Government introduces Postcodes in Ireland..?

turgon wrote: »

Lets look at my phone number (+35321487****) in parts, off the top of my head

+353 - nothing there to indicate Ireland
2 - nothing there to indicate Cork County
14 - once again, no hint to Cork City and outlying areas
8 - nothing says to me this means west of the City
7 - no hint that this could mean Ballincollig

Yet I still remember my phone number easily, do you?? You dont need to have the number/letter relate to areas - to do so would only increase expense, you can remember it easily already.

Not sure you are comparing like with like there. First of all, numbers are easier to remember than alphanumeric codes anyway.

A better comparison would be old style (pre 1987 car regs). Were they easier to remember and more useful than the present system? Which system do people prefer? I would wager that people prefer the present one, because the information in the code is more meaningful. It is very easy to remember the year, and county, then all you really have to think about is the 4 or 5 digits after that.

I dont think you really grasp how good this system is in that It does not really on say 'C' for Cork. Because of this you dont need to have a database on every GPS you get. Which means you dont have to pay a yearly subscription whatsoever, and also means your GPS is up to date every second you use it, with the exception of the map which is just an image representation and not totally necessary.

I think I grasp pretty well what the pros and cons of various kinds of system are. For instance, a system that uses a county code, and then a geographical co-ordinate within each county, would meet all the needs you mention above, ie. since county boundaries don't change, then your GPS would never need to be updated.

I think it's understandable that people have latched onto Gary's system because it's the only proposal that is in the public domain right now (apart from latitude/longitude, ITM etc.) But it would be wrong not to consider other possibilities.

ob wrote: »

Is this your personal opinion, or has there been studies to prove it?

I'm not aware of any studies, but it has been my personal experience. To me, it makes sense, since there are 26 different letters, but only 10 digits. The human brain is set up to remember words quite well, but a word is really recognised as a complete symbol in itself, whereas random collections of letters are not recognised as such.

garydubh wrote: »

Ok good - so you have no problem with
Cork W5K 59VN or
C-W5K 59VN
- which is already dealt with here: http://www.gpsireland.ie/index.php?option=com_content&task=view&id=47&Itemid=79&limit=1&limitstart=1 and earlier in this thread....

In our detailed spec for PON codes we define a County Code prefix as an optional element as per the link above and we include 2 character codes for Northern Counties also.

I do have a problem with redundant information in a code. It creates problems for software (and people) that have to deal with it. What if someone says their code is D-W5K 59VN?

Rather than trying to convince people that your code "does everything", you should just accept that like all kinds of design, it involves compromise, and argue for the compromises that you have made.

brim4brim. If you think the thread is going round in circles, then don't bother reading it. You're not contributing a whole lot anyway.

Personally, I think the best way forward for the government would be to define three or four different systems (including PONC), write up a detailed document describing the costs and benefits of each. Then, ask the public which one they prefer. If all other things are equal, it might come down, to a matter of preference.

[edit]Actually, I agree with brim4brim, the thread is going round in circles. This is my last post on it. Gary can keep bumping it every few weeks if he wants to. I won't intervene....

Article on Postcodes from The Sunday Business Post (7th Dec)

Postcode system could save public bodies €22 million

07 December 2008
By Niamh Connolly, Political Correspondent
The introduction of a national postcode system could mean benefits of €22 million for public bodies, according to a new report.

Ireland is the only country in the EU that does not use national postcodes, and the Minister for Communications, Energy and Natural Resources, Eamon Ryan, plans to bring proposals on the issue to government shortly.

A report by PA Consulting, which evaluated the monetary benefits of national postcodes at the request of the minister, estimated that it would save public bodies, including the emergency services, €22 million and would create efficiencies in all areas of social and economic planning.

Ireland’s postal sector is due to liberalise fully on January 1, and postcodes would form an important piece of infrastructure for competition, according to the report.

Cost savings were identified in a number of government departments, including Revenue and the Department of the Environment, along with Health, and Social and Family Affairs.

A postcode system would facilitate cross-departmental sharing of public data and information dramatically cutting waste and duplication.

The country’s rapid heterogeneous population growth also increases the need for efficient database based on postcodes reducing inefficient service delivery and infrastructural planning.

In addition, postcodes are seen as vital to efficient spatial planning and aiding health research, education, housing social care and employment integration.

The business sector has said that a postcode system would lead to increased efficiencies, while the insurance sector has estimated that it would result in annual savings of around €40 million by improving their risk management assessments.

In the voluntary sector, the charity Breastcheck said it relied on accurate postal services for efficiencies in its own service.

The Irish Exporters Association previously estimated that the lack of postcodes added up to 30 per cent to An Post’s sorting and routing costs.

But An Post opposes a national postcode, saying it is an unnecessary expenditure that may not be widely used.

Former communications minister Noel Dempsey aimed to introduce postcodes by next month, but this was deferred to allow for further consideration of the economic benefits. The National Postcode Project Board has said that the cost of implementing and promoting a national postcode system would be an estimated €15 million.

Annual maintenance costs for a postcode management licence holder are estimated at about €2.5 million, but the minister said last week that ongoing costs would be covered by income generated by the eventual licence holder.

Previous suggestions were for multiple numeric and alphabetical postcodes, similar to those used in other parts of Europe.

These might include the one and two-character county codes currently used in vehicle registration plates, but Conradh na Gaeilge wants a postcode that recognises both languages.

© Thomas Crosbie Media, 2008

and a letter from The Irish Times (9th Dec)

Adopting postcodes
Tue, Dec 09, 2008


Madam, - Ireland currently is the only country in the EU that does not have a postcode system. However, it has been reported that the Minister for Communications, Eamon Ryan, is to bring proposals to the Dáil shortly with regard to the introduction of postcodes in Ireland.

The naming convention applied to postcodes by our European partners often consists of a series of letters and numbers that frequently have little real meaning.

When Mr Ryan makes a final decision with regard to postcodes we should use our "last mover" status to our advantage and introduce 21st century postcodes that use GPS location codes.

This could be done by using more user friendly shortened four-digit longitude and four- digit latitude codes that indicate the general area.

This will help business and consumer deliveries to better locate their consumer, avoid delays and aid productivity. - Yours, etc,

CIARÁN TANNAM,
Asgard Road,
Dublin 2.

© 2008 The Irish Times

It is very hard to imagine a useful postcode which provided a unique number for each house without an address database. There would be no way to validate the postcode against the address. There would also be no way to take an address (even an urban address) and figure out accurately what its postcode would be.

It is very hard to see how ordinary people could accurately determine even their own postcodes without the aid of a national database of some sort.

It would also be pretty much impossible to figure out how to sort mail to deliver it to the address based on postcode, without the database. 'Rating' parcels, a key issue for international carriers of various sorts would also be difficult, unless the code was either hierarchically structured, or there was a database.

Having a database is actually not a big deal. There are a number of national address databases in existence in the private and public sectors, and it is a matter of gradually merging them together over a period of time. Many of these databases have common reference points, so this is very feasible.

The big players in the direct mail business will tell you that they don't need a postcode, that they can do everything they need with GeoDirectory. This is more or less true. In practice, the businesses that lose out are the local and medium sized ones that aren't big enough to bring in a fancy DM agency. Certainly, direct mail in Ireland is a lot less sophisticated than in the UK.

But would it make a difference to the amount of direct mail in Ireland? Probably some difference, but not that much. The reason is that the direct mail revolution is over. The world of direct marketing has moved on to the Intenet for a wide variety of reasons.

The reason why you get so many unaddressed leaflets is complicated, and has more to do with An Post's IR situation than anything else.

Our laws on the privacy of electoral lists are tighter than in the UK and this would certainly slow down 'new' direct mail. What you would see, however, is carefully targeted leaflets coming in with your regular mail (ESB bills, magazines, eircom, etc.)

There is one further issue. It is different to the old style direct marketing. It is the 'postcoding' process that Dunn Humby/Tesco carry out - the demographic profiling of relatively small groups of homes -. My own view is that certain elements of what DH are doing are wrong and that these practices should be prohibiited in law if they are not already illegal.

There is an explanation of 'postcoding' here (written by me) - http://www.eire.com/2007/02/17/tesco-loyalty-cards-privacy-what-does-it-all-mean/.

Having postcodes or not having them will not ultimately make any real difference to the ability of DunnHumby to apply their model to Ireland.

Using web-based maps and satnavs to survey the country is just not going to cut it. A satnav or web-based map will just not give the level of accuracy required to determine lattitude and longitude. Gary's scheme goes down to 4m squares, but Google maps are only accurate to around 10m at best, and in rural areas, can be 100m or more out of kilter. A code accurate to only 100m is better than nothing, but not great.

Handheld or vehicle GPS's realistically have an accuracy of around 10m, maybe a bit more, maybe a bit less depending on the location and the conditions. (The highest claimed accuracy is 7m, but you will be very lucky indeed to get this in a built-up area.)

Even with a proper survey using a WAAS GPS system or based on an accurate map, Gary's scheme will also give the same property, even quite a small property, a number of different postcodes. The average suburban home will end up with around 10 postcodes, assuming it sits on a 160 sq m plot. It could be said that this is a 'feature'. I am not so certain. I think it would make things more confusing. The code would not be much use to the ESB for instance, in determining whether a particular property already had a supply.

Lack of accuracy is not just a cause for confusion or an academic problem; if the survey point used (as a result of poor surveying, inaccurate GPS's and inaccurate maps) falls a little behind the house, for example, it could direct a driver to the house behind, which could be ten or fifteen minutes drive away. This seems like a small amount, but if a driver has a number of deliveries to make, it will upset his route and force him to backtrack. And this is not a problem that can be sorted out easily when it occurs - it will continue to occur every time that particular code is used, and there is no way to eliminate the code from use, because there is no central database.

Gary's scheme also results in the same postcode applying to a number of properties in urban areas (some properties are considerably less than 3m wide). In the case of remote rural properties, it will not necessarily indicate the appropriate way to access the building if the site is not surveyed correctly (i.e, to the entrance rather than to the building).

Also, there is no way to manually sort items addressed using Gary's scheme, because the code does not read from left to right (is not 'hierarchical' to use the jargon).

There is also no way to 'rate' a package (to figure out how much to charge for delivery) without a fairly complicated piece of programming - a conventional lookup table (basically a spreadsheet giving a price for every postcode area cannot be created). There has to be a simple way to determine shipping rates, without doing distance calculations and complicated lookups on map systems to determine a shipping price.

To deal with the above problems, the user will most likely have to fall back on the 'old' system of addresses. But the old address system is inadequate (and this is the basic reason why we need a postcode system in the first place).

Brevity is also an issue. A useful lat/lon code is just too long. The only way to make it short is to include letters as well as numbers, and letters are very prone to mix-ups when the codes are called out over the phone or handwritten.

Some comments on how I would suggest establishing a postcode and Gary's criticisms of it:

Electoral divisions (which is the administrative area I would suggest using) really have not change to any significant degree in 170 years of use. Some have been subdivided, which is not a big issue, but that's about it. (Subdivision, which is infrequent, can easily be dealt with in an orderly way if it is planned for at the outset.)

Using an existing division vastly reduces the amount of surveying required and therefore the cost. It also means that the new codes can be immediately linked to existing administrative districts used by Garda, health services and local authorities. (Incidentally, this type of code is nothing like the codes introduced by monopoly postal operators in the 50's.)

Overall, Gary's view is that a completely new system of addressing is needed and that it should have no relationship to the old one. I think his scheme falls short as a complete addressing scheme in itself.

I agree with him that the system of addresses we have at present has big flaws. But my view on how to fix the problem is different. Rather than replacing or sidestepping the system, I think it makes more sense to use a code which supplements and puts order on the existing addresses that we have.

I'd be pretty confident that satnav manufacturers could incorporate a postcode based on administrative areas like ED's without significant expense. These areas are already in GeoDirectory, a data source that most, if not all of the satnav companies licence already.