Hacker News Reader

Comment by corbinpage | Hacker News Reader

I mean, all software has bugs.

To me, the major difference between Apple's OS's and the Windows/Linux/Androids of the world is that Apple _fails smoother_.

A typical bug may result in an iOS app freezing and then quitting on the user. However frustrating this may be, Apple's "something went wrong" user experience is much better than the annoying pop-up alerts, blue-screens of death, and cryptic error messages of the other platforms.

These error messages are only relevant to a small, small subset of power-users, so why show them to everyone else? The other 99% of users are just going to quit and restart the application anyway (as we've been trained to do), so why not smoothly lead them down this path?

View on HN

"Apple _fails smoother_."

Umm.. not any more apparently

http://daringfireball.net/linked/2014/10/01/it-just-works

"My iPhone 6 rebooted after I changed the home screen wallpaper. Tapped a new image in the wallpaper settings, and poof, it rebooted. Worse, it never stopped rebooting. Endless reboot cycle."

The other 99% of users are just going to quit and restart the application anyway (as we've been trained to do), so why not smoothly lead them down this path?

The fact that 99% of people want to be completely ignorant about the systems that their lives are centered around, and partially controlled by, does not mean they should be restricted from any opportunity to do otherwise. To answer your question,

so why show them to everyone else?

It gives the opportunity for someone to be curious and search for a solution, or at least mention the error to someone who can help. Even if they don't remember the message completely, telling someone "it couldn't foo the bar" even if they don't know what that means is far, far better than "something went wrong". We are told to be as detailed as possible when asking questions or writing bug reports, but these opaque error messages encourage the exact opposite.

This is from someone who has been asked to troubleshoot others' Apple laptops.

> I mean, all software has bugs.

That's a cool mantra to repeat but it do not seem to match reality.

For instance, in over 10 years of daily usage, I do not think having ever noticed any bugs in vim, grep and PostgreSQL. Just to take grep, for instance: if it says the string is not in the file, I have 100% confidence it is not in the file. So not only grep do not freeze my PC, it is also deterministic, which means perfectly reliable when one talks its language.

OSes are more complex you will tell me. Yep, but you said "all software has bugs".

Well, if we're being pedantic, the fact that you have never noticed a bug in Vim, grep, and PostgreSQL does not mean they don't exist. If you google, e.g., "postgresql bug" you will find evidence of this.

The charitable reading of GP's comment is, of course, that virtually all software has bugs, or that all complex software has bugs. Or, even better, that it is extremely difficult to write bug-free code, and this difficulty rapidly increases with the complexity of the software.

You can't acknowledge that an operating system is more complex, but then not give it the benefit of the doubt of halting a system on a bug. If grep or vim fail, they just abort. If Postgres fails, you maybe get a corrupted database and it aborts. If OS X kernel panics, the machine halts. All software has bugs.

Also, if Postgres' bugs were unnoticeable, why did 9.3 get to 9.3.5 before the went to 9.4? Just because you didn't notice a bug doesn't mean someone else wasn't affected by it, and just because you notice bugs in OS X (which I'd bet you didn't) doesn't make them big issues.

Reply on news.ycombinator.com