Hacker News Reader

Comment by mattbrewsbytes | Hacker News Reader

I think you're glancing over the fact that in the olden times it was too hard to update things. If a piece of software had a bug is the company going to ship out a floppy disk to 1 million users to patch it?

The nature of software is that it can constantly be changed, updated, etc. I like it, its like no other industry. To be honest and completely selfish, there is no end in sight, it pays an absurd amount of money (comparing all jobs for all humans on earth) and I could probably do this type of work well into "retirement" age range.

The level of quality to deliver a working software service is much lower because we can change and update it at will. If you don't like working on that type of software, there are jobs out there like writing software for MRI machines and other things that require much higher levels of quality.

View on HN

> If a piece of software had a bug is the company going to ship out a floppy disk to 1 million users to patch it?

In the very early days, software companies had BBSes that they distributed patches through. Then it was FTP sites. They usually offered the updates on floppies (and, later, CDs) as well, but that wasn't what they preferred you do.

But there are two main differences that make me think those days were better -- such updates weren't needed as often as they are now, and you could choose whether or not to apply a particular update.

These days, software is often shipped half-baked because they rely on the ability to update frequently over the internet. And, partially as a result, software tends to demand an internet connection or at least regularly phone hone if it has one -- and it's often hard or impossible to skip an update.

> In the very early days,

Not everyone was aware of the BBS'es and then later (dialup/net accessed) FTP sites. Besides the hassle getting onto some of these FTP sites was time consuming as well, putting aside state interference, like redirecting your own dialup call to their own dialup FTP site.

> But there are two main differences that make me think those days were better

Simpler times, less variables, people hadnt found the weaknesses in RFC protocols, phishing attempts online and offline and other forms of deception and criminal intent, people were more trusting back then, less exposed to negative situations of varying magnitude. A classic example of ignorance is bliss.

One thing could be considered absolute or constant, the speed at which data can be obtained has increased, and the range of ideas people can access has increased. Will this affect humans in a negative or positive way?

>These days, software is often shipped half-baked because they rely on the ability to update frequently over the internet.

There is only so much testing you can do in the lab, so why not get early adopters to do the testing for you? It could be cheaper.

> There is only so much testing you can do in the lab, so why not get early adopters to do the testing for you? It could be cheaper.

Because that's abusing your customers. However, you're correct that there should be real-world testing by real users. That used to be a regular practice in the industry, and done without abusing customers. It was called "beta testing".

We still do things we call "beta testing", but "beta testing" doesn't mean what it used to mean.

Reply on news.ycombinator.com