3 ms·

Here's what's funny: Rands begins with a story about Borland, and how it was this amazing software company run by hacker barbarians who by 1992 were converting

by mapgrep 15y ago

Here's what's funny: Rands begins with a story about Borland, and how it was this amazing software company run by hacker barbarians who by 1992 were converting their products with "complete object-oriented rewrites." What he doesn't say is that Borland's hacker ethos is what greatly diminished the company. The company indulged a well-known programmer tendency to bulldoze and rewrite code. The company tried to rewrite Arago into dBase for Windows and in the meantime was surpassed by Microsoft Access. It tried to rewrite Quattro Pro for Windows, ended up with few new features to show for years of work, and was surpassed by Microsoft Excel. Joel Spolsky wrote about both cases in one of his best known essays http://www.joelonsoftware.com/articles/fog0000000069.html http://www.joelonsoftware.com/articles/fog0000000069.html Rands is making the point that companies need hacking, or, as he puts it, "well-maintained Barbaric chaos inside the company." In this he is right. Facebook is trouncing Google with a more hackerish ethos, and is not making the same "rewrite everything" mistake as Borland. But Facebook /is/ making some fresh mistakes. If you were to ask yourself, "Is Facebook maybe taking an excessively hackerish view on anything like Borland did?" it would not be hard to find affirmative examples, including a series of features that use private data in a way that tends to offend non hackers, and a series of privacy defaults that do likewise. By more fully acknowledging the ways in which the hacker ethos burned Borland (beyond the understatement that Borland's products were "running late"), and by acknowledging the problems hacker culture contributed to at Facebook, Rands would have an even better essay, one that both asserted the importance of the hacker way and that acknowledged the potential downsides of hacker thinking. I'd be interested to hear his thoughts on how to embrace hacking while also instituting some checks on the less useful impulses of programmers.

krig 15y ago
Aren't you simply reinforcing his point, though? There's a deep and fundamentally unsolvable conflict between the need to nurture the existing good versus throwing it away for what may turn out to be a failure. That companies fail attempting risky projects is not an indictment of risk-taking: Taking risk is the only way to succeed in the long term. At the same time, risk is, well, risky. Also, you cannot look at a single company in isolation. The barbarians at Borland may have been bearded and fearsome, but that didn't mean that there weren't any other barbarians around, hacking away at other companies at the same time. There's just no fool-proof path to success. There's plenty of fool-proof ways of failing, though, and to stop taking risks and attempting to innovate has to be one. I don't think it's really all that useful to try to learn lessons from individual mistakes made by companies in the past. They made their mistakes because they attempted things they hadn't tried in the past. With hindsight, they probably shouldn't have. Hindsight is the one thing I can guarantee you'll never have access to before you set out on a project.
- mapgrep 15y ago
  Hindsight won't fix old mistakes, but looking at the past helps avoid new ones. If there are lessons in the success of Borland and Facebook, surely there are lessons in their failures, too. Or maybe we disagree on that.
  - krig 15y ago
    Well, I don't agree with you entirely at least. It is not at all evident to me that looking at the past helps avoid new mistakes. It might help avoiding the exact mistakes that were made before, but it is highly probable that the mistakes that, again, in hindsight, should have been avoided are not the same mistakes as those that others made before. In fact, I think there are plenty of examples throughout history of how it is the desire to avoid certain previous mistakes that directly lead to new mistakes being made. I do agree that the lessons to be learned can be found equally among previous failures as in successes. In the same way that I think it is problematic to focus on individual previous mistakes, I also think it is problematic to focus overly much on individual previous successes. Simply copying the strategies that worked for someone else won't lead to similar success, as hundreds of failed attempts at replicating the Apple iPad can attest to.
RyanMcGreal 15y ago
> a series of features that use private data in a way that tends to offend non hackers, and a series of privacy defaults that do likewise. Really? I find it's the opposite: hackers tend to be both conscious of and bothered by Facebook's flippancy toward privacy, whereas non-hackers don't seem to know or care either way.
narag 15y ago
Borland's hacker ethos is what greatly diminished the company. The company indulged a well-known programmer tendency to bulldoze and rewrite code. The company, as in management, took wrong decissions. But the blame still manages to fall on poor programmers :-)
- mapgrep 15y ago
  I actually blamed the "hacker ethos," not the rank and file programmers, but the CEO of Borland was very much a coder: http://en.wikipedia.org/wiki/Philippe_Kahn http://en.wikipedia.org/wiki/Philippe_Kahn
  - narag 15y ago
    I don't think so. The Wikipedia article you link to doesn't support that claim anyway. I'd say he was a hacker, in the broad sense of the word, but not a programmer. That would be Anders. A comparison with Jobs and Wozniak comes to mind.
    - mapgrep 15y ago
      > The Wikipedia article you link to doesn't support that claim anyway "As a student, Kahn developed software for the MICRAL, the earliest non-kit personal computer based on a microprocessor."

by mapgrep 15y ago

krig 15y ago

Aren't you simply reinforcing his point, though? There's a deep and fundamentally unsolvable conflict between the need to nurture the existing good versus throwing it away for what may turn out to be a failure. That companies fail attempting risky projects is not an indictment of risk-taking: Taking risk is the only way to succeed in the long term. At the same time, risk is, well, risky. Also, you cannot look at a single company in isolation. The barbarians at Borland may have been bearded and fearsome, but that didn't mean that there weren't any other barbarians around, hacking away at other companies at the same time. There's just no fool-proof path to success. There's plenty of fool-proof ways of failing, though, and to stop taking risks and attempting to innovate has to be one. I don't think it's really all that useful to try to learn lessons from individual mistakes made by companies in the past. They made their mistakes because they attempted things they hadn't tried in the past. With hindsight, they probably shouldn't have. Hindsight is the one thing I can guarantee you'll never have access to before you set out on a project.

mapgrep 15y ago
Hindsight won't fix old mistakes, but looking at the past helps avoid new ones. If there are lessons in the success of Borland and Facebook, surely there are lessons in their failures, too. Or maybe we disagree on that.
- krig 15y ago
  Well, I don't agree with you entirely at least. It is not at all evident to me that looking at the past helps avoid new mistakes. It might help avoiding the exact mistakes that were made before, but it is highly probable that the mistakes that, again, in hindsight, should have been avoided are not the same mistakes as those that others made before. In fact, I think there are plenty of examples throughout history of how it is the desire to avoid certain previous mistakes that directly lead to new mistakes being made. I do agree that the lessons to be learned can be found equally among previous failures as in successes. In the same way that I think it is problematic to focus on individual previous mistakes, I also think it is problematic to focus overly much on individual previous successes. Simply copying the strategies that worked for someone else won't lead to similar success, as hundreds of failed attempts at replicating the Apple iPad can attest to.

RyanMcGreal 15y ago

> a series of features that use private data in a way that tends to offend non hackers, and a series of privacy defaults that do likewise. Really? I find it's the opposite: hackers tend to be both conscious of and bothered by Facebook's flippancy toward privacy, whereas non-hackers don't seem to know or care either way.

narag 15y ago

Borland's hacker ethos is what greatly diminished the company. The company indulged a well-known programmer tendency to bulldoze and rewrite code. The company, as in management, took wrong decissions. But the blame still manages to fall on poor programmers :-)

mapgrep 15y ago
I actually blamed the "hacker ethos," not the rank and file programmers, but the CEO of Borland was very much a coder: http://en.wikipedia.org/wiki/Philippe_Kahn http://en.wikipedia.org/wiki/Philippe_Kahn
- narag 15y ago
  I don't think so. The Wikipedia article you link to doesn't support that claim anyway. I'd say he was a hacker, in the broad sense of the word, but not a programmer. That would be Anders. A comparison with Jobs and Wozniak comes to mind.
  - mapgrep 15y ago
    > The Wikipedia article you link to doesn't support that claim anyway "As a student, Kahn developed software for the MICRAL, the earliest non-kit personal computer based on a microprocessor."