Systemic failures that seem to be minor, but point to a larger problem or issue

Here is a chance to discuss minor SYSTEMIC problems that have irritated you, but may have been dismissed by others as trivial, or perhaps you have been told that it can’t be changed.

This is not about individual businesses or people. These are things on a larger scale, so it might be industry wide for example.

To start the ball rolling, I would like to highlight the Australian Government’s inability to enter anything more complex than the 26 letters of the alphabet and numbers into any of their data bases. Now this may not seem to be a problem to people from English speaking backrounds, but sometimes you can’t even enter an apostrophe in a celtic surname such as O’Neil. The databases certainly can’t handle the variety of accents, tildes, diaeresis, umlauts, etc. that are used in many ‘western’ alphabets.

Languages that use other than the western scripts add a layer of complexity to to the mix because they need to be first translated into our western script first, and then recorded to accurately reflect the correct name.

Given that Australia is a land of migrants with about 29% first generation emigrants and about 21% being second generation emigrants in 2020. Australia’s population was 25,694,393 people at 31 December 2020. As England only accounted for 3.8% of our emigrants, this means that up to say 46% of Australia’s population could have a foreign name that potentially cannot be accurately recorded. In other words there could be almost 12 million peoples’ names recorded incorrectly.

So you think we are a multicultural society? Think again.

Any others??

6 Likes

Well Unicode has been around for a long time in computer systems and is the standard encoding in many modern languages like Java. So there is no excuse at all for not handling language specific characters such as you mention, including the acute in your name.

I have a problem every year doing my tax using the My Gov system. It happily prefills the account number and interest paid on one of my bank accounts, but then at final calculations decides it doesn’t like a dash in the account number and throws an error.
That is just basic validation silliness.

5 Likes

Well yes it could be that many. It could also be a lot less. You give no way to determine how big this problem is. For a forum that regularly pings writers for using language like “up to 90 % off” I don’t know how you typed this one.

Nonetheless you do have a point. It isn’t that hard to write software that allows extra characters. There are some problem though if you think this will be a big revelation and change. One is the degree that those whose native language (using the basic Roman script) use special characters have adapted to using the nearest standard equivalent. Do you know how many people that this bothers or matters to?

Also what proportion of users know how to get such characters out of an English language keyboard? If it bothers them you might expect the owners of such names to learn how but what about those who have to transcribe them or hand write them? Another question, that I don’t know the answer is to, is to what degree are the keystroke combinations for these things standardized across systems and applications?

So while it is technically feasible and might be nice for some people who are very keen on maintaining their ethnic heritage we don’t know if it is a sizable problem nor what it would cost to solve it.

1 Like

Accepting them is easy but the overall costs can be significant. Consider sorting orders for each unique alphabet and language as the tip of the iceberg, not counting those languages that use pictograms or script characters rather than letters.

Some are

https://www.secureclub.net/Help/acm/UG/ug_Chars.htm

and there are some embedded functions English speakers usually are never aware of since they don’t need them. Japanese would buy Japanese keyboards as would many of the other linguistic groups, but there are ‘helpers’…an example

and on it goes.

edit:

3 Likes

I can see that sort order might be a problem. This looks to me to be another aspect of the standardization question. If you need to pass data from one system (possibly on another platform) to another you had better have all your ducks in a row.

1 Like

Unicode was designed and is globally specified to address issues like sorting and cross system compatibility.

2 Likes

That is true but ‘it’ is not ‘free’ in practice.

https://unicode.org/reports/tr10/

2 Likes

Nice link, but after reading it my brain hurts. Panadol time. :grimacing:

3 Likes

I can relate to the observation.
Which problem are we trying to solve?
It might be ‘Tilting at windmills’ if my experiences of family history searches are considered.

Excluding errors in the records, of which there are many, English efforts to standardise names have been with us for several hundred years. Despite the best efforts of those in public service failures abound. Even for common names of English origin.

EG Are you a Mc or a Mac, a Haig or a Hague? And are you one who enjoys ‘ghoti’ Fridays.

For names derived from other languages or written languages the common alphabet of English has always been a limitation. There are however conversions based on the creation of new or alternate phonetic interpretations of the 26 letters in the modern English alphabet. There are many examples now taken for granted, from the names of WW1 battles named for local villages, the use of X in names of Chinese origin, or ‘wh’ to represent a Maori language ‘f’ sound.

If I had a choice, it would be to keep just the 26 + 10 common alpha numeric convention. For names that are not naturally written using the English 26, it would seem simpler to convert to the common 26, and add an extra box where one could note which conversion or language of origin applies.

P.S.
For any wondering about other options, there may be a lesson from Japan which has 3 different systems of writing, only two of which are phonetic. Many systems add the English alphabet and numbers to the 3 existing Japanese. More is not less.

International agreement for Passports is to include key details in English or French as well as the language of the issuing nation. The identifying details need to be machine readable. The conversion of names to suit the limitations of 26 characters is established.

1 Like

Without being seen to be making light, I empathise with Tamás and understand his lifelong frustration dealing with our English-character-centric world. Yet even when there is no linguistic issue and someone has a single character name, say ‘K’ not ‘Kay’ nor ‘Kaye’, the minds that validate names cannot conceive of such being right so invariably turn it into an initial and always print it out as ‘K.’ not ‘K’. K can then expect to be questioned further, although K is not in the same league as some of the absurd (to many of us) names some billionaires and celebrities bless their offspring with.

Some find the foibles irritating and others find them amusing or challenging. I can imagine the programmers who ‘own’ how it works often tip their glasses in self satisfaction of what they have wrought, and giggle.

1 Like

One that always intrigues me is the pronouncitation of the US states of Kansas and Arkansas.

Same 6 letters but totally different pronouncitation.

image

2 Likes

One failure I noted was the assumption that everyone had a middle name (second given name). When Qld required people to have a Blue Card before interacting with children, the form required at least 2 given names and a surname. If no middle name was entered the form was sent back ad nauseam to be ‘corrected’. It was only in recent years that the form has a tick box for ‘no middle name’.

3 Likes

Weren’t there legions of Queenslanders whose official name was of the form John None Smith?

2 Likes

My father only had one given name, and that plagued him for life. My mother suggested he invent one, but cross checking with other records meant the first attempt was rejected with a request to “provide your middle name” even though (presumably) cross checking revealed no middle name.

Qld Blue Card applications were returned to the sponsoring body eg Scouts, youth group, school, for correction. The applicant might not know the status of their request as it bounced back & forth to (for example) Scouts HQ in an endless “provide middle name” loop.

1 Like

Data validation of names is a minefield and too many programmers are not up to the task. The worst attitude is to make assumptions about the scope of names based on your own experience. Until you have spent some time examining names from a wide corpus you cannot imagine the strangeness that can come about in real life.

Just a few kinds of errors:

  • Rejecting Ng or other family names as too short
  • Rejecting Featherstonehaugh or Balasubranamiamnanananana etc as too long
  • Rejecting composite names; eg the surname “The Red”. Yes I have met Eric The Red, he told me he often used Thered to avoid trouble.
  • Assuming uniqueness. It isn’t just John Smith who has many twins. In a large city like Sydney you will find several men with the names George George, George Georges, George Georgamlis, George Georgiades etc. You will also find groups such as Salim Bechara, Salim Salim, Bechara Salim and Bechara Bechara (all unrelated) living together. I once had 30 odd staff and two had the same name, fun for young and old, I won’t name it as both are still around. One had a middle name so we started to use it, the other didn’t so we manufactured one for him. When you refer to people in a meeting as either Wallace Trimble Peabody and Wallace Alligator Peabody each time you get funny looks.
  • Speaking of middle names I knew a man called Syd Bridge. Yep his name was Sydney Harbour Bridge. Really it was.
  • And the old favourite: if somebody gives their name as Van Fan Tran which is the family name? What if their English isn’t sufficient to tell you?
1 Like

My other half used to put in a ‘.’ (full stop) where middle name was mandatory on a online form. In most cases it worked as the field wasn’t set for alphanumeric characters only.

Systematic failure from one of our neighbours. She lost her husband about 12 months ago and has had no end of trouble getting his name removed from joint accounts (bank, electricity, insurance etc). Even after sometimes contacting several times to successfully instigate the change, she still gets marketing materials from some of the organisations in her late husband’s name. It appears the marketing sections use either different contact databases or old versions. Every time a letter is addressed incorrectly, it brings back last year’s loss.

1 Like

I totally agree, but I can’t think of any way to quantify this.

The problem is the blinkered approach that has been taken when developing Government forms and databases which does not allow for any diversity outside the 26 alpha +10 numeric characters. I do not blame the programmers for the strictures, but more the senior eschelons of the corporate structures who have not considered it as something that should be incorporated as a matter of course.

Many of the characters can be readily represented using ASCII, much as we did back in the pre-Windows days. The problem is that very often these ASCII characters are rejected by the systems, probably because they are on mid or large scale systems which work with different operating systems and database software.

If you remember the bifurcated option of male/female. No other option was ever considered. Nowdays there are a number of other choices people can use to identify themselves. So I have hope that one day things may change.

It probably is. Someone has to be the first to speak out about any issue. I don’t expect that there will be any change, but I hope that by raising the issue more people will be aware of this.

Coming from a first language that is totally phonetic, the vaguarities of the English language is a constant source of amusement.

2 Likes

I have rarely heard it put so politely.

2 Likes

I was just lucky, I think! I have no middle name and when applying for my first driving licence I was told firmly and clearly “no middle name, no licence”! I then proffered my confirmation name and this was accepted so now I have some documents showing a middle name and some not. When I attempted to delete this middle name for my last licence renewal (hoping for parity across the board) I was refused! Ah well I am still me…

3 Likes