Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Eircode discussion

Options
13567…15

Comments

  • Options
    #62
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    BoatMad wrote: »
    Its been argued by many that remembering a 7 digit alphanumeric code is not difficult, People remember codes that are important to them and in general use, Eircodes may or may not be in that criteria for different people. Arguably remembering a hierarchical code is no better.
    In all, this is a very very weak argument against Eircode .
    No point in continuing to rehash the same points, but suffice to say this could have been something that the public could have been asked their opinion on
    Apart from that, the main value of an open structured postcode is that people can make use of it without having to artificially license and pay rent for a product or data set purely to decode something that shouldn't need to be decoded. Anyone who knows anything about IT, or has more than one brain cell can see this.
    NO , I dont agree, in order to at least try and implement data protection laws, there is a need to have (a) fully anonymised codes and (b) control access to the underlying address database. Other wise you have uncontrolled telemarketing etc ( and you have some of this issue in the UK) , also you have post code lottery type issues as well. IN my view, technically, Id prefer a complete anonymised code.
    Eircode is not anonymous. On the contrary, it's the first postcode in the world to identify households and in many cases people too. Your point that the data needing to be licensed mitigates the data protection risk will not be borne out. If someone's Eircode turns up on a list that was thought to only contain anonymous data, anyone will be able to convert that to a full postal address on the eircode finder.
    If you run a small business serving a particular set of areas, you can put a map on the wall of the areas you cover and you just look at the map and you can tell where your customers are without needing a database or software. It's quite simple really. If software adds value to that, brilliant. In some cases it will. In others it won't.
    This is not the purpose of a post code. With eircode and a licensed copy of GeoDirectories its easy to to what you say. a very small business can just analyse the text address anyway and buy a box of thumbtacks !. Your issue is more a function of licensing costs not Eircode per say
    Again the point is you shouldn't have to license anything - no more than you have to when you build up a database of customer addresses. You say it's not the purpose of a postcode, why isn't it?


  • Options
    #63
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    Personally, I think some of these issues could be mitigated if they made a dataset available for free that comprises the complete set of Eircodes and only the small area code for each. It wouldn't affect the core address verification/lookup business, for which you'd still need ECAD or ECAF. It would facilitate applications where a data controller doesn't want to deal with Eircodes, but would like reasonably fine grained location info, but which is still anonymous. A web api could redirect to some third party to do the lookup and only the small area code would be returned to the user. Shape files for small area boundaries exist and someone could add some limited geocode info to that.

    It would also facilitate simple applications for which the actual small area boundaries make sense. For anything more complicated, you would need ECAD and the detailed location info which that provides. Opinions?


  • Options
    #64
    GJG
    Banned (with Prison Access) Posts: 158 ✭✭GJG


    plodder wrote: »
    Personally, I think some of these issues could be mitigated if they made a dataset available for free that comprises the complete set of Eircodes and only the small area code for each. It wouldn't affect the core address verification/lookup business, for which you'd still need ECAD or ECAF. It would facilitate applications where a data controller doesn't want to deal with Eircodes, but would like reasonably fine grained location info, but which is still anonymous. A web api could redirect to some third party to do the lookup and only the small area code would be returned to the user. Shape files for small area boundaries exist and someone could add some limited geocode info to that.

    It would also facilitate simple applications for which the actual small area boundaries make sense. For anything more complicated, you would need ECAD and the detailed location info which that provides. Opinions?

    I think that's an excellent idea. That level of granularity is fine for many uses and - most importantly - it's easily doable.


  • Options
    #65
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    See post #9 in this thread.


  • Options
    #66
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    Sam Russell wrote: »
    See post #9 in this thread.
    Good point. It's pretty much the same idea. I think the small area code is the statistical code you referred to. But, its use isn't just for statistics. It would be useful for other things as well.


  • Advertisement
  • Options
    #67
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    plodder wrote: »
    Good point. It's pretty much the same idea. I think the small area code is the statistical code you referred to. But, its use isn't just for statistics. It would be useful for other things as well.

    Well, I thought it could be used as an anonymous post code. If the format mimicked the Eircode, it would be checked as a genuine, but annonymous, post code. EG if your Eircode was D04 AB12, then the anonymous one could be D04 XY66, which would be the same for all your neighbours.

    Of course, if it caught on, everyone would be using the substitute one instead of their own one. An Post would not care, but others might.


  • Options
    #68
    GJG
    Banned (with Prison Access) Posts: 158 ✭✭GJG


    Sam Russell wrote: »
    Well, I thought it could be used as an anonymous post code. If the format mimicked the Eircode, it would be checked as a genuine, but annonymous, post code. EG if your Eircode was D04 AB12, then the anonymous one could be D04 XY66, which would be the same for all your neighbours.

    Of course, if it caught on, everyone would be using the substitute one instead of their own one. An Post would not care, but others might.

    I don't think that giving everyone two postcodes is a runner. The likelihood of the average punter understanding the difference between the two, let alone remembering which was which, is remote.

    Plodder's idea, I think, is a constructive suggestion. It would just take an API that would allow an interface to take in an Eircode and return a small area code. That would be perfect for, say, website store locators, coverage areas for pizza delivery or franchises.


  • Options
    #69
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    Sam Russell wrote: »
    Well, I thought it could be used as an anonymous post code. If the format mimicked the Eircode, it would be checked as a genuine, but annonymous, post code. EG if your Eircode was D04 AB12, then the anonymous one could be D04 XY66, which would be the same for all your neighbours.

    Of course, if it caught on, everyone would be using the substitute one instead of their own one. An Post would not care, but others might.
    Okay, not exactly the same idea then. The small area code is a completely different identifier (9 digit number I think). I don't think they would be used directly by the public.


  • Options
    #70
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    plodder wrote: »
    Okay, not exactly the same idea then. The small area code is a completely different identifier (9 digit number I think). I don't think they would be used directly by the public.

    Well, it is much the same. Maybe pushing it a bit far with an alternative Eircode.


  • Options
    #71
    clewbays
    Closed Accounts Posts: 148 ✭✭clewbays


    plodder wrote: »
    Good point. It's pretty much the same idea. I think the small area code is the statistical code you referred to. But, its use isn't just for statistics. It would be useful for other things as well.

    Small areas are too small for statistics. I doubt the CSO publish anything more than very summarised age groups by sex etc.


  • Advertisement
  • Options
    #72
    FishOnABike
    Registered Users Posts: 4,460 ✭✭✭FishOnABike


    plodder wrote: »
    Personally, I think some of these issues could be mitigated if they made a dataset available for free that comprises the complete set of Eircodes and only the small area code for each. It wouldn't affect the core address verification/lookup business, for which you'd still need ECAD or ECAF. It would facilitate applications where a data controller doesn't want to deal with Eircodes, but would like reasonably fine grained location info, but which is still anonymous. A web api could redirect to some third party to do the lookup and only the small area code would be returned to the user. Shape files for small area boundaries exist and someone could add some limited geocode info to that.

    It would also facilitate simple applications for which the actual small area boundaries make sense. For anything more complicated, you would need ECAD and the detailed location info which that provides. Opinions?
    This idea appeals to me. Even without the small area code shape files and mapping software an api that returned an approximate area (say down to 10km or 1km square based on the national grid) would be useful for grouping deliveries, analysing statistical data etc...

    If more accurate data (down to an individual delivery point) is needed it is reasonable to have to pay to use the data for a high volume of lookups.


  • Options
    #73
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    clewbays wrote: »
    Small areas are too small for statistics. I doubt the CSO publish anything more than very summarised age groups by sex etc.
    The 2011 census was the first one to be published down to small area level. They were designed to be as small as possible while retaining sufficient anonymity. That makes them useful for other purposes as well. If they are too small, you just combine them (or you license ECAD and build your own grouping). A square grid could be used either, but if you go too small you will probably lose anonymity in some places. Too large and they are less useful.


  • Options
    #74
    ManAboutCouch
    Registered Users Posts: 225 ✭✭ManAboutCouch


    clewbays wrote: »
    Small areas are too small for statistics. I doubt the CSO publish anything more than very summarised age groups by sex etc.

    The entire census is published at Small Area level, see this for an example.

    There is also a 1km grid square output which only contains age groups by sex with some anonymity protection built in. That is available here.


  • Options
    #75
    clewbays
    Closed Accounts Posts: 148 ✭✭clewbays


    ManAboutCouch wrote: »
    The entire census is published at Small Area level, see this for an example.

    There is also a 1km grid square output which only contains age groups by sex with some anonymity protection built in. That is available here.

    Thanks for link. They are very basic tables e.g. the very small number of individual nationalities listed and even then there is only one Lithuanian living in that small area. Hence nationality was not cross-classified by age or employment status etc. as that level of detail was probably considered too detailed for areas with around 200 persons.


  • Options
    #76
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    I'd be curious to know if there are any countries that charge money for postcode data like this. We already know that they are free in the UK.

    I just checked the US, and it's ironic that at the hub of world capitalism the government is required to release all this kind of information to the public domain (for free). I suspect they realise that governments don't always have the best record at commercialising their own data and it's best to just release the information as a public good, and let the private sector commercialise it then.

    It also means that voluntary organisations and individuals who would never be in a position to pay for access, get to use such data.

    I also found a site which has free postcode data for around 75 other countries

    https://www.aggdata.com/freedata-category/postal-codes


  • Options
    #77
    ManAboutCouch
    Registered Users Posts: 225 ✭✭ManAboutCouch


    clewbays wrote: »
    Thanks for link. They are very basic tables e.g. the very small number of individual nationalities listed and even then there is only one Lithuanian living in that small area. Hence nationality was not cross-classified by age or employment status etc. as that level of detail was probably considered too detailed for areas with around 200 persons.

    The detail published in the 2011 census at Small Area level is exactly the same as that published in the 2006 census at ED level. ED's are (usually) significantly larger than Small Areas, there are around 3,400 ED's in the country, while there are approximately 18,500 Small Areas.

    I disagree with your claim that Small Areas are too small for statistics. The level of detail is sufficient for complex and detailed analyses to be carried out.


  • Options
    #78
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    If there were 3,400 routing keys or maybe 2,000, then many of the complaints about Eircode would evaporate. There are about 2,000 post towns in Ireland, but only 139 were chosen for some reason.


  • Options
    #79
    ukoda
    Registered Users Posts: 2,084 ✭✭✭ukoda


    Sam Russell wrote: »
    If there were 3,400 routing keys or maybe 2,000, then many of the complaints about Eircode would evaporate. There are about 2,000 post towns in Ireland, but only 139 were chosen for some reason.

    I disagree that the complaints would evaporate. We've seen in the original thread that once one myth got busted people moved on to complain about something else
    First was the tender
    Then won't fit on sat nav
    Etc etc


  • Options
    #80
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    ukoda wrote: »
    I disagree that the complaints would evaporate. We've seen in the original thread that once one myth got busted people moved on to complain about something else
    First was the tender
    Then won't fit on sat nav
    Etc etc

    The Tender and other aspects were pursued by vested interests on both sides.

    There were a number of people who consistently railed against the poor design and the secrecy of the design criteria and the fact that the design specification from the Oireachtas was ignored.

    When it was revealed that An Post had no intention of using it (since denied) and that Shannon Airport was in Limerick, then it was a lost cause as far as they were concerned and every bump in the road became another mountain.

    However, the large and variable size of the routing codes, their complete meaningless nature, and the use of nonsense random codes all caused most to abandon the whole concept.

    Making routing keys of consistent size (as to the number of addresses each one covers, and adhering to county boundaries would remove a large group from the nay-sayers camp.

    As to fitting on a satnav - well it does not fit any so far.


  • Options
    #81
    oscarBravo
    Technology & Internet Moderators Posts: 28,799 Mod ✭✭✭✭oscarBravo


    Sam Russell wrote: »
    ...that Shannon Airport was in Limerick...
    You know that Shannon Airport isn't in Limerick, right?
    Making routing keys of consistent size (as to the number of addresses each one covers, and adhering to county boundaries would remove a large group from the nay-sayers camp.
    I've pointed out before that there are downsides to having routing keys adhering to county boundaries. Next-door neighbours in Charlestown would have different routing keys.

    You can argue that there are workarounds to this problem, to which my retort ought to be obvious...


  • Advertisement
  • Options
    #82
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    oscarBravo wrote: »
    You know that Shannon Airport isn't in Limerick, right?

    So did everyone else. That is why that sort of thing was an own goal for Capita.


  • Options
    #83
    oscarBravo
    Technology & Internet Moderators Posts: 28,799 Mod ✭✭✭✭oscarBravo


    Sam Russell wrote: »
    So did everyone else. That is why that sort of thing was an own goal for Capita.

    You think Capita should have hand-edited the Geodirectory in order to avoid the sort of stupid criticism that was levelled at them?


  • Options
    #84
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    oscarBravo wrote: »
    You think Capita should have hand-edited the Geodirectory in order to avoid the sort of stupid criticism that was levelled at them?

    I am saying that is the kind of own goal that is hard to counter. It would not have happened if the routing keys covered smaller areas, and were of consistent extent.


  • Options
    #85
    ukoda
    Registered Users Posts: 2,084 ✭✭✭ukoda


    Sam Russell wrote: »

    As to fitting on a satnav - well it does not fit any so far.

    so both Garmin and TomTom have made public statements that they are implementing eircode. As shown on the implementation thread.

    So are you trying to refute that it will fit? Or is the purpose of that comment to highlight the delay?


  • Options
    #86
    Sam Russell
    Moderators, Science, Health & Environment Moderators Posts: 19,679 Mod ✭✭✭✭Sam Russell


    ukoda wrote: »
    so both Garmin and TomTom have made public statements that they are implementing eircode. As shown on the implementation thread.

    So are you trying to refute that it will fit? Or is the purpose of that comment to highlight the delay?

    I am saying it is a moot point until it is announced. For example, If Garmin or TOM TOM say it will only fit on new 2016 ones, then it will not fit on my one.

    So moot.


  • Options
    #87
    ukoda
    Registered Users Posts: 2,084 ✭✭✭ukoda


    Sam Russell wrote: »
    I am saying it is a moot point until it is announced. For example, If Garmin or TOM TOM say it will only fit on new 2016 ones, then it will not fit on my one.

    So moot.

    Is there 20-30MB of storage on your device? Odds are there is. It's highly likely to be available as an upgrade to all models. We are talking about something 1/10th the size of the Facebook app on your phone


  • Options
    #88
    oscarBravo
    Technology & Internet Moderators Posts: 28,799 Mod ✭✭✭✭oscarBravo


    Sam Russell wrote: »
    I am saying that is the kind of own goal that is hard to counter. It would not have happened if the routing keys covered smaller areas, and were of consistent extent.

    Hang on - if the routing codes were smaller, looking up V14 EE06 wouldn't have returned "Limerick" as the county?

    Seriously?


  • Options
    #89
    gctest50
    Registered Users Posts: 9,605 ✭✭✭gctest50


    plodder wrote: »
    .........

    I also found a site which has free postcode data for around 75 other countries.........

    meh postcodes

    this works worldwide, accurate to a good few metres

    what3words functions without a data connection.

    This solves a perpetual constraint when in remote and unaddressed locations, or in areas with poor connectivity.

    http://w3w.co/spirited.curtails.chimps





  • Options
    #90
    clewbays
    Closed Accounts Posts: 148 ✭✭clewbays


    ManAboutCouch wrote: »
    I disagree with your claim that Small Areas are too small for statistics. The level of detail is sufficient for complex and detailed analyses to be carried out.

    So why don't you contact the CSO for a more complex and detailed analysis of the small area that you highlighted (around Jervis street in Dublin 1). For example, ask for the age group and sex of the 1 Lithuanian living in that small area. Could you also ask for the number of persons of Greek nationality living in that small area - nothing about their educational attainment or socio-economic group just one figure of how many Greeks lived in that small area.

    Looking forward to you giving us those complex and detailed statistics.


  • Advertisement
  • Options
    #91
    plodder
    Registered Users Posts: 7,220 ✭✭✭plodder


    oscarBravo wrote: »
    I've pointed out before that there are downsides to having routing keys adhering to county boundaries. Next-door neighbours in Charlestown would have different routing keys.
    Yet, these neighbours already travel to different places (Sligo or Castlebar) for their county council business. They potentially pay different property tax. They elect different councillors to a different council. So, it's hardly all that big a deal if their postcodes have different prefixes.

    Not that aligning with county boundaries is an essential feature of a postcode, but it's funny to see people arguing that it's a bad thing.


Advertisement