Hacker News Reader

Comment by notreallymee | Hacker News Reader

Using a throwaway account.

A few years ago I did some consulting work for British Airways. It was interesting in that as far as the planes themselves, and the operational risk of the planes was concerned, the company had their shit together. Decisions made from a safety first point of view. Engineers making the calls. Everything else no object.

But ALL the software around the planes, from the ops to ALL of the b2c was outsourced to hell. Cue a few months later their entire datacenter went down for 3 days, then a massive security breach.

Their idea of agile was to release stuff once every 3 months in a 2 week window.

The thing that I found odd was that they didn't really perceive software engineering as engineering. They had time, money and patience on the aero engineering, but software to them was just another cost center.

View on HN

This was my experience as a contractor for the Navy, years ago. We’ve had radar systems that can detect planes after a reflection on another plane, but the software is dogshit.

Someone once told me, it’s because generals and admirals can see hardware, and fell like they got value for their money. Software, not so much.

Perhaps it’s something innate in humans. Remember how Beats headphones used to add weights to feel more substantial?

This reminds me of when I gave an tour of the software development company I was working for (we had about 250 employees at the time) and his reaction was "so many people, but you guys don't produce _anything_".

He was right, we didn't ship anything physical.

We just lined up electrons. :)

"We just lined up electeons" :)

Adding weight to headphones does have a practical advantage: it keeps them firmly on your head.

I had a relatively expensive - but very lightweight - set of headphones a while ago and I swapped them for a cheaper but heavier pair because they just wouldn’t stay in place.

That doesn't sound right to me. I don't think gravity is or should be the most important or even merely a significant factor keeping headsets in place. That, and less weight on the head should feel better to the head-holding muscles (not to your biased conscious self evaluating "feel" for the only brief time while you concentrate on it).

I would even say it is the opposite: The heavier the headset the tighter its grip has to be so that it stays in place during head movements. A tighter grip means more pressure around the ears, which for most people means less comfort.

I could see that while you concentrate on the feeling of headsets you may have that same bias as most people probably have that heavier equals better (quality, more solidly built), but I doubt that is true for the rest of your brain and your body.

I tried to find something, anything of substance on this concrete subject but this time my Google-foo failed. Links to research would be appreciated, if somebody else has better luck. A possible confounder is that heavier headsets - if it's not a cheat like in the discussed example - might indeed be better built, but I think we are talking about similar built quality here that only differ in weight, and objective criteria, not self-reported biased by already known to exist weightier-is-better value bias.

I only found this, which is really just a statement: "Relation between weight and comfort?" -- https://www.headphonezone.in/pages/headphone-weight

The weight of objects has interesting effects on our judgement: https://www.discovermagazine.com/mind/holding-heavy-objects-... -- but obviously that does not mean the body sees weight the same, physiologically.

Here's an explanation by a rating company about their headphone comfort scoring: https://www.rtings.com/headphones/tests/design/comfort

It seems the padding matters a lot more than the weight, but lighter headphones need less padding, so for a given price point the lighter headphones are more likely to be comfortable.

> Remember how Beats headphones used to add weights to feel more substantial?

This was not really true, the original tear down that found weights was of a counterfeit pair of Beats headphones.

https://www.gizmodo.com.au/2015/07/are-beats-headphones-real...

That article says that while the first pair was counterfeit, real Beats still include weights to feel more quality...

"As a few folks have pointed out, the version of the headphones we tore down was likely counterfeit. We have posted an in-depth comparison of these and a pair of authentic Beats headphones here. TL;DR: the authentic pair is nearly identical to the counterfeit one."

https://medium.com/@BenEinstein/how-it-s-made-series-beats-b...?

> Remember how Beats headphones used to add weights to feel more substantial?

Beats didn’t do this. Cheap counterfeit headphone manufacturers did.

I feel the need to support your point. British airways software has been utter s* in my experience. It's broken multiple times a year for days, making it impossible to buy tickets. Sometimes it's only the website, sometimes the mobile app, more often than not it's both.

My sister has been trying to get a ticket to visit me for Christmas. Couldn't do it because British Airways checkout is broken, tried in multiple browsers. Didn't work yesterday, didn't work today.

Yep - I've complained numerous times about being able to go through the upgrade with points flow on an existing booking, only for something to go wrong after entering my payment details and then just having to wring a call center to process the upgrade.

I wonder what the cost of annoyed customers who now fly Virgin and call center salaries is compared to proper engineering?

That's an interesting contrast to what I remember Richard Feynman saying about NASA after the Challenger - that the software was the area that was impeccably run compared to everything else.

Programming was a different game back then, when every kb counted. The bar was much higher. Niche PhDs chasing the cutting edge of technology. Now we have web dev.

What’s wrong with controlling spacecrafts with JavaScript?

Lack of a built-in integer type makes it clear that measurements are inaccurate, so there's a good reason for using JS in that domain :)

You know, you're right. We could even connect it to the net via satellite for real time updates from a headless browser.

As someone who is familiar with BA IT operations but not connected to B2B/C in any function I can say what a shitshow from experience. Engineering has put up a decent firewall to keep the MBAs from mucking it up. But shit rolls downhill and it’s only a matter of time.

Hopefully the Boeing fiasco is causing other engineering lead companies to refocus their priorities.

This mirrors my experience in automotive regarding the sentiment for hardware vs software.

I think that's often the case in any company that doesn't sell software as a product/service. I know automotive companies are really terrible at it. That said, I'm not sure what level of engineering they do anyway. It seems like at best they design the chassis and maybe the engine. Everything else appears to be contracted out to some other firm. Is that the case?

Not sure about other car companies but my last job was at GM where I worked on their android auto platform. GM had before outsourced their development, but found the quality to be lacking so they have started to move software development in-house.

Agreed, my company's biggest problems all come from a lack of basic software/programming skills which they've decided is not something they need to spend any effort on. In the words of my managers, "we are not a software company" and "learning to code is easy" and "eughhh I don't want to learn any of that computer science stuff". That latter one was when I mentioned something about clean code.

"learning to code is easy"

Oh boy, that's a good one. Like learning to write and learning to paint are easy. You can easily get SOME result...

By the way, I have automotive experience. They do, and have to, pay well.

The one industry where car analogies are appropriate :)

To echo what others have contributed, I see this as well in the aerospace industry. Project managers are often willing to accept software as a black box component but wouldn’t dare accept hardware without a high level of quality pedigree. I think some of it is the tangible nature of hardware but also because systems engineers on these large projects almost never come from a software background.

I often wonder if the cavalier attitude towards treating software as an engineered product is because the software industry is relatively young. I imagine a lot of the cowboy attitudes we see in software were also prevalent in, say, pressure vessel design 130 years ago before adoption of strict design standards became common practice. I’ve heard it said that software is very much in the cave drawing days by comparison.

"Their idea of agile was to release stuff once every 3 months in a 2 week window."

Can someone explain why this is actually bad?

(EDIT: I'm being serious, so please don't just down vote it without explaining. For airline industry with legacy systems, I don't necessarily think it's bad, but maybe I'm missing something.)

Agile Manifesto[1], principle #3: Deliver working software frequently, from a couple of weeks to a couple of months, with a preference to the shorter timescale.

There's nothing inherently wrong with this, if 3 months are considered "frequent" in your industry (true for mission/safety critical systems, or systems that should not be upgraded frequently for some reason).

The idea is to deliver small, non-breaking increments to production frequently, rather than infrequent, big bang releases.

1. https://agilemanifesto.org/principles.html

I took it as more of a "only did this one aspect of it" statement more so than a "the one thing they did was bad" statement.

Reply on news.ycombinator.com