benefits.

We've seen this already with Eircode, where the initial license model is widely acknowledged as unsuitable for navigation devices. It is way too restrictive and cumbersome. The confusion and doubt concerning google does not help either, as the google mapping platform has a much simpler license model and is free for many use cases. Potential users of Eircode are probably better off waiting to see what (if anything) emerges from this before committing to using ECAD.

In the meantime, probably the only people to benefit are those who have developed competing location code systems, and who do not want to see Eircode succeed.

oscarBravo

plodder wrote: »

In the meantime, probably the only people to benefit are those who have developed competing location code systems...

...and those of us who have posted repeatedly here how Eircodes are a godsend for our business.

ukoda

So here we are now almost a year after eircode has launched.

As of today, here's who is using it:

An post
Nightline
The Census
Facebook (for admin pages)
123.ie
Dominos.ie
The CRO
Electric Ireland
Imagine broadband
SUSI
Revenue
Dept of agriculture
PayPal (requesting it but not validating)
Power city

This is to name a few, the list is much longer

These companies have made public statements to say they will implement it:

DPD
Fastway
TomTom
Garmin

We have the first eircode app launched by Autoaddress

The FTAI have got their heads around the SAC's in eircode database and have said they are "very impressed with it"

Two thirds of respondents in the CBRE/KPMG supply chain confidence index 2016 declare eircode was positive for the industry and 73% of shippers said it was positive

Looks like eircode is useful to a very wide range of industries and we can only hope its adoption rate increases

plodder

Should Eircode be scrapped?: A closer look at the national postcode system

Hadn't seen this quote before:

So, what should happen next? If this system is not fit for the purpose outlined by Eircode themselves, should it be scrapped?

"Eircode should be “parked” as a unique property identifier, essentially becoming the PPSN for property. It should be treated in the same way as a PPSN: not publically displayed, but used on all Government (and utility-provider) correspondence with the individual or business. DCENR should then revisit the NPPB report, and roll out a national postcode consistent with its original research findings from 2006," says McDonnell.

Can't see that happening tbh.

Sam Russell

plodder wrote: »

Should Eircode be scrapped?: A closer look at the national postcode system

Hadn't seen this quote before:

So, what should happen next? If this system is not fit for the purpose outlined by Eircode themselves, should it be scrapped?

"Eircode should be “parked” as a unique property identifier, essentially becoming the PPSN for property. It should be treated in the same way as a PPSN: not publically displayed, but used on all Government (and utility-provider) correspondence with the individual or business. DCENR should then revisit the NPPB report, and roll out a national postcode consistent with its original research findings from 2006," says McDonnell.

Can't see that happening tbh.

Well, all the heavy lifting has been done in the design and mapping. All it needs is a recoding of the actual post code using the SACs and adjusting them to make sense of estuaries and other non-contiguous adjacent codes, and other discontinuities.

To base the original design on An Post's major postal towns, ignoring county boundaries, and using the original Dublin post codes was just daft. Even more so when you take into account An Post always maintained it did not need them and would not use them.

So, maybe. The controversy has not gone away you know. It would not cost €38 m to redo it.

byrnefm

If say, the code was to be consigned to being a Property # in effect, I wonder how many years it would take...

It is a good database ID for looking up addresses but for day to day use, not easy for the average person to remember or to guess where something is by simply looking up the Eircode like you could with a UK postcode, for example.

Don't get me wrong - I use mine, I just don't think it is the most user-friendly or memorable system.

oscarBravo

plodder wrote: »

Hadn't seen this quote before:

He also said "The bottom line is, there is literally no benefit in using it." In this he is, objectively, factually wrong. Which means the entire article is agenda-driven and therefore suspect.

byrnefm wrote: »

It is a good database ID for looking up addresses but for day to day use, not easy for the average person to remember or to guess where something is by simply looking up the Eircode like you could with a UK postcode, for example.

If I see an F23 postcode, I know it's in the Castlebar area. If I need more detail, I can look it up. I'm at a loss as to why people are so utterly hung up on being able to pinpoint a postcode without a lookup. It's not like looking it up is an insurmountable challenge.

plodder

oscarBravo wrote: »

He also said "The bottom line is, there is literally no benefit in using it." In this he is, objectively, factually wrong. Which means the entire article is agenda-driven and therefore suspect.

If I see an F23 postcode, I know it's in the Castlebar area. If I need more detail, I can look it up. I'm at a loss as to why people are so utterly hung up on being able to pinpoint a postcode without a lookup. It's not like looking it up is an insurmountable challenge.

The bottom line is cost. We could have developed a location code for free that didn't need any licensing or we could have released some simple to use subset of data for free.

Given that this has been pointed out numerous times before I can only guess you are suggesting that the Autoaddress app compensates for that somehow. In that case:

a) there are too many questions about the app's business model to be sure it will always be free

b) logistics companies will have their own software and systems and many would prefer to have access to the data rather than someone else's app.

c) there are many other uses for the same kind of information which the app doesn't help with at all

Releasing the underlying data as opposed to an app gives greater confidence also that they can't start charging for it in future. Apps can be switched off at a moment's notice.

byrnefm

One other thing I was wondering about .. one of the complaints about Eircode is that many people's addresses' counties change (I know it doesn't really). Why the need for the county to be included at all in the address if the Eircode is included? It's more granular than a county. It works in the UK .. the county is effectively optional.

I know people have already tried it with the Dublin district codes .. by forcing people to still write that with the Eircode just makes the code stand out a bit too much, as in, it doesn't appear to add value if included on an envelope since your post might get delayed if you leave out the Dublin district, despite including the Eircode.

oscarBravo

plodder wrote: »

The bottom line is cost.

If he had said that, we'd be having a different conversation. He didn't say that; he said something that's objectively untrue.

If he had even said that the benefits don't justify the cost, we'd be having a different conversation. But he didn't say that either. He said something that's objectively untrue instead.

ukoda

plodder wrote: »

We could have developed a location code for free that didn't need any licensing

No we couldn't have developed a location code for free. Developing things costs money. No code is free to develop.

eircodes design does not *need* a licence. It has one by choice. As does Loc8code who similarly make you pay a licence fee via in app purchase if you want more than 15 look ups.

plodder

ukoda wrote: »

No we couldn't have developed a location code for free. Developing things costs money. No code is free to develop.

Wrong. One of the earliest reports on the subject pointed out that simple geocodes (like loc8) are no more than a mathematical conversion of geo-coordinates into some alphanumeric alphabet. Anyone could have developed a code like that with less effort than was put into the Postcode board by its members (I assume for free).

What costs money are things like websites, and publicity. The state is prepared to fund these for all kinds of public functions. So, why not postcodes?

eircodes design does not *need* a licence. It has one by choice. As does Loc8code who similarly make you pay a licence fee via in app purchase if you want more than 15 look ups.

A geocode (not loc8) would have been completely license free, because there is no substantial IP that could be licensed.

That's not to say that a code like Eircode which has associated IP, does not have value. I am simply refuting your suggestion that the Eircode way was the only way. We could have developed a cheap/license free postcode. Anyone who suggests otherwise is being disingenuous and dishonest.

Oscar Bravo wrote:

If he had said that, we'd be having a different conversation. He didn't say that; he said something that's objectively untrue.

If he had even said that the benefits don't justify the cost, we'd be having a different conversation. But he didn't say that either. He said something that's objectively untrue instead.

So, he said Eircode is useless. So, what? That doesn't invalidate every other criticism made by him and others. Michael O'Leary spouts a lot of nonsense, and it doesn't invalidate everything he says about the transport business

oscarBravo

plodder wrote: »

So, he said Eircode is useless. So, what?

So, he's wrong. Either (a) he knows he's wrong, in which case he's lying, or (b) he doesn't know he's wrong, in which case he doesn't know what he's talking about.

Either way, I think it's safe to say that his opinion on Eircodes is suspect at best.

plodder

byrnefm wrote: »

One other thing I was wondering about .. one of the complaints about Eircode is that many people's addresses' counties change (I know it doesn't really). Why the need for the county to be included at all in the address if the Eircode is included? It's more granular than a county. It works in the UK .. the county is effectively optional.

I know people have already tried it with the Dublin district codes .. by forcing people to still write that with the Eircode just makes the code stand out a bit too much, as in, it doesn't appear to add value if included on an envelope since your post might get delayed if you leave out the Dublin district, despite including the Eircode.

If An Post "used" Eircodes in any meaningful sense then it would mean you could address a letter with a name and an eircode only, but they say you can't do that. You would also be able to drop parts of the address like the county.