Debunking 13 anti - Eircode arguments
Posted by Pat Donnelly
| Apr 15, 2015
Brian Lucey's latest blog post 13 things that are wrong with Eircode is a comprehensive re-publishing of anti-Eircode arguments. What's wrong with the analysis? Lots.
1. Loc8 is Free, Eircode will cost up to 100 million
This is a classic case of comparing apples to oranges. First let's look at the claim that Loc8 is free.
If it is the intention of the Department of Communication Energy and Natural Resources to notify 2.2 million addresses of their individual postcode then this is likely to cost €2.2 million to deliver when postage and printing are taken into account. If a sizeable number of these locations were able to self‐service the creation of their own postcode this would mean that the savings in postage alone could pay for the acquisition of the portion of Loc8 Code that is not already owned by the state via Enterprise Ireland.
That doesn't look free to me. The "saving" identified is based on the assumption that a self-service option isn't available with Eircode. This is false. The decision to disseminate all Eircodes by post was not forced by the lack of a self-service option, so no "saving" would accrue.
Let's now consider the actual cost of the Eircode design in the context of the build phase of Eircode. How much of the €16 million project phase was spent on the Eircode design? This figure versus the amount Loc8 want to buy them out would be an apples to apples comparison.
2. Poor Design (No Design?)
According to the article
No competition was ever held for the design of Eircode.
An inexpensive design competition would have offered the DCENR an overview of the innovative code technologies already in existence.
A Competitive Dialogue Tender Process was run by DCENR that lasted more than two years. The results of this comprehensive process informed the final tender requirements issued by DCENR. Participation in the Competitive Dialogue Tender process was at the bidders own expense. I think that qualifies as an inexpensive design competition. Autoaddress can confirm we assessed alternative code technologies, including an exhaustive evaluation of a grid-based design similar to Loc8. The design we submitted in the final tender submission, now Eircode, was designed by Autoaddress for rapid adoption, sensitivity to postcode discrimination, memorability and ease of validation.
3. No Field Trial/Pilot
What would happen in a pilot? Mrs Miggins, Ballysmallplace, West of Ireland receives a "pilot" postcode in the post. How would she use it? It's still the pilot phase, no one has updated their systems to accept postcodes. Are we piloting An Post's ability to deliver post?
4. Database Driven
The full Eircode database does not require 2Gb of space to store on a device. This is another apples and oranges example, using the size of the full text export of the entire ECAD database. The file size of an off-line version of the Eircode database that compares with Loc8, i.e. a postcode and a coordinate, is 30Mb. Without compression. If you want to have the address as well, and do so in the least optimised manner, the file approaches 100mb. An optimised file format would be less than 70mb. Top selling mobile games approach 500mb in size. Eircode will fit on a Sat-Nav, and is easily downloaded to a mobile phone for off-line use.
5. Major Data Security Risk
The Eircode database of addresses and Eircodes is public, not hidden. Everyone will be able to lookup any address or Eircode. For €60 you can purchase an entire copy, no need to resort to espionage. This is equivalent to the UK PAF, which has the address and postcode of every address in UK. There is no personal data in the Eircode database, simply addresses, Eircodes and their locations. The source data has been available and used by numerous companies in Ireland for the last 15 years.
6. No Logic, easily confused
The article states the following
One example would be a manual parcel sorting depot where humans read the address on a parcel and then place it in a particular bay or cage for onward delivery, random codes will only add to address confusion unless every worker has a handheld computer with access to the database, this would be too slow and too expensive to use day to day.
Emergency services especially will suffer from confusion with a code with no built in safeguards and that needs database access at all times. Having to always look up a code before knowing even the rough destination will introduce delays and may cost lives.
The clear implication here is that a Loc8 code can be used to group parcels for delivery and that a Loc8 code can provide a reliable "rough destination" for emergency services without access to technology. Grid-based postcodes are incapable of providing this functionality, because they ignore accessibility. The following example was created by Autoaddress to illustrate that Loc8 codes are not suitable for manual postal sortation or grouping of deliveries.
Loc8 responded with a much better image that illustrates the problem definitively
Loc8 recommend grouping by the squares on the map. The 5XX squares group properties either side of the river Corrib. This image is all that is is required to see that Loc8 is completely unsuitable for manually sorting post, grouping deliveries, or providing reliable visual clues for emergency response.
It works sometimes but not others? That would be the definition of unreliable.
7. Requires constant updating / Delayed Updates
Debunked in point 4 above.
8. Difficult to integrate with software / devise
Debunked in point 4 above.
9. No use for Road Traffic Accidents
Mobile phone geolocation> is the problem, the solution is the LocateMe112 system, designed by helicopter pilot Lt Colin Gallagher, which enables the GPS chip on an emergency caller's smartphone to be activated remotely – via a link sent via SMS – by rescue services, who can then pin-point the exact location of the injured person.
The key feature of LocateMe112 is that it doesn't require the installation of an app, only needing the caller to have a smartphone with location services turned on.
It should be obvious that asking a user to click on a link sent to them by SMS is far superior to having them go to an app store, find an app, download the app, get a code and read it back to the operator.
10. Emergency - Communication Errors
International best practise is to ask for both an address and a postcode, independently without prompting, to verify the postcode is correct. The Eircode design ensures that similar addresses have very different Eircodes to aid verification. Loc8 "checker" does not check the two digits in the middle of a Loc8 code, so if digits were reversed, e.g. 19 instead of 91, then you could easily end up the wrong side of a river as illustrated in the River Corrib example above.
11. Emergency - Hidden Code
Eircode is a public database, accessible to all. It is not hidden, everyone will be able to determine the Eircode of any building. The rest of the point is debunked in point 9 above.
12. Limited use for Tourism
Rock of Cashel, Newgrange, Hill of Tara, etc. will all have an Eircode. They have visitor centres, etc. that receive post.
13. Loc8 can be used as a Postcode
The postcode design becomes irrelevant if it isn't ubiquitous as people won't have recall of their postcode for other purposes. An Post deliver millions of letters every day, and have sophisticated OCR scanning capabilities but no system will accurately scan 100% of the time (otherwise we wouldn't be bedevilled by "Captcha" on websites). A successful postcode design must enable efficient manual sortation of post, otherwise it isn't a postcode. This is the function of the Routing Key. You cannot have a postcode without a Routing Key. Loc8 cannot be used to manually sort post.
Loc8 is a very simple algorithm, only 50 lines of code. It cannot be modified if it inadvertently assigns an offensive code to a dwelling. Large swathes of the country have codes that begin with inappropriate or offensive terms, and numerous examples exist of codes ending in offensive terms.