National Postcodes to be introduced

spacetweek

iantheteacher wrote: »

Every single day I have to give directions. Couriers who are paid good money to deliver often give up and ask me to come collect from them at the nearest town. Tesco won't deliver unless I select an address from their list that is incorrect.

Move to a town

thirty-six dragons

Not exactly a workable national spatial strategy.

SPDUB

I would however think that it's long past due that An Post insisted that people put a mail box on their gate way.

COMREG won't allow them

I find it ridiculous that Irish postmen/women have to walk up driveways, bend down to low letter boxes

It took a postman winning a legal action against An Post for the Government to change the building regulations to ban low level letterboxes for new buildings .However An Post is still obliged to deliver to any pre existing ones and the pre existing ones have to be taken into account when planning routes .

And you can give up talking about Postman Pat and how he wants to be able to sort his little bag of mail according to districts and roads - because your friendly Postman Pat already uses a postcode for your address - your address is part of the big address directory with geolocation coordinates

No they don't in local delivery offices and for a lot of material in the mail centres .

Only junk mail and big mailouts by Eircom, ESB etc use it and even then it's not on any material that reaches the local delivery office .

A lot is still sorted by district then road and number .You could put the Geodirectory number on a letter If you knew it and it won't get delivered unless somebody recognises it as Geodirectory ( which for the most part isn't likely ) and goes off to try and find what that number means .

Tesco won't deliver unless I select an address from their list that is incorrect.

That's a failing on the part of Tesco though in using software that is presumably for the UK and not willing to adapt it for the Irish market

thirty-six dragons

The notion that roadside postboxes are going to make addressing easier is a little urban-naive. I have my postbox at the entrance to my property - quite a walk from the house and nowhere in sight of my house. It is a courtesy to the postman. It is not required. An Post are paid to deliver - not to approximate. However I would still challenge you to find my postbox based on any address.

The problem with Tesco is no a UK thing - it is based on their use of the An Post address directory which lists me incorrectly and I seem to have no control over this information.

antoinolachtnai

If you email Geodirectory, they might fix it for you, if you explain your situation fully. They will contact the local postal delivery office to check.

As you have pointed out, Geodirectory is not particularly accurate. I am not knocking it, it is a good production and has improved a lot over the years, but it is nowhere near definitive. (There is also no guarantee whatsoever on the coordinates it lists for the buildings. This is not because they are doing a bad job, it is more to do with how difficult it is to compile a really fine-grained database of addresses.)

thirty-six dragons

I think it's a smashing directory - but it falls down as it isn't in the public sphere so you'd never know if your entry was correct or not unless you knew about such things and there seems no way to share your location (no public postcode - even though each entry is numbered in the database) and no procedure for updating your record.

(I have emailed. No response)

I could add the OpenPostcode to their directory in about 12 minutes (to pick a random shorttime) and everyone would have a geolocation code. Refinements would continue to be made over time. People then at least would also have the option to define their own OpenPostcode - to be adopted by a directory like GeoDirectory or not, depending on them. But at least Tesco would let me get my address right and the ambulance will be able to find me.

antoinolachtnai

If you weren't in the directory though, Tesco still wouldn't deliver to you, would they?

The coordinates for each address have never been checked and are potentially incorrect. This is likely to be a major problem for non unique addresses in rural areas. Every single code you would create would be subject to change.

SPDUB

antoinolachtnai wrote: »

If you weren't in the directory though, Tesco still wouldn't deliver to you, would they?

The coordinates for each address have never been checked and are potentially incorrect. This is likely to be a major problem for non unique addresses in rural areas. Every single code you would create would be subject to change.

I'm not sure what you mean about not being checked

Do you mean at the local delivery office because the Geodirectory staff do visit the local delivery offices with maps and ask the postman what building is at various locations and is there new construction under way etc .

antoinolachtnai

Sure, they do that, and that is fine up to a point in urban areas, though it still has issues. In rural areas though, the delivery system works on the basis of local knowledge of postmen, not what's in the directory. It has to be this way because the addresses are non-unique. The whole thing depends on the relationships that the postman has with his customers.

If you go and actually try to match up buildings with geodirectory entries in a rural area, you will have a tough enough time.

Also, when the geodirectory staff visit the delivery office they are only checking the general run of things like street names, etc. They aren't checking that the individual map coordinates for each individual delivery point are correct. No one checks this. It's unverified, and although it's highly useful, I don't think it would be a very good idea to base a coding system upon it.

thirty-six dragons

antoinolachtnai wrote: »

If you weren't in the directory though, Tesco still wouldn't deliver to you, would they?

The coordinates for each address have never been checked and are potentially incorrect. This is likely to be a major problem for non unique addresses in rural areas. Every single code you would create would be subject to change.

Makes no sense. Tesco deliver - just since they use the GeoDirectory they use an incorrect address (and then I give them long complicated directions).

Whether coordinates are incorrect is not an algorithm problem. That is an implementation problem - everything gets implemented. It makes no sense to insist on knocking everything because it is not complete.

thirty-six dragons

antoinolachtnai wrote: »

Sure, they do that, and that is fine up to a point in urban areas, though it still has issues. In rural areas though, the delivery system works on the basis of local knowledge of postmen, not what's in the directory. It has to be this way because the addresses are non-unique. The whole thing depends on the relationships that the postman has with his customers.

If you go and actually try to match up buildings with geodirectory entries in a rural area, you will have a tough enough time.

Also, when the geodirectory staff visit the delivery office they are only checking the general run of things like street names, etc. They aren't checking that the individual map coordinates for each individual delivery point are correct. No one checks this. It's unverified, and although it's highly useful, I don't think it would be a very good idea to base a coding system upon it.

The benefit of a geolocation code is that it actually doesn't need a database - but since so many people natter on and on and on and on about the imaginary needs of An Post - I just say that their own database could be converted instantly.

People should try the GeoDirectory for themselves. It does also pinpoint addresses in rural areas.

bk

antoinolachtnai wrote: »

If you weren't in the directory though, Tesco still wouldn't deliver to you, would they?

The coordinates for each address have never been checked and are potentially incorrect. This is likely to be a major problem for non unique addresses in rural areas. Every single code you would create would be subject to change.

That doesn't make any sense. If we had geolocation codes, Tesco would stop using and requiring matching Geodirectory addresses.

Instead they would just:

1) Ask you to enter your geolocation code (maybe show the code location on a map to verify it.
2) Check that the location is within a certain distance from a depot that they deliver from.
3) Deliver to exactly that geolocation.

All much simpler and cleaner then their current Geodirectory based system.

This would also apply for any other delivery company.

I see the only need for Geodirectory as a quick and easy way to inform people of their new geolocation code. Generate a postcard for every address in the Geodirectory db with the new corresponding geolocation string and perhaps a small area map showing where the resulting geolocation is and post it to everyone.

If the person who receives the postcard sees that the geodirectory/geolocation is wrong (but still received the card due to local postmans knowledge), then give them a website that they can go to and correct the geodirectory/geolocation entry.

Of course people could create their own geolocation codes for anything they like. Location of a historical monument in the middle of a field, etc. You only need geodirectory to get it started and to help inform people of it.

larchill

Wonder where the postcode thing is at? I see the DCNR have the tender out for the National Postcode Operator. This will be the body responsible for managing postcodes. This has been out since Jan last. Has this secuumed to the whole EU/ECB/IMF thing? :rolleyes:

spideog7

I looked at the Open Post Code and it seems nice and it has a hierarchical aspect which allows for a certain amount of privacy, although I didn't do the math to see how big the regions would be?

The only issue with using Geocodes is that for the post office they may not necessarily line up with the post offices distribution network i.e. two people who's houses are very close might be on two completely different roads (due to a lake, bay, estuary, mountain etc.) and therefore on two completely different post office delivery routes. That being said without some sort of database system this will always be a problem. Given that almost everything is done on computers now Geocodes make more sense than Postcodes.

antoinolachtnai

Two houses which are adjacent 'as the crow flies' may very often not be close together at all in practical terms, as regards walking or driving distances. This is common in suburban areas.

In practice this means that unless you have a way of verifying the longitude and latitude for every delivery point to a fairly high level of accuracy, depending on this code will have a fair few significant mix-ups, i.e., the code will misdirect the service provider.

The only way to verify the coordinates for all these delivery points is to establish a register or database of correct codes.

spideog7

antoinolachtnai wrote: »

The only way to verify the coordinates for all these delivery points is to establish a register or database of correct codes.

Hmm, why would you need a database? The maths can't be wrong, you put in a code and it gives you a geographic position accurate to a few metres. If your navigation equipment can't find a way to get there then it's not really the codes fault.

I don't understand why Loc8 requires you to "verify" your postcode or how you can "create" and "manage" the postcode, surely it's just an algorithm with a bunch of rules, you pump in a geographical location and it gives you a code. I can't own that code and if I want to give the middle of a bog a postcode and I want to give that postcode to my DHL man then that's my problem.

antoinolachtnai

The problem is, if we as a country adopted one or other of the coordinate-based codes, where would we get a list of accurate coordinates to begin with, and how do we distribute them to the people who need them? We would need a database in order to do this. We could use GeoDirectory or a similar database to resolve this.

In practice, we would have a big problem with inaccurate codes. The GeoDirectory codes are just not going to be all that accurate for various reasons (nor do GeoDirectory claim that they are completely accurate). Over time, codes would have to be checked and corrected. This would require a database. People who were already using the older, inaccurate code for the premises would also have to have a way to update their database with the new, more accurate code.

Incidentally, we would also depending on the map in the device being accurate and comprehensive enough to figure out what is the appropriate way to get to the coordinate. if the roads on the device are even slightly off, it may direct you up the wrong road. (For example, if you are looking for a mews on a lane running parallel to an adjacent major road, as is common in older urban areas, then if the GPS doesn't know about the lane, it will direct you to the building on the major road.)

spideog7

This is the case with any system that is a location based code rather than a post code. Post codes however will have even greater problems as they have the same issues mentioned above along with additional problems such as their complete reliance on a database and their lack of accuracy.

I would be very impressed if there was any post code system which somehow encoded the exact location along with the route required to get there regardless of where you were coming from... into 8 digits.

antoinolachtnai

A classic postcode would not have this issue since it would be based on postal operations. (This has many other drawbacks.)

A post code based on electoral divisions and small areas would not have this problem, as every building would clearly be within one or other electoral division or small area. There are some exceptions, but not that many. There are existing databases. They are not perfect, but they are a very workable starting-point. Even if the coordinates in the database were wrong, the codes would nearly all be correct.

You can very easily create a code within 8 digits which will encode the street and locate the delivery point. To get you to the location, the map would need to tell you how that street relates to other streets, but it doesn't have to be plotted completely accurately in terms of coordinates. (2 digits for the county, 3 digits for the small area, 1 digit for the street segment within the small area and 2 digits for distance along the street). However, to code each individual delivery point or building, I think you would probably need to survey or manually plot locations for accuracy, over a period of time.

spideog7

antoinolachtnai wrote: »

A classic postcode would not have this issue since it would be based on postal operations. (This has many other drawbacks.)

That doesn't help you get directions there unless you're a postman or you follow a postman.

antoinolachtnai wrote: »

A post code based on electoral divisions and small areas would not have this problem, as every building would clearly be within one or other electoral division or small area. There are some exceptions, but not that many. There are existing databases. They are not perfect, but they are a very workable starting-point. Even if the coordinates in the database were wrong, the codes would nearly all be correct.

How is being accurate to a small area more helpful at getting you correct directions than being accurate to within 5 metres. It seems these solutions still rely on your navigation method, be it a map, GPS etc. and are still open to the routing algorithm making a mistake.

antoinolachtnai wrote: »

You can very easily create a code within 8 digits which will encode the street and locate the delivery point. To get you to the location, the map would need to tell you how that street relates to other streets, but it doesn't have to be plotted completely accurately in terms of coordinates. (2 digits for the county, 3 digits for the small area, 1 digit for the street segment within the small area and 2 digits for distance along the street). However, to code each individual delivery point or building, I think you would probably need to survey or manually plot locations for accuracy, over a period of time.

But your small area might not necessarily be along a single street. So two adjacent houses with different ways to get to them might be in the same small area, thus you're back at square one be it a small area or a tiny area.

antoinolachtnai

As I stated, 8 digits would give you a lot more than the small area. Five digits would get you to the small area.

The remaining three digits would identify the road segment within the small area (in the urban example, a street) (1 digit) and would indicate where exactly on the road segment (remaining two digits).

The problem with the coordinate based codes has nothing to do with the quality of the routing algorithms. It has to do with the accuracy of the delivery point and road data. If both are not perfectly accurate, there is a strong possibility that the routing algorithm will snap the delivery point to the wrong road and direct you to the wrong place.

spideog7

But if you can go to the hassle of building a database of small area post codes, then you can use that same effort to accurately map everyone's location based post code. Ok it's not as cheap as it would be if everyone did their own but it means that it will be perfectly accurate (and it's probably still cheaper than designing a dedicated system). It's still an open system and it means people can still make their own code which would be in the same format and would work where pinpoint accuracy is not required i.e. a code for a field.

I'm not sure how you'd intend to encode the street using a single digit without an enormous and cumbersome database which listed every possible street within every possible small area and labelled them. Again you're basically putting in a huge amount of effort to make a system which is designed to counteract the inaccuracies in the road data that the routing algorithm is using when half the amount of effort would probably just fix the road data in the first place.


I don't think this is really an issue for urban areas as they will still have the street name as part of the address, rural areas only have the town land, which typically isn't very helpful. By the way I'm not really talking about a system for An Post here, theirs works just fine, but rather a generic system for the multitude of other uses.

I also don't think we're going to be able to agree on this, I just think a location based system is far better because it requires no knowledge beyond the algorithm used to create it. In a day after tomorrow scenario where all we have left is a pen, paper and a map, we'll still be able to use this location based code. Plus it's based on latitude and longitude which don't readily change, small areas are too dynamic and the system would require constant refinement and maintenance.

antoinolachtnai

spideog7 wrote: »

But if you can go to the hassle of building a database of small area post codes, then you can use that same effort to accurately map everyone's location based post code. Ok it's not as cheap as it would be if everyone did their own but it means that it will be perfectly accurate (and it's probably still cheaper than designing a dedicated system).

It would be very expensive. It is hard to see how you could survey this for less than 15m euros.

It's still an open system and it means people can still make their own code which would be in the same format and would work where pinpoint accuracy is not required i.e. a code for a field.

A system based on small areas could equally be an open system and I very much agree with you that the system should be open.

You can also come up with your own code for the access point to a field with the system I described.

I'm not sure how you'd intend to encode the street using a single digit without an enormous and cumbersome database which listed every possible street within every possible small area and labelled them.

This database already exists. The cost of recoding it as above would be less than two or three million euros for sure. That would include extensive checking using satellite and maps.

Again you're basically putting in a huge amount of effort to make a system which is designed to counteract the inaccuracies in the road data that the routing algorithm is using when half the amount of effort would probably just fix the road data in the first place.

It is an awful lot less effort than getting accurate coordinates for every premises and home in the country.

I don't think this is really an issue for urban areas as they will still have the street name as part of the address, rural areas only have the town land, which typically isn't very helpful. By the way I'm not really talking about a system for An Post here, theirs works just fine, but rather a generic system for the multitude of other uses.

That is what I am talking about too. Small Areas have absolutely nothing to do with how An Post delivers their mail.

I also don't think we're going to be able to agree on this, I just think a location based system is far better because it requires no knowledge beyond the algorithm used to create it.

This is untrue. You need a lot of geographical databases to make any use of it at all. You need an accurate road map to give meaningful context to the code.

In a day after tomorrow scenario where all we have left is a pen, paper and a map, we'll still be able to use this location based code. Plus it's based on latitude and longitude which don't readily change, small areas are too dynamic and the system would require constant refinement and maintenance.

Small areas are not dynamic at all. Small areas are designed so that they will not change.

Any system is going to require constant refinement and change, as new areas are developed and access and delivery points change. Change is something you have to factor into any system.

In the DAT scenario, coordinates-based system is very burdensome to use with pen and paper and without GPS. It would take 10 minutes or longer to convert a code from the condensed post code format into latitude and longitude and there are many possibilities for errors which will result in an incorrect conversion. It would then take another 5 or 10 minutes to plot the location on a map, if you had an accurate map. It would be very difficult to design a system of signposts to help people find their way to a particular code. It would be very difficult indeed to sort the mail using this code (let alone avoid zombies and make contact with other survivors).

It requires a great deal of patience to plot the exact latitude and longitude of an isolated point on an island with cloudy skies. You could map off reference points, but the math required to convert from whatever local grid system you adopt to true longitude and latitude is pretty big and complex.

spideog7

antoinolachtnai wrote: »

You can also come up with your own code for the access point to a field with the system I described.

Not if you don't have an accurate map/database of the small areas.

antoinolachtnai wrote: »

Small areas are not dynamic at all. Small areas are designed so that they will not change.

Aren't small areas designed to have constant population? Population is dynamic, imagine the changes that would have taken place if these small areas had been designed and laid out before the Celtic Tiger and the one off housing boom. We can't anticipate the future, while it's unlikely that kind of situation will happen in the near future it is more likely that the opposite will happen and we will begin to see greater urbanisation.

antoinolachtnai wrote: »

In the DAT scenario, coordinates-based system is very burdensome to use with pen and paper and without GPS. It would take 10 minutes or longer to convert a code from the condensed post code format into latitude and longitude and there are many possibilities for errors which will result in an incorrect conversion. It would then take another 5 or 10 minutes to plot the location on a map, if you had an accurate map.

You can use a database for the location based system also but if you don't have that then you calculate it, worst case scenario. It's not ideal but it is possible. Without the database of small areas this is basically impossible. Once a grid system is mapped to lat/long then you don't need to do it again.

antoinolachtnai wrote: »

It would be very difficult indeed to sort the mail using this code (let alone avoid zombies and make contact with other survivors).

Zombie's can't do math (commonly known) therefore the location based system is zombie proof :pac:



Out of interest how are these small areas indexed? Is it purely based on numbering, I mean can you infer that two small areas are adjacent.

Either way we should adopt some unified system before commercial interests come up with a multitude of different incompatible systems.

antoinolachtnai

spideog7 wrote: »

Not if you don't have an accurate map/database of the small areas.



Aren't small areas designed to have constant population? Population is dynamic, imagine the changes that would have taken place if these small areas had been designed and laid out before the Celtic Tiger and the one off housing boom. We can't anticipate the future, while it's unlikely that kind of situation will happen in the near future it is more likely that the opposite will happen and we will begin to see greater urbanisation.

No, it is not intended that they change to meet the population. I guess it might happen that they might be split into two, but the boundary would stay intact.

You can use a database for the location based system also but if you don't have that then you calculate it, worst case scenario. It's not ideal but it is possible.

In practice though, service providers want to deliver to verified addresses. They don't want to send drivers out to coordinates. The best way to do this would seem to be to have a database.

I accept that there are times when you just need to specify a location which is away from boundaries. But I think this is quite a different requirement to delivery to a point.

I think this requirement could be met using MGRS (Military Grid Reference System). This system is already incorporated into many gps units, so there is no need for any new software for it, and it is obviously global in scope. (See http://answers.google.com/answers/threadview/id/556142.html)

Without the database of small areas this is basically impossible. Once a grid system is mapped to lat/long then you don't need to do it again.

The database of small areas should be open for sure.

Out of interest how are these small areas indexed? Is it purely based on numbering, I mean can you infer that two small areas are adjacent.

I am not clear on what the numbering is. I do not think it will give a readily usable code. You would have to renumber them so that you could infer that two small areas which have similar numbers are close together.

Either way we should adopt some unified system before commercial interests come up with a multitude of different incompatible systems.

The practical problem is that you really need to tie to an address database to make any system feasible.

NFD100

What's happening with this? Wasn't the tender to operate the system supposed to have been awarded on 1 September?

thirty-six dragons

In the OpenPostcode (http://tinyurl.com/openpostcode), the first two characters represent a zone about 13km by 17km. Enough for general statistics, marketing calculations, to give some a general idea of where you live, etc.

The first four digits together make an area-code of about a half-kilometre squared (.46km x .60km, 27 hectares) - suitable where privacy is important. It acts as a general description of an area - for statistics, service-available calculations, and a rough idea of what part of the city or country someone is living in. It is good enough as an address validation code on a letter or online form. For many rural locations it may in fact be postcode enough for a single address.

The first six digits together make a site-code, an area five times wider that the complete code, approximately 16m x 21m, which is enough to mark a single address in many cases (certainly larger premises and campuses will not have to use complete location codes). It is enough for credit and other address verification checks. It is enough in most cases for navigation.

With all digits in place accuracy to extended to its full +/- 1.6 and 2.1m - precise enough to indicate garden sheds, garages, entrances, postboxes, greenhouse, indeed any geolocation information.

e.g. JVGB-PK3-P
http://tinyurl.com/openpostcodegoogle/?opc=JVGB-PK3-P

Zone: JV (general statistics)
Area: JVGB (anonymous address pointer)
Site: JVGB-PK (immediate area)
Location: JVGB-PK3 (spot-on at a particular place)
Confirmed Code: JVGB-PK3-P (including the final check character)

garydubh

bk wrote: »

Yes, I'm afraid I'd prefer that vanity addresses such as 999 weren't possible.

But it isn't a big deal, I didn't realise your system offered greater fine level detail then loc8, that is obviously a big advantage.

But in the end, I believe the openness of your proposal is the biggest advantage.

but let's take a closer look before you make up your mind ...............



Wasn't it Brian O' Driscoll that said: 'Knowledge is knowing that a tomato is a fruit; wisdom is knowing not to put it in a fruit salad"

This would appear to suggest that it is very often easy to just repeat a statement but, without the requisite understanding or practical know-how, actaully fail to interpret or use it properly.

This could be said of some of the statements made on this thread over the last while regarding the accuracy and nature of location codes.

As opencode does not control the source and quality of the location information being used to create the codes;- then it most definitely cannot claim any "greater fine level detail" than others or even any related accuracy at all! In fact the truth is that the accuracy of an opencode is in truth "uncontrollable" or "unknown" and not +/- 3m as claimed! Its accuracy is at the mercy of the contributing Lat/Longs. Those created from Google Maps/Earth can be in error by up to 30 or 40 metres in parts of Ireland (image ground truthing and coverage issues). Lat/Longs published on the web are often kilometres out and Lat/Longs without the associated datum taken into account can be multiple kilometres in the wrong place. (multiple datums exist and Lat/longs can be created in different datums depending on the source!) Not to mention that Lat/longs come in multiple degree fraction formats and the average user does not know one from the other and in the confusion can create a code 10's of kilometres in the wrong direction. Lat/Longs generated by GPS receivers can also have significant postion accuracy and datum related errors. The understanding and control of all of these potential error sources being a classic case of how Wisdom is separated from Knowledge and how professionals are separated from opportunists. Quite honestly anything calling itself a postcode and at the same time not making any effort to control the quality of the source from which the postcode is to be created is more than a little "fruity" to say the least.

It is for these reasons together with 2 years of beta testing and 31 years of related experience that the Loc8 Code process does not allow anyone to enter Lat/longs or Grid coordinates on its public site in order to prevent the generation of Loc8 Codes of dubious quality.

To be very clear, the source of Loc8 Codes is absolutely controlled and in the main limited to www.loc8code.com where OSI/OSNI mapping is used and a code must be generated at 1:1000 scale where the map precision is in the order of +/-1 metre. Therefore, a validated Loc8 Code created in the site and used via a web application (the only way that an opencode can currently be used so a direct comparison) is potentially +/- 1m accuracy and therefore better than any other Location code promoted in this thread or elsewhere. In a non-web scenario on disconnected SatNavs, (no other has this capability) the validated +/- 1m accuracy web created Loc8 Code has an accuracy in the order of +/- 5.2m on average and controllable as such. That is why the more than 5,000 Loc8 Codes currently appearing on commercial websites across the island can all be absolutely relied upon to get you there.

As you also know, on the web Loc8 uses a system of star ratings to further qualify the reliability of the codes. www.loc8code.com/help and www.loc8code.com/W8L-82-4YK and this is the reason that Loc8 encourages users to validate and take responsibility for their codes.

So opencode suffers from being "open source" - i.e. because of its "openness" it can be created from any Lat/long from any source of any quality and therefore could never be relied upon by emergency services/couriers/service providers or users of any sort;- in fact it offers very minimal benefits over the Lat/long itself from which it is calculated in the first place. Some might say that an error of 30/40m doesn't really matter - well yes it does if people are to find places efficiently including if the destination is adjacent to one-way systems, dual carriageways, rivers, canals, motorways or back gardens of houses in another estate where such errors could mean detours of several kilometres. Equally, if promoters of their own code are claiming unsubstantiated 3m accuracies, then it most certainly matters for that reason also!

It also appears that Loc8 Code has been criticised by one promoter on this thread for being "closed source". It is in fact this "closed source" design that ensures the quality of the Loc8 Code created and allows others to depend upon it. It is for this reason that Garmin have implemented it (having chosen it in a head to head competition between 3 contenders from Ireland including some mentioned on here) and many state and commercial organisations are already using it. "Open Source" or "openness" therefore, just means potentially poor quality and unreliability and is definitely not the "biggest advantage" or indeed any advantage at all by any definition of the tomato!

Furthermore, there also seems to have been a suggestion from the same quarter that Loc8 Code hides the means by which the code is created. This is not the case, the construction of the code is in the public domain for quite some time and it is here for anyone to see: http://www.sumobrain.com/patents/wipo/Navigation-method-using-geographic-location/WO2011073965A1.pdf The process that led to this being published started nearly 6 years ago and not just a few weeks ago!

From the detailed specification you will see that a Loc8 Code is not a random set of characters like others - it is carefully engineered so that certain things can be guaranteed:- i.e. accuracy, predictability, validation capability (helped by being guaranteed alphanumeric amongst other things and definitely a "big deal"), sortable, has qualities of adjacency, no unwanted combinations can be created and is self checking as well as many other qualities. You will also note that Loc8 Code also eliminates characters such as V appearing after V as its looks like a W or V and W appearing side by side unless it is known which is first for the same reason (VW) and for obvious reasons WWW cannot appear either. Control of the use of C,G, H,M etc is also imposed unlike others promoted here to avoid confusion in hand writing and these and other controls favour quality and the elimination of sources of possible confusion over and above the drive for higher resolution. Also possible confusion between 1,L,5,S, Z,2, B,8 etc etc is solved by the unique design of the checker code - again something not done by others and all the reasons that Loc8 is superior and is in no way random like others. Many of the designed-in qualities of Loc8 Codes are synopsised here: http://www.loc8code.com/about/what-is-a-loc8-code In short, Loc8 Code was designed to satisfy certain minimum requirements rather than being just the random output of a basic calculation.

Equally, in the Loc8 Code specification you will note reference to protection against confusion with the BT postcode for Northern Irish areas. Also note the use of dashes to distinguish between and emphasise elements, the use of 8 characters in the code and the introduction of a checker character - some of which were implemented for the first time ever in a location or post code globally by Loc8 and some of which have now also appeared in recently conceived solutions. It is very notable that protection against confusion with the BT postcode was highlighted on 18th July last by me in this thread and very shortly afterwards opencode was adjusted to include this as a feature. Some call this flattery, but too much more of this type of flattery might be hard to take!

Yes indeed, Loc8 Code is protected by copyright (an automatic entitlement) and by a patent priority to ensure the quality of all Loc8 Codes and also to protect the interests of those who have supported and invested in the solution. (Enterprise Ireland included) Those who get to use Loc8 Code calculations in applications and services are controlled because the aim is to protect the quality and reliability of all Loc8 Codes used in the public domain. Loc8 Codes created from Lat/longs (or grid coordinates) that nobody knows the origins or quality of are not permitted. Loc8 Codes must be reliable and that is why Loc8 has invested substantial resources and time in developing, field testing and refining the nature of the code and licensing high accuracy Ordnance Survey Mapping (not free) to deliver them in the first place. So if it is alleged as if a crime in this thread that Loc8 Code is a "protected closed source code" - yes indeed it is and this is absolutely the case to ensure its reliability and quality and therefore there is nothing bad about this at all as appears to have been suggested.

Does this mean that you have to pay to use it - not necessarily - but it does mean that those who can use the algorithm to create Loc8 Codes and how it is used are rigidly controlled but absolutely contrary to claims.........

• Anyone can go to www.loc8code.com and generate a Loc8 Code for any location from the Map to +/- 1 metre accuracy and at no cost !

Do companies wanting to use Loc8 Codes for commercial purposes have to pay ? - yes they do! Why? - well because of the controlled quality of the code and the investment and infrastructure and ongoing full-time support behind it, which guarantees time and fuel savings for them as well as other very significant benefits to their operations.

If you want to use Loc8 Code in any developed solution then just talk to Loc8 Code, as many have done already, and this may not mean that you have to pay - but all users can be sure that Loc8 will control the way Loc8 Codes are created so that users from Ireland (North and South) and those visiting Ireland can rely on them to get where they want to go quickly and efficiently.

Finally, I would like to refer to discussion about a postcode for Ireland. Loc8 Code is not currently delivered to Ireland as a postcode. It could be and has been offered as one but in the circumstances of it being a National postcode it would be delivered to the country in different ways. Loc8 Codes are currently offered as high quality, reliable location codes which can offer efficiencies to many in the Logistics, Tourism and Emergency Services businesses. In the event of them becoming a national postcode, each property owner/occupier would be delivered their Loc8 Code manually as well as electronically using An Post's GeoDirectory (now subject to postal liberalisation rules) as the basis. The cost of this would be substantial but this would be costed into the service implementation and the transfer of related ownerships to the state. (An Post companies are already using Loc8 Codes )

However, it must be said that postcodes are a limited and dated technology (only property and groups of propreties related) and the current Department of Communications stalled plan still proposes to have a postcode which will only define the centre of up to 50 properties (up to a whole townland in many areas) - and really would be of no use for finding places. Loc8 has a signed letter from the previous Minister indicating that the planned wide area design would not change (and is still the case) and it is as a result of that emphatic letter that Beta PON Codes and later the full release Loc8 Codes (after 2 years of field testing) were launched as a precise alternative to the stalled and illconceived National postcode plan. It is worth noting that the recent associated legislation had the words "precise" and "unambiguous" removed from the final definition of the proposed National postcode! As the plan initiated in 2005 has not really progressed much since then (probably because it has little to offer potential users including An Post as it was designed solely for mail sorting which is a diminishing industry), implementation of a national postcode really is not expected for several years yet!

1.8 million Loc8 Codes have already been allocated to addresses in the GeoDirectory for some commercial users but Loc8 Codes right now for the general public are elective;- you create and use them if you feel they are going to be beneficial to you or to a service provider on which you are depending - they are not imposed on anyone and there is no cost to create one! However, Loc8 does invest a lot of resources in making sure the means by which Loc8 Codes are created are as precise as possible. This makes "closed source" a good thing in this case rather than a bad one as some seem to have suggested on this thread.

Loc8 Codes are the quality "fruit" of proven research, design, testing and an implemenation plan with investment, standards and infrastructure behind it and are the only location code in proven and active day to day use by real users (including state agencies) on the island of Ireland.

If you have any questions about Loc8 Codes - rather than speculating and potentially getting it wrong - just ask at info@loc8code.com

recedite

garydubh wrote: »

Also note the use of dashes to distinguish between and emphasise elements

I discovered by accident that, when punching loc8 codes into the Garmin satnav, I can use a "space" instead of a "dash" in the code. The advantage to this is I don't have to keep changing from the querty screen to the symbols screen where the dash is found, therefore quicker and easier.

If I am going to some obscure place, I look it up on the OSI map provided on the loc8 website the evening before, in comfort at my PC, then write down the code generated there. Next morning I get in the car, punch in the code, and set off early for a hassle free journey. I can also text or phone the code if I want to meet someone there.
I have to say, I find it's a great free (to me anyway) service. Well done.

thirty-six dragons

GaryDubh is being deliberately incorrect and confusing about the maths and about mapping.

http://tinyurl.com/openpostcode

The OpenPostcode, freely and easily, using an open free-to-use algorithm, converts any location data, however accurate, into an easy to read, write and use, hierarchical postcode - accurate as a representation of the location coordinates to within a 3 metre radius. No other code does this.

But being Ireland, we will probably adopt the worst performing code and the one that create layers of unnecessary bureaucracy and profiteering.

There is nothing wrong with the calculation that the OpenPostcode much more accurate an algorithm than Loc8codes. The calculations are on site at http://tinyurl.com/openpostcode - please feel free to find errors or corrections to the code. It is open. It is not "fruity".

What is "fruity" is to claim that anyone can make a loc8code to +/- 1 metre accuracy. This accuracy does not exist in the code. +/- 6 metres is the stated accuracy on the site, though it is not sure how much this fails as you move down in latitude. Of course, it is a closed code.

Try comparing the two codes. Don't forget the GoCode - another closed-source variety.

To claim that "open" is a failing is truly amazing and I doubt anyone sees this as any more than a desperate attempt to hold water in your hands.

However the GaryDubh boy is right. GPS coordinates can be incorrect. So what? That's not the purpose of the code. The OpenPostcode algorithm converts coordinates to a hierarchical, easy to read and understand and compare and remember and write, code. The proof-of-concept website uses Google Maps to prove concept. But of course you can also get your coordinates from the OSI maps as Loc8code's website does (see http://maps.osi.ie/publicviewer/). You can use your SatNav with +/- 3 metre accuracy (at least with my Garmin anyway when it has nearly all its satellites and mine is ancient). Or, for the future, the government can use the Geodirectory coordinates as used by AnPost.

No one bar GaryDubh understands why a postcode needs to be verified. I created one with Loc8code with as little accuracy as the OSI map could supply (it draws an odd representation of my actual house). I arbitrarily chose a location in my yard; it could have been way down the laneway at my gate; but I chose the location. And then I verified it. Woop-di-do.

There is no stated way in which a Loc8code would in reality be anymore "verifiable" or accurate or effective. It would however be less useful - it wouldn't be convertible to coordinates by programmers and users. The conversion is a closed algorithm. It is not a geolocation code in fact at all. It is only a geolocation code if used with licensed products.

I, nor any other ordinary user, is permitted to convert a Loc8code to coordinates or to create a Loc8code in a database, spreadsheet or other program (except using the Loc8code website or other licensed products). There's no way I want my address organised like that.

I, nor any other ordinary user, is permitted to sort Loc8codes logically, to create navigation routes, to use Loc8codes for statistical or marketing purposes - to us it will always be nothing more than a random assortment of letters and numbers.

By all means in a democracy let people feck up their postcodes if they wish. They just won't get their mail and their houses will burn longer waiting for the firemen. People feck up their addresses all the time and there is no law against. And there is no law which says which 3 metre radius circle is the exact location of my address: I decide.

And as for people who continue to talk about postman pat and his little route along a street and how that can make a postcode. GO DO IT! Come back to us when it is done. Meanwhile postman pat is already using his own method based on the Geodirectory at AnPost. http://geodirectory.ie/GeodirectoryMap.aspx

Postcodes/Geolocation codes are NOT navigation - they are points, not vectors. Navigation is for something else to work out and as we well know navigation changes (even from hour to hour and depending on how you are getting there and whether you want to use toll roads or not, whether you are walking across mountains or not, whether you are cycling and avoiding hills, whether a new road has been created, whether an old road has been removed or blocked or whether a one-way traffic system is changed, as they so often do). It cannot be fixed into points.