Email or username:

Password:

Forgot your password?
Top-level
Bartek Jasicki

@blacklight While I agree that Voyager is a marvelous example of code, I disagree that we lost something over time. πŸ˜‰

1. In those days there are also created a bloated or buggy software, but today we don't remember it. Using the comparison to masonry, because it collapsed. πŸ˜€

2. It is hard to compare two completely different types of software. If desktop software would be created in the same way as airspace one, it would be a few thousand times more expensive and slower to deliver.

11 comments
Paul Schoonhoven πŸ‰ πŸ‹

@thindil indeed... Over 50 years only the good stuff survives. It doesn't matter if that is software or a building. The bad versions are buried and forgotten.
@blacklight

Bartek Jasicki

@vosje62 Time is the best medicine... but also the best Q&A department. 😁

Natasha Nox πŸ‡ΊπŸ‡¦πŸ‡΅πŸ‡Έ

@thindil @vosje62 So basically 99% of stuff on the market today will eventually be forgotten and lost because all the dependencies ("cloud") will break, proprietary code will be forgotten and the binaries become dead weight. πŸ™ƒ

DiscreetSecurity

@vosje62 @thindil @blacklight Partly true. We would remember either of the Voyager craft even if they'd died due to a bug after 25 years. As long as the other/both haven't, at least, we will remember both, for a long time.

Fabio Manganiello

@thindil good point - thanks for pointing out the potential "survivorship bias" there.

But even when we take into account the "fallen buildings" from that era, I still feel like some of that craft has been lost.

When I look at some of that old code (I'd say up to mid 1980s), I see mastering of bitmasks, hammering data into registers and memory blocks, craft of memory allocation and usage (some of the coders from those years even wrote their own little garbage collectors before Java and friends came around, or were masters in optimizing stack usage/unwinding). In short, doing great things with the little resources that were available to them.

Even if most of those "buildings" have "fallen", and we admire the few that are still functioning at the edge of our solar system, it doesn't mean that the techniques of the time weren't authentic engineering wonders that we've lost in an age where we take so many abstractions for granted.

@thindil good point - thanks for pointing out the potential "survivorship bias" there.

But even when we take into account the "fallen buildings" from that era, I still feel like some of that craft has been lost.

When I look at some of that old code (I'd say up to mid 1980s), I see mastering of bitmasks, hammering data into registers and memory blocks, craft of memory allocation and usage (some of the coders from those years even wrote their own little garbage collectors before Java and friends came...

Bartek Jasicki

@blacklight True, when I'm looking at a code, especially which coming from HIS, almost always it is a bit magical code. πŸ˜‰ There are reasons for that:

1. Programmers have a lot of time for tinkering with the code.
2. Many programmers there are very experienced.
3. When creating that kind of software, there is an access to really powerful tools, often even unknown for others.

I think, young people of today will someday do something better, as we improved in other fields over time. Eh, 500 limit.

@blacklight True, when I'm looking at a code, especially which coming from HIS, almost always it is a bit magical code. πŸ˜‰ There are reasons for that:

1. Programmers have a lot of time for tinkering with the code.
2. Many programmers there are very experienced.
3. When creating that kind of software, there is an access to really powerful tools, often even unknown for others.

πŸ³οΈβ€βš§οΈ Fiona πŸ³οΈβ€βš§οΈ

@blacklight @thindil But now compare the performance and the average number of bugs with modern code that was written for an optimizing compiler. You may find that there are reasons for why we don't do that stuff anymore.

Bartek Jasicki

@Fiona I'm not sure do I understand correctly, but the average number of bugs didn't change over time. It is still approx. 1 bug per 100 lines of code. πŸ˜‰ The problem is the size of modern software. 20 years ago, Linux, the kernel, has around 500k lines of code. Today around 30M. Performance, that's the Wirth's Law: en.wikipedia.org/wiki/Wirth%27 πŸ™‚

We still do that things, just they didn't get their legends yet. They are too new. πŸ™‚

Go Up