Open Code = Closed Code?
It was certainly one of the most interesting parts of this most recent "exchange" to be told, by a coder I respect no less, that binaries "reveal all without revealing the source code." The implication drawn from this was that having access to the source was not important. Indeed, it was a "distraction." That was a bit of a shock to the way I look at the world. I'm of course not a coder (I did that for many years in my youth, and indeed, for two years earned my keep coding , but my mind has been corrupted by east coast code so I wouldn't pretend to know anything interesting about coding anymore), but I wonder, has anyone told RMS this? I mean, why are we worrying so much about open or free code, if in fact binaries "reveal all"?
The folks at Oops helpfully note that at least "there's more material for study in a book than in a [binary only] softrware program." And that in turn may be responsible for the "quite poor" state of software development. That seems right. But is this interesting view of this coder (who I want to be quick to add has done a great deal of good advancing all things good and great) shared by others? I mean, I understand the sense of binaries "reveal all" in the sense that all functionality might be inferrable from a binary (not sure if that's true, but I'll assume it). But is the open/closed line really so unimportant?
| Permalink | technorati
Comments (7)
This perhaps maybe a moot point. I am no super coder myself. I think the point of your friend was to point out that any binary code can be reverse-engineered into some rough semblance of its original self. However, this rough semblance isn't for the faint of heart to modify. Herein lies the distinction. Open code, is begging to be modified, improved, made for human coder consumption, because it was written by a human. Binary code, reverse engineered by a machine to some modifiable code, will take a buttload of time to understand, and for most intents and purposes, isn't as open=modifiable as true 'open' source. Reverse engineering in this way, isn't for the faint of heart so I don't see this avenue for 'opening' closed code to stir the embers of some new anti-M$ revolution.
Peace, love and linux(shameless plug)
When you're writing software for the mass market, the effort goes for the most part into designing the interface, and most of the value is in the interface. Now, hide the code if you will, you can never prevent the features of the interface from being seen, understood, and reused.
I think Sebastien makes a good point that designing the user interface is one of the keys to mass market software success.
Given the difficulty of using copyright to protect user interfaces, I have always expected that there would be more patents relating to user interfaces. I would expect both utility patents and design patents, since it is permissible using design patents to patent software icons.
If 'Open Code = Closed Code' than cou can't comply with Non-Disclosure-Agreements which prohibit you from disclosing the knowledge gained in wirting the programm.
E.g. when you want do develop a deriver for special hardware the hardware manufacturer will give you the technical specifications only if you sign a NDA keeping you from giving away the sourcecode of the driver you are developing. This hurts many Freie Software Projects. But if 'Open Code = Closed Code' you even can't give away the binary driver.
can open code resolve issues raised by the napster controversy?
I applaud these excellent principles. Free Culture should remain that and not be subsumed into agencies, authorities or anyone else. See Cyberlaw for more ethics discussion
continuous heavy traffic or data sessions" almost assuredly means "running a popular HTTP or KaZaa server to the point that we notice it and want to shut you down".