Eircode - its implemetation (merged)

plodder wrote: »

The companies that you mentioned earlier offer APIs. That means you need to obtain software that uses those APIs (as well as pay the money).

You're wrong. They do need a license. When you click on the T&Cs of a website you are agreeing to a license. They need a license. They need additional software (beyond a spreadsheet) and most importantly they need to pay over 200 euros. With the UK postcode they could do it for free, and I would guess every other postcode in the world that doesn't use a hidden random code which needs to be decoded with a licensed product.

This is like the time you said SUSI had to be saving two hours work for each application by using Eircode and it turned out to be 2 minutes. You're not fooling anyone and I am tired of this conversation (for now). Have the last word if you wish.

They can just get the data back without any API's or anything. The API is only there if needed

You said they couldn't get a licence, you were wrong, also implying they would have to licence from eircode. Now you agree you were wrong but now you move your complaint to the fact that they have to agree to terms and conditions from the resellers website? That tick box we all do all the time is now the thing you want to complain about? Seriously?

SUSI published the claim not me, I agreed with it and I wasn't wrong about it. The claim of savings is valid. You just couldn't understand it and declared it wrong.

ukoda wrote: »

So it's the same amount of work either way? A while ago you were off on a rant about them having to write software. At least it's dawned on you now that they wouldn't have to do that, just a simple sort by whatever field they wanted.

The companies that you mentioned earlier offer APIs. That means you need to obtain software that uses those APIs (as well as pay the money).

So we're back to the issue you started with, "they wouldn't be able to get access to a licence" which you were wrong about.

So what's you point now? The entire the process is too impractical and I'm stupid for saying it could be done?

Or the process is the same but you thought they couldn't get a licence and now you realise they don't need one?

Are you going to go back to talk about the cost now?

You're wrong. They do need a license. When you click on the T&Cs of a website you are agreeing to a license. They need a license. They need additional software (beyond a spreadsheet) and most importantly they need to pay over 200 euros. With the UK postcode they could do it for free, and I would guess every other postcode in the world that doesn't use a hidden random code which needs to be decoded with a licensed product.

This is like the time you said SUSI had to be saving two hours work for each application by using Eircode and it turned out to be 2 minutes. You're not fooling anyone and I am tired of this conversation (for now). Have the last word if you wish.

Sam Russell wrote: »

That is a nonsense argument. ......
.

No, it's not, put magic Eircode into magic website and it'll show you where it is

plodder wrote: »

You don't have to plot anything. You just put them all in a spreadsheet, sort them and count the number of codes in each area. Anyone who knows spreadsheets could do it.

Even if they paid the money to get the SACs out of ECAD, they would have to do this anyway.

So it's the same amount of work either way? A while ago you were off on a rant about them having to write software. At least it's dawned on you now that they wouldn't have to do that, just a simple sort by whatever field they wanted.

So we're back to the issue you started with, "they wouldn't be able to get access to a licence" which you were wrong about.

So what's you point now? The entire the process is too impractical and I'm stupid for saying it could be done?

Or the process is the same but you thought they couldn't get a licence and now you realise they don't need one?

Are you going to go back to talk about the cost now?

Sam Russell

gctest50 wrote: »

They're routing codes, doesn't matter, put magic Eircode into magic website and it'll show you where it is

That is a nonsense argument. The Eircode (or rather the postcode system) was to be designed as a neutral design - not favouring one supplier. And here we go - it is designed to favour An Post (who did not want it and said they would not use it). They could have used 2,000 routing codes matching their existing postal towns (- they just chose to use only main postal towns outside of Dublin) which would be much larger than SACs but useful for other uses, without having to bother with decoding the random bit.

Your neighbours mobile phone number probably bears no relationship to yours.
If you don't have your neighbours mobile phone number, you'll have fun trying to call him on it.

What has that got to do with postcodes?

His landline number would be useful because that is a location based system. In fact the first five digits of any landline number give a better indication of location than an Eircode - assuming it is based on Telecom's STD system from the nineties.

ukoda wrote: »

By your own words the impractical part is the work they've have to do after they got the postcode data. Even if it was free. Someone would have to sit down and plot it all out and do all that work. Yet another example of you trying to attribute an issue to eircode that doesn't fly.

You don't have to plot anything. You just put them all in a spreadsheet, sort them and count the number of codes in each area. Anyone who knows spreadsheets could do it.

Even if they paid the money to get the SACs out of ECAD, they would have to do this anyway.

Sam Russell

gctest50 wrote: »

They're routing
Switzerland = about 15000 sq miles

Switzerland has a population of over 8 million - that is twice our population.

They are quite good at infrastructure and their trains run on time as do their buses.

plodder wrote: »

No, you've missed the point. The point was this is a good example of where postcodes could have been useful, but eircode is impractical, not that it cannot be used.

By your own words the impractical part is the work they've have to do after they got the postcode data. Even if it was free. Someone would have to sit down and plot it all out and do all that work. Yet another example of you trying to attribute an issue to eircode that doesn't fly.

You will really go to extraordinary lengths to find problems with eircode that don't actually exist

ukoda wrote: »

You've missed the point. I said they CAN do it. Which you said they couldn't, and you were wrong.

It's up to them what they WANT to do.

No, you've missed the point. The point was this is a good example of where postcodes could have been useful, but eircode is impractical, not that it cannot be used.

No one has to licence a product to locate a postcodes location.

I think you'll find that you have to agree to certain terms and conditions before you can do 4,500 lookups, as well as paying over 200 euro.

12Phase wrote: »

Plodder, while the phone number would give you a unique ID, beyond the actual area code, the local number's first few digits no longer mean much as numbers are just assigned to end points almost like the way URLs are ultimately pointing towards an IP address and that can be changed depending on how the DNS database is setup.

In older technology, the number was actually used directly to route the call as the network had little intelligence. That hasn't been the case for a long time. If you are on any provider that isn't using OpenEir exchanges, you could have any number in any given area code.

When you dial a number, the exchange actually performs a look up equivalent to what a browser does with a DNS server to figure out how to route the call.

Also, the % of households with active voice landlines is shrinking every day.

I know that, but it is still the case that the vast majority of landline numbers correlate with the old routing system for the telephone network. Over time it will increasingly become less true I accept. I think for the purposes of this discussion - that there are any number of things that the parents group would do, before they would dream of spending money on eircode lookups, it is valid.

12Phase

gctest50 wrote: »

How many hours do think that'll take then ?

There's a diddlyeye dance at the crossroads n stuff this evening etc

No, most likely they simply won't bother doing anything as they probably won't see the point and the costs are prohibitive.

Carawaystick wrote: »

~Another defect is the delay in getting an eircode for new builds.
They should be allocated on grant of pp, to allow deliveries of construction materiel.
Having a postcode and only using it for letters is silly.

They should be allocated when pp is applied for, it'd make it easy to tag the site as the process goes along

plodder wrote: »

It was 4,500. So, that would cost £180. It would be more economic to just buy a few maps and spend some hours looking up addresses and pin pointing them on the maps. So, eircode wouldn't be much use in this situation imo.

How many hours do think that'll take then ?

There's a diddlyeye dance at the crossroads n stuff this evening etc

12Phase

Plodder, while the phone number would give you a unique ID, beyond the actual area code, the local number's first few digits no longer mean much as numbers are just assigned to end points almost like the way URLs are ultimately pointing towards an IP address and that can be changed depending on how the DNS database is setup.

In older technology, the number was actually used directly to route the call as the network had little intelligence. That hasn't been the case for a long time. If you are on any provider that isn't using OpenEir exchanges, you could have any number in any given area code.

When you dial a number, the exchange actually performs a look up equivalent to what a browser does with a DNS server to figure out how to route the call.

Also, the % of households with active voice landlines is shrinking every day.

plodder wrote: »

I honestly can't tell sometimes whether you actually believe the stuff you write here. I'm beginning to think that you do.

Do you seriously think it is a practical proposition for the parents group we're talking about to spend £180 on eircode lookups, or worse that it is actually a good thing? They would also have to write software to do it. A group trying to setup a school????

They would be better off asking for people's fixed line phone numbers. Area codes and exchange codes would be a far simpler basis for determining locations that costs zero euro.

Or else they would do a rough sort based on the area part of address. I can think of half a dozen other things they might do. There is no earthly way that a group like that will write software and then spend over 200 euros on Eircode lookups. It's simply laughable to suggest they would.

You've missed the point. I said they CAN do it. Which you said they couldn't, and you were wrong.

It's up to them what they WANT to do.

ukoda wrote: »

Sometimes. But to be fair, most of the time you're making up negatives that don't exist. I.e. "A group of parents wouldn't be able to look up eircodes because it's unlikely they'd get a licence." When in fact they can do it and don't need a licence at all. Of course you immediately found a problem with that too, which I could overcome. You then moved on to tell me my attitude means I'd except any old rubbish as a post code. Not true.

I honestly can't tell sometimes whether you actually believe the stuff you write here. I'm beginning to think that you do.

Do you seriously think it is a practical proposition for the parents group we're talking about to spend £180 on eircode lookups, or worse that it is actually a good thing? They would also have to write software to do it. A group trying to setup a school????

They would be better off asking for people's fixed line phone numbers. Area codes and exchange codes would be a far simpler basis for determining locations that costs zero euro.

Or else they would do a rough sort based on the area part of address. I can think of half a dozen other things they might do. There is no earthly way that a group like that will write software and then spend over 200 euros on Eircode lookups. It's simply laughable to suggest they would.

12Phase

Sam Russell wrote: »

The defects of the Eircode design imo are:

1. The routing code areas are far too large. They are based on a subset of An Posts postal towns, and the historic Dublin postal codes. 139 areas are used - and those areas vary in size geographically and in population hugely. At least one routing code is intersected by another so is not contiguous.

To a casual observer the routing codes are just totally illogical. They might make sense to some internal An Post systems but, they're utterly daft for any other purpose.

On top of that it's blatantly Dublin-centric. The codes in Dublin and surrounding areas are quite small and probably actually vaguely useful as you could follow them to a large degree on a map. There are at least 30 codes around the Dublin area.

The equivalent Greater Cork Area was given only a handful of codes and it is not 10 times smaller than greater Dublin area, by a long shot. Then there is basically one code for County Galway which is a vast area and Limerick City is also contained in one code covering a huge area.

So, basically they've just created a somewhat useful system for Dublin and to hell with the rest of the country.

You glance at an eircode in Galway and you basically know it's somewhere in the West of Ireland unless you actually key it into something to get the location or smaller areas from the database, which to be frank isn't very practical.

The system is deliberately hobbled and just reenforces the whole "Dublin" vs "Down the Country in Culshie Land" and the old notions of the Part 1 and Part 2 phone books etc etc. Even linking the eircodes in Dublin to a county letter, and giving every other area totally random looking letters just rubs people's faces in it even more.

I have already heard someone in Dublin refer to someone with a K67 code in North County Dublin as having a "Bog code".

plodder wrote: »

Mobile phone numbers are not tied to locations, whereas postcodes are.

Random postcodes are much the same as if An Post insisted on random house numbers on the pretext that someone could build a new house in their garden, but where you had to license a commercial product from them, to find out the actual locations.

Nobody would stand for that.

No one has to licence a product to locate a postcodes location.

gctest50 wrote: »

Your neighbours mobile phone number probably bears no relationship to yours.

Mobile phone numbers are not tied to locations, whereas postcodes are.

Random postcodes are much the same as if An Post insisted on random house numbers on the pretext that someone could build a new house in their garden, but where you had to license a commercial product from them, to find out the actual locations.

Nobody would stand for that.

Sam Russell wrote: »

The defects of the Eircode design imo are:

1. The routing code areas are far too large. They are based on a subset of An Posts postal towns, and the historic Dublin postal codes. 139 areas are used - and those areas vary in size geographically and in population hugely. At least one routing code is intersected by another so is not contiguous.

They're routing codes, doesn't matter, put magic Eircode into magic website and it'll show you where it is

Sam Russell wrote: »

2. The second part of the code is purely random in all four positions such that my neighbours code bears no relationship to mine. The logic for this has not been explained but it makes it more difficult to remember mine and his.

Your neighbours mobile phone number probably bears no relationship to yours.

Sam Russell wrote: »

3. It has done nothing to solve the non-unique addresses. It is basically a PPS number for properties. Now if you have the Eircode then you can find the property using a computer connected to the internet, but if you do not have the Eircode, then the location is impossible to find. This is not necessarily the fault of Eircode, but it was an opportunity lost to go some way to solve it. For example, many addresses in urban areas are non-unique because they refuse to number their houses or display their house names.

If you don't have your neighbours mobile phone number, you'll have fun trying to call him on it.

Sam Russell wrote: »

4. The licencing appears to not be necessary if a simpler design had been chosen. For example, the Swiss have a 5 digit code plus a two character canton identifier.

I know that Eircode is here to stay for at least a decade so by then it will be obvious if it has succeeded or not. Hopefully some move will be made to eradicate the non-unique addresses by then.

Switzerland = about 15000 sq miles

FAILSAFE 00

HelgaWard wrote: »

Using Eircodes constantly at work and have to say they are very useful. I think they need to make the data more open and accessible to all in order for their implementation to be really successful. Also think they need to implement a system whereby people/developers can apply for Eircodes to reduce the costs of keeping the system up to date. I think these changes will happen and it will ultimately become a really useful national piece of infrastructure.

It will absolutely become useful once it works however it's a shambles at the moment.
They need to become transparent and highlight what their strategy is to make it work accurately.

It's not working at the moment and that needs to be accepted and then worked on. Working for a few isn't enough for a system like this.

HelgaWard

FAILSAFE 00 wrote: »

You forgot the people that see things as they actually are. Not head in the clouds dreamers or dooms day apostles.

The system is as useful as tits on a bull.

Case closed.

Using Eircodes constantly at work and have to say they are very useful. I think they need to make the data more open and accessible to all in order for their implementation to be really successful. Also think they need to implement a system whereby people/developers can apply for Eircodes to reduce the costs of keeping the system up to date. I think these changes will happen and it will ultimately become a really useful national piece of infrastructure.

FAILSAFE 00 wrote: »

Case closed.

Oh right thanks for the update. Let the mods know to close the thread, you've given your opinion and that's it so. Case closed. Go ahead and ignore all the people and companies who find it useful. Bury your own head in the sand and ignore that.

plodder wrote: »

but at least I'm suggesting improvements to it.

Sometimes. But to be fair, most of the time you're making up negatives that don't exist. I.e. "A group of parents wouldn't be able to look up eircodes because it's unlikely they'd get a licence." When in fact they can do it and don't need a licence at all. Of course you immediately found a problem with that too, which I could overcome. You then moved on to tell me my attitude means I'd except any old rubbish as a post code. Not true.

FAILSAFE 00

ukoda wrote: »

Some people look at something and immediately see the potential and oppurtinities

Other people look at the same thing and can only see problems and negatives

I'm glad I'm the first kind of person

You forgot the people that see things as they actually are. Not head in the clouds dreamers or dooms day apostles.

The system is as useful as tits on a bull.

Case closed.

HelgaWard

Carawaystick wrote: »

~Another defect is the delay in getting an eircode for new builds.
They should be allocated on grant of pp, to allow deliveries of construction materiel.
Having a postcode and only using it for letters is silly.

Totally agree, their current update process is to wait until the property receives mail, then the GeoDirectory gets updated, then an Eircode gets created. Too much of a delay and things being left to chance. If for example someone builds a new house, but directs their mail to a different house, there will be a huge delay before an Eircode is allocated. You should have to to apply for an Eircode when you have submitted a commencement notice to your planning authority. Developers building large developments should have to submit plans or the estate layout and Eircodes be requested accordingly.

Carawaystick

~Another defect is the delay in getting an eircode for new builds.
They should be allocated on grant of pp, to allow deliveries of construction materiel.
Having a postcode and only using it for letters is silly.

Sam Russell

plodder wrote: »

A geocode was not the best choice technically. But, if the state wasn't capable of getting a reasonable licensing system from geodirectory (an entity that it actually owns) then perhaps a completely free geocode would have been better.

I don't think it's perfect either, but at least I'm suggesting improvements to it.

The defects of the Eircode design imo are:

1. The routing code areas are far too large. They are based on a subset of An Posts postal towns, and the historic Dublin postal codes. 139 areas are used - and those areas vary in size geographically and in population hugely. At least one routing code is intersected by another so is not contiguous.

2. The second part of the code is purely random in all four positions such that my neighbours code bears no relationship to mine. The logic for this has not been explained but it makes it more difficult to remember mine and his.

3. It has done nothing to solve the non-unique addresses. It is basically a PPS number for properties. Now if you have the Eircode then you can find the property using a computer connected to the internet, but if you do not have the Eircode, then the location is impossible to find. This is not necessarily the fault of Eircode, but it was an opportunity lost to go some way to solve it. For example, many addresses in urban areas are non-unique because they refuse to number their houses or display their house names.

4. The licencing appears to not be necessary if a simpler design had been chosen. For example, the Swiss have a 5 digit code plus a two character canton identifier.

I know that Eircode is here to stay for at least a decade so by then it will be obvious if it has succeeded or not. Hopefully some move will be made to eradicate the non-unique addresses by then.