Sure, the software has a bug. Software does. Almost always. What is broken here is that it is possible to get stuck in an airport with no possibility of escalating sensibly to a chain of human beings with agency to resolve it. You could have a software process which works 99.9% of the time, and have staff at hand which could solve any solvable remaining issues like this one. They call the Home Office internal hotline, explain the issue, resolve it by manually interfering on the spot (and have a bug filed), and everyone carries on.
Instead, we demand a 100% correct system, and save money on actually having people around who can deal with the edge cases and make sure everyone has a good time. Of course, getting a project from 99.9% to 100% costs a lot of money, because that includes literally all edge cases ever (getting to 100% is as likely as reaching ∞ by multiplying any given starting number by 2 until you reach it), and so it will likely end up being more expensive in budget overruns than the 99.9% plus humans system.
Obviously, politicians must choose for the fictional 100% case, because they can't afford the cheaper 99.9% plus human one.
Similar things happen when you require fingerprints or retinal scans. "Well, I don't have a right hand anymore. It's in some hospital's incinerator."
And not having an override or bypass is a software issue. Designing a system that doesn't permit exceptions or overrides is bad software design.
I think his point is that the designers of the software would have been aware of this, but they weren't the designers of the system, who wanted software to do All the Things.
You cite three meaningful changes in biometrics: disfigured, loss of fingerprint and retinal issues.
In all three cases surely the right and proper thing do to is to get your details updated on the system ? It may even be a legal requirement for you to do so in the case of official ID documents.
Of course, if you have a temporary mismatch due to being injured on holiday that's different.
But for a permanent thing I just don't see why you would not just get stuff re-issued or updated.
> Monique Hawkins, the acting chief executive of the immigration campaign group the3million, said it had seen previous cases similar to Webb’s, adding that it highlighted vulnerabilities in the system and it was “crucial that people have a genuinely stable means by which to prove their status”.
> She told Politico the eVisa rollout had been “rushed”, adding: “We regularly see people not only denied boarding, but also compensation, as carriers and the Home Office play a blame game shifting responsibility between each other.”
"Oh, that's an edge case. 0.1% of users will encounter it. P2 bug, not a ship blocker."
Previous company had full payments/accounts team, a few QAs, a couple accountants and SRE working on the automation of a specific payment edge case that was normally handled through support, and was taking 5-10 mins to handle per case. Whole thing took four-five months.
In the first year, less than 100 people used it, out of a few million. It was B2C, so after taxes, that amount was maybe enough to cover a couple days of lunch for the people who worked on it.
The vision that it evokes is terrifying!
What's really stupid is not being allowed to get on the flight at all.
Like: she's not labelled as a "terrorist", so just let her on the flight as she's not likely to make the plane explode, and let the customs folks take her as a one-off to be dealt with on home soil. Why is it necessary to be pre-approved to even get on the flight?
This is another instance of the usual hubris of management in thinking that automation has solved all the problems without creating any of its own, therefore all the humans are now made redundant.
This may be possible some 15+ years ago. Home Office today is very different. Without a prober procedure written down I doubt this could be solved manually. Nor would anyone want to solve it manually.
This is extremely common
I don't think it is good enough for either, but I also think we should practice, and run our companies, as we preach.
This is a fundamental problem with "running government like a business," one of several that point to the fundamental flaws of the entire idea (profit motive, unitary unit of execution, etc.).
Why can't the same system work here?
so yes, they can and do override the system.
Also the person wasn't stuck in the airport, they were denied boarding of the plane because there was an issue with their identification documentation.
It's not good, its a travesty that you even need an ID to fly between two european countries, let alone a visa, but the picture you paint is at odds with reality.
I seem to recall a time when you could ring up, get through to a human that could actually solve your problem.
Now humans are just front ends for software, and there's noone to take responsibility.
I would put the blame more of systems and cost saving. They don't want humans to have any responsibility, or pay them for that.
For example, I went to the bank down the road and tried to withdraw a few thousand dollars. This isn't atypical for me - I do this every few months. I had my ID and knew my account number, but I didn't have my debit card. They essentially told me "system says no" and that there wasn't anything they could do. This is a feature for the bank - a human teller might be talked into pressing the wrong button by fraud. Instead, I had to go home and get my debit card. By presenting that bit of plastic, system said yes, and I was good to go with my cash.
For another example, I was recently at an airport listening to a family struggling to get on a flight. They were checked in with seats assigned (confirmed by the gate agent) but the system refused to give the green "ok to board" light for them. Rather than just put them on the flight and file a ticket to fix the record after the fact, the agent just stood there churning for a while until a supervisor was able to sprint over from the other end of the airport. The supervisor didn't use their superior access level to override the broken system. Instead, they used their superior knowledge on how to push the system through its bug. They kicked the family off the flight, unassigned their seats, put them on standby, approved them for (random) empty seats, and then let them board... but now they weren't sitting together in the seats they had before. I ended up missing that (extremely delayed) flight for boring reasons related to ongoing engine problems and upcoming connections so I don't know how it worked out but I hope they got to sit together in the end!
As a someone in tech, I feel this a classic case study of how NOT to roll out a release and develop a solution.
The release was chaotic, none of the airlines staff were trained on how to use this. So they still insisted on the expired cards and then picked on the expiry dates.
Next, almost everyone makes you generate a verification share code in front of them by logging into the e-visa website even if the rules say I can print it out and just use the code. It takes a few minute to get on the website and generate one.
This solution assumed that everyone has access to a smart phone and a good wifi connection to make this work.
Whoever built this didn't test it on the ground.
So how many people did not successfully use it? Could be 100 million for all we know. Do they think just throwing out a large number without anything to compare it to is somehow proving something?
I had recently a case where my title (Mr) ended up as my middle name in the system of the operating airline (it was perfectly correct in the system of the airline that I booked the ticket through). Two hours of talking to those two airlines did not help me. I had to buy another ticket...
I think this starts to remind me the movie Brazil...
They'll have paid one the many software/technology companies to build this thing. Who will be hiding their shoddy crap behind their client (the government department that paid for it). There are some patterns that I'm sure would start to arise from naming the consultancies that are churning out crap.
So basically both US multinationals (Accenture have a Dublin HQ as a fig leaf)
I was losing my sanity over all the refusals (app refusing the photos).
I have had to take photos of each of my two kids at six months old for passports. That was not a lot of fun (though it's nice to look at the baby photo for the older one now).
The hilarity of having to compare a baby picture with a 5 year old kid stood in front of you is a separate problem.
Now I live in the US.
The UK government has decided that to travel to the UK, a UK citizen MUST use a UK passport. They also, helpfully, don't offer passport services at their US embassies and consulates.
So the process is this: I begin the process on gov.uk, and validate my identity to them, to their satisfaction, using questions only I should know (where did my parents marry, born where and to whom, what time of the day). Once I have done that, I get to upload a passport photo of myself... and then I have to find, and nominate another UK citizen who -does- have a passport, -who is not related to me-, who "has known me for two years", and "has the risk of loss if they were to make a false statement", to go through the same process themselves, validate their identity, and then state that the photo they have seen is of me, that they have known me for two years or more, what the nature of our relationship is, and more.
Then the UK will issue my passport.
Fun story of the app refusals though, id.me, as used by the IRS:
"Scan the front and back of your Driver's License."
[upload scan of front of DL @ 200DPI]
"Unable to find a face in the image you uploaded."
[upload scan of front of DL @ 300DPI]
"Unable to find a face in the image you uploaded."
Huh. Maybe I'll try with a lower resolution.
[upload scan of front of DL @ 72DPI]
"Thank you, now please upload the back of your Driver's License."
Hmm, 72DPI worked for the front, so...
[upload scan of back of DL @ 72DPI]
"Unable to read a barcode in the image you uploaded."
[upload scan of back of DL @ 200DPI]
"Unable to read a barcode in the image you uploaded."
[upload scan of back of DL @ 300DPI]
"Thank you for verifying your Driver's License".
I myself recently had a 10-minute or so delay due to a visa issue. I was departing from Johannesburg and was going to use my Italian passport for entry into New Zealand. I had acquired an NZeTA visa about 15 months before but didn't think to consider until checking in that it was against my previous passport number which I had since renewed.
When the agent said they were just having to double check my visa (while my wife's and children' were fine), the fact the passport had been renewed occurred to me and I told the agent about it which probably saved a couple of minutes of them having to work out this was the issue. They phoned New Zealand, explained the situation and within about 10 minutes the New Zealand side updated their system and I was able to complete my check-in.
Comedy gold.
Laws against persona; freedoms should be called out as unethical.
Jokes aside, when helping someone I know return to the UK, I wasn't sure if I have a stroke or someone who wrote the guidance had a lobotomy.
Clear as mud.
What's described here is not an engineering failure, but a process failure.
Yes it's a process failure but just like health and safety it is everyone's responsibility.
The real world is messy.
> “I had no idea the issue had been corrected until I was allowed to board my flight home,” Webb said. “Had I not spent hours on the phone trying to resolve the matter, I have no idea how long I could have remained stranded abroad.”
Sounds like there were indeed channels to escalate this.
- a valid “immigration document” which satisfactorily establishes identity and nationality or citizenship.
- and, if the person requires a visa, a visa of the required kind or other permission to travel. This visa, or permission, may now be in either physical form, or a digital e- Visa; or
- a valid Electronic Travel Authorisation (ETA), if the person requires one.
https://www.gov.uk/government/publications/document-checks-a...
I'm glad you said "ethically" because the current US administration is openly admitting that they're making decisions on things like school funding, disaster relief based on whether that state went for Trump or not.
God this happens all the time. Software written by people that don't use the software themselves. It has happened to me so many times, I come across a piece of software with weird bugs or interactions that would be immediately fixed if the developer just spends 5 minutes using the software like an end user would. Dogfood all the things!
My experience is completely contrary: developers typically do care deeply about such things. The problem is rather that very commonly managers actively disallow these poor developers to apply their diligence to the software that they develop.
I really have a strong feeling that you want to participate in the ugly vendetta against software developers that has been going on for decades ("Replace all software developers"; current season: "Replace them with AI").
Don't do that. Rather fight the people who are really responsible (who are often (project) managers or Scrum masters).
Not sure if this situation falls under the "push it to prod and let the users QA it" or if this is truly edge case situations. It's easy to bash and make fun of, but I've been in the trenches of catching what is an edge case for most people but run of the mill for those providing the data to me.
To be fair to the airlines here, if they get it wrong then the UK government fines the airline. This is why they are pretty risk averse, and also why I'm surprised that the screenshot worked.
And now that they are the airlines do not have the system to check them so you have to print the webpage which also explicitly said should _not_ be used instead of the share code...
Back in my day, visas were handled by init scripts. And we like it that way!
Computer says no.
Welcome to the last 20 years of UK public policy communications. Politicians know that people do not question big numbers or have the skills to put them in context.
I remember the May government shouting that "more children go to schools rated good+ by Ofsted than ever before". Me and you would say, well, there's more children alive than ever before, so it's hardly an achievement.
Proportionally speaking most of the failures will surely have been non-crossing uses (electronically proving right to work/receive benefits/not have to ultimately pay/claim on insurance for non-emergency care).
It sounds to me like this number indicates broad success.
The rough estimate seems to be that only about two per cent of border crossings use an eVisa anyway (British passport holders don't and non-immigrating visitors don't).
Even over summer that would probably only mean maybe 2.5 million border crossings, so the other uses are not at the border — they will instead be employment/landlord/benefit/healthcare checks using the online systems.
So for example, renting a new property, starting a new job, applying for a new benefit, or needing non-emergency healthcare etc.
The strongest indicator that the system is riddled with issues is in what they did not say. If they had less than 100 cases where the system initially failed and the issues got resolved within a few hours I'd bet they would brag about it.
p99 is ok for web stuff, not for systems that impacts people in their life, but I’ve seen too often things being dismissed because they are edge cases (which becomes an actual real issue at the scale of a country of course)
Obviously the solution is to accompany facial recognition with DNA tests.
That's a bit of a straw man though isn't it — could there ever have been as few as a hundred problems in a system of this complexity with four million users?
(It also moves away from your "hundred million people" somewhat does it not?)
I am sure it has not been a flawless rollout but I don't really understand the outrage here.
It's worth understanding that much of the news media coverage of this is tinged with anger, regret and disappointment: this is a system we would not have needed in this formulation if we had not done Brexit.
Small problems with existing immigration and border systems would not have made any news but they could also have been happening at a higher rate, yes?
We shouldn't have done Brexit, but we do need a system, and therefore we have a system that is less joined-up than it could be.
Either way, the fact that it has been used ten million times successfully strongly suggests it is basically doing OK, since it's hard to imagine many more than ten million uses since the alternatives became unavailable.
Glad to have your obviously authoritative review of the subject, though. I am suitably chastened for trying to show the broader picture.
The government did not even stop issuing physical visa stickers until the beginning of last month.
So again, this ten million figure does not seem to indicate a likelihood of massive, systemic failure.
If there has been massive failure, then I guess the government department and the two US tech multinationals (Accenture's global HQ is only in Dublin for tax purposes) will have to figure it out.
In this case I suspect that either there was a recent mix-up or that person hasn't travelled since she came to the UK or got settled status and never noticed the problem.
EU citizens in the UK since before Brexit only got an eVisa when their status transitioned to "Settled status" for instance and never had anything in their passports. But as long as their online account is up-to-date (i.e. linked to current passport) they can even use the automated passport check gates on arrival into the UK. Only issue since ~last year and especially now that ETA is mandatory for non-residents is to prove status to airline staff on departure otherwise they deny boarding, this is done through online account.
In my experience with large regulated systems, the engineers writing the software don't have the expertise to even know what the correct specifications should be in most cases, and especially in edge cases. Software engineers are not experts in customs and immigration legal requirements and would be relying on heavily on guidance from experts in those fields to determine the proper rules to apply.
In my experience with bugs like this, it is usually scenarios that all parties involved in the design did not consider... especially when implementation is rushed due to tight deadlines.
Other stakeholders typically don't apply such caution.
... I saw last week that they just had their (latest) 25% headcount reduction, followed by their (latest) "executive offsite retreat" to pat each other on the back and figure out what they'll do between now and the next one.
This is like the argument "developers should use the software that they write" (dogfooding).
Let me give one example: the previous person who designed and developed the previous version of the software which I now work on really could use "his" program insanely well. The problem was rather that he was basically the only person who really did understand it sufficiently well to not to be triggered by problems other users had. Yes, a huge part of my job is over the years to turn this piece of software into something that is understandable by other people, too,. :-)
Paranoia?
There are a lot of devs who don't really like what they develope so don't spend any time dogfooding it. Has nothing to do with devs who are blocked by management.
That said I have also encountered lazy engineers who are totally fine with implementing obviously broken software, those are a perfect match for the egotistical product manager.
Note that in my comment I wrote nothing about dogfooding - for a very good reason:
It is well-known that dogfooding leads to software that has rather few bugs, but also dogfooded software often shows a tendency to be "biased towards programmers and power users". So, there often exist very good reasons if the programmers actively don't dogfood their software, and rather trust someone else (who is very trusted by the programmers) that the workflow that the programmer should implement does make sense for the user, even though most programmers would implement this aspect very differently.
If you mean that all humans are related, that is true but a completely useless fact in this context.
Lessons that you should be taking away are rather:
- Every tool that possibly might be good for improving the quality of the software is just a tool that can also cause harm if used in a wrong way.
- Deeply distrust keynote speakers and thought leaders who sell their method for improving the quality of the software or improving software development as a panacea. Every method is only useful under specific circumstances and is harmful of used wrongly. Keynote speakers and thought leaders who don't go into details when their methods are helpful and when they are harmful are simply scammers and snake oil salesmen and should be called this.
