As someone who writes some very thorough unit tests... and also have had to have mandatory training... I find "this isn't a drill you can run" to be _very_ wrong.
It would not surprise me that their back-office operations are likewise economized and some things are just not done because "they can never happen."
Meh, it doesn't even have to be "never". It just has to be cost multiplied by frequency is less than the cost to prepare.
If they lose $100m every five years due to a system failure, and it would cost $30m/year to plan for those failures, they it's just cheaper to let it happen.
And I don't mean this in a judgmental, Fight Club-car recall speech kind of way. It's just business reality. At some point every business has to decide that the cost of planning for something is higher than the cost of letting it happen.
Sometimes businesses end up on the wrong side of that bet. They see only the costs but not the benefits of preparedness (by the time it fails, there will probably be a different CEO in charge) and make a bad call.
The airline industry feels like one where each year it's a different carrier who has some catastrophic scheduling failure. Today, everyone says they're never flying Southwest again. But if you fly semi-regularly then it won't take very long before you don't have any airlines left to fly on.
For people who weren't affected, I doubt very many are even going to remember this. Personally, I remember that this kind of thing has happened recently with other carriers but I couldn't even tell you who.
And people who were affected can mostly be bought off if you need to. Some vouchers & hotel reimbursement and it's just the cost of doing business.
Plus, the airline industry has proven over and over that people are willing to put up with a lot when you have the cheapest prices.
It's different from an industry that's built on reputation and trust. Like, a password manager, the only real thing you're selling is your reputation. Losing trust is a real existential threat. Security costs need to be in the bucket of either "yes, we will do it" or "it's so expensive that if we do it then we don't have a business anyway, so we'll skip it and pray."
I think reputation impacts in this industry from anything other than crashes don't hold much staying power.
I'm actually quite tempted to buy the SWA dip...
Southwest isn't a particularly budget airline compared to modern budget carriers like spirit and ryanair that haven't copied the open boarding policy. I suspect the opportunity to upsell seats / luggage and have distinct classes outweighs the turnaround time costs of assigned seating.
They're also the "friendly airline", they easily have the most personable and friendly staff. I don't know what they do different, but Southwest employees treat me human and all the rest generally treat me like human trash. It's got to be a company culture thing, maybe connected to Southwest not having a first-class section.
Usually I fly with Southwest whenever possible without thinking twice about it, but this outage and the outage last year are forcing me to reconsider. Better to deal with rude people than to have my flight delayed..
Definitely. Normal American customer service is to treat you like human trash, so obviously Southwest has decided to do this differently and so probably does things like only hiring friendly people, training them to be friendly and positive even in difficult situations, and checking on them somehow to make sure they're doing this and not just faking it for the interview and training. Other airlines obviously don't, but it's not just airlines, it's everywhere in American customer-facing business these days. Rudeness is just a normal part of America culture now.
I have a backpack that has a phone holder on the shoulder strap. When you click your phone in it, the camera is visible. I've had many a situation where someone was about to go full New York on me, but noticed the camera staring at them and toned it down. I've never once been recording.
training them to be friendly and positive even in difficult situations
That's what happened with Dr. Dao in any case. Not long after that shit show, I booked a transcon on American. Turns out it was overbooked, and they were desperate to get people off the plane. The gate agents basically asked how much money it would take to get people to volunteer to take another flight. They got their volunteers, everyone went home happy because American empowered their staff to resolve problems.
United is awful. American isn't much better.
With the exception of United I'd say most airlines I've dealt with have been pleasant whether or not I'm flying up front or have status. Yes, even RyanAir. We're in something of a golden age since (with the notable exception of Southwest) American airlines have mostly put their mergers behind them and have mostly shed themselves of CEOs who view employees as adversaries. Take a look at the bad old days of the late 80s through mid 2000s. Smisek. Lorenzo. Parker. Bastards and crooks, all of them. It's difficult to stress just how toxic airline leadership was and how that trickled down for a long time.
A fortune, they only just finished an 8-year migration to Amadeus
Thorough unit and integration tests include failure modes. Mine include things like "what happens if the OS reports that the storage was unmounted during read/write" (because that was a failure often seen in production with some flaky SAN devices) and "what happens if the server stops responding" (because networks are generally unreliable) and "what happens if invalid (random) data is given" because data corruption is a frequent occurrence for similar reasons.
> Perhaps you have a false sense of security about what you’ve really been testing?
Perhaps. On the other hand I've seen a _lot_ of other developers test only the happy path and call it a day then spend days/weeks/months debugging failures.
As a IT-VP/CIO, the statement of "there's no way to test it" is not acceptable.