Follow Slashdot stories on Twitter


Forgot your password?
For the out-of-band Slashdot experience (mostly headlines), follow us on Twitter, or Facebook. ×

Comment: Re:Fucking Lawyers (Score 3, Interesting) 174 174

That would be sufficient to make the APIs stop working. Perhaps you should think again about what information is required and copyrightable.

I think that this decision may mean that Google will need to do something like alphabetize the API. (Customized organization can be copyrighted, but alphabetical order can't.)

Comment: Re:Open source isn't enough (Score 1) 246 246

The last time I looked into using Objective C on a Linux system the documentation of both Objective C and GnuStep were lamentable. So bad I picked up something else.

I'll agree it's a difficult problem, but Python, Ruby, D, and even Smalltalk (well Squeak, anyway) and Scheme (Racket anyway) have addressed it reasonably. Objective C documentation seems to only work if you're on an Apple, and GnuStep documentation doesn't seem to work anywhere. It's great if you just want something to remind you of how to do something that you already basicly know how to do, otherwise not.

Comment: Re:Linux Support (Score 2) 246 246

I'm sorry, but I don't get the reasoning behind your evaluation. I'm not greatly familiar with Objective C, but the only reason that I didn't pick it up in the past was that other languages had better documentation. (I'm not using an Apple).

Even GnuStep wasn't that bad. It was better at handling unicode than C libraries were. But too many pieces assumed that you had someone at hand to explain basics to you.

So I used Python and D and Java. I looked at Vala, but it seems too tightly bound up with gtk, and when the maniacs began pushing gnome3 I gave up on it. I avoid C because of excessive use of pointers. That's also a problem with C++, but it's really a matter of excessive ambiguity (and poor handling of unicode) that cause me to avoid it. (The unicode problem also affects C and Java, and possibly objective-C, though I never really checked. Java doesn't even HAVE a unicode aware ispunct() function. I needed to fake one up using the general character classification, which Java made numeric for some stupid reason. The unicode version would have made things a lot easier as I could just have checked the first character of the classification.)

So, objective-C has problems with documentation (if you aren't on a Mac), and I'm not thrilled with the GnuStep choice of 16-bit unicode (unless they've changed that since I looked), but what's wrong with the language?

Comment: Re:And counting... (Score 2) 236 236

The question is, how could they possibly restore trust?

They had trust, the secretly betrayed it, using techniques that were not evident. So if they reform, how do you know that they've actually reformed rather than just changed their techniques?

And for that matter, there is plain evidence of shipments being intercepted and altered without the manufacturers knowledge. So you also need to verifiably reform the methods of shipping. How do you verify their security? The only thing I can think of is something analogous to key signing for hardware, but I can see no way to implement that.

So you say "they need to resolve this issue", and I agree that the need is present, but I don't see a possible mechanism.

Comment: Re:The NSA fallout here is astonishing (Score 2) 236 236

No. Read your history. Actually the US has been remarkably peaceful for a dominant world power. Compare it to Imperial Rome or Alexandrian Greece or the Persian Empire. I suspect that this is because wars are no longer profitable. Trade driven empires were rare in ancient times, Egypt is the only one I can think of. (Sorry, I don't know enough Chinese or Indian history to include them in this summary.) But this US "empire" is more similar the the Egyptian empire than even the the British Empire. And the British Empire was peaceful compared to it's predecessors (though wars were still slightly profitable up until around WWI).

My hope is that the sucessor to the US will be even more trade driven that is the US "empire". But that *is* only a hope, and would be quite unusual historically. OTOH, wars are now much more costly and less rewarding that they have ever before been.

Comment: Re:Alternative? (Score 1) 236 236

Depends. If you are a company doing business with your government, you buy locally. If you want to be secure from your government, you buy as foreign as possible, i.e. from a company in an area controlled by a government that has as little interaction locally (to you) as possible. And you still can't trust it, because the shipment could have been intercepted during importation.

Comment: Re:Disagree with stupid wording (Score 1) 236 236

I'm not sure you're right, though it's certainly possible. It's also possible that the current owner of the site has some default opinions that are automatically emplaced. Certainly, though, there are plenty of astroturfers on various topics. There's little evidence that it's even often the government. This, of course, doesn't mean that it isn't. It merely means that it's an unnecessary hypothesis.

Comment: Re:Controlling the message (Score 1) 172 172

This is an inherent problem with systems that have a centralized control. Including GitHub. git is inherently decentralized, so it has different problems, which the designers of GitHub tried to resolve through centralization. If you'll notice, it worked. The problems with decentralization were solved. But now we have the problems of centralization.

I don't know what a real solution would be. Google is a good example in another area. Web pages are decentralized, but Google makes it possible to find what you need...but Google is centralized, so if Google doesn't want you to find somthing, it is even more difficult to find than before.

This is the inherent problems of monopolies, even when they aren't abusing their power. But monopolies always eventually abuse their power. Sometimes not until the first generation of management retires, but eventually.

I was worried about SourceForge from the first time that I heard about it. But it was so useful...

Comment: Re:Honest, trust us... (Score 1) 143 143

Well, trust isn't a single bit operation. More nearly a float. Actually, trust along a single dimension is reasonably considered a float, but there are multiple dimensions.

Yes, it's safer if you use your own trusted compiler. But it's also safer if you build your own CPU, and the rest of your computer. And I doubt that MS would have bothered to build a custom compiler that would hide back doors when it was compiling the MSWind OS. It clearly *could*, it's just unlikely. Of course, how unlikely you consider it depends on what you are worried about, and I'm not planning on using any of their software, so I can afford to be unworried. I worry more about SOHO router vulnerabilities.

So the question becomes "For what purposes are they considering using MSWind?". This is still probably only security theater, but your proposed objection is likely to be unreasonable. One should never be certain, so one operates on the balance of probabilites of cost and gain.

Comment: Re:They have no concept (Score 1) 145 145

You're missing the point. The other guy *is* evil/hateful/fascist/$badBadBad, just in a slightly different way than the guy you were convinced to vote for. This is usually true even for the candidates offered by the minority parties, though that may well be because only loons will run after an office (and spend the effort) when there's just about no chance they'll get it.

Every election I witness I become more and more convinced that a lottery would be a much better way to select a representative. Three adults at random from each voting district. And penalties for declining (say, triple your tax bill for the next 20 years). And when you "retire" from office, you get a pension of twice the median income in your district, and are prohibited from accepting "favors" (how to phrase that to eliminate loopholes) from anyone you regulated or passed laws concerning while you were in office.
This would require a bit of internal restructuring of the government to remove the ability of a single person to really screw things up (as occasionally a real winner would get selected) but that needs doing anyway (as occasionally a real winner gets elected).

Comment: Re:What about the cost for enrichment waste? (Score 1) 169 169

I understand that you think that Synroc is a solution. Maybe it "sort of" is. I've long considered glassification to be a reasonable approach to develop. But I said "develop". I don't believe that there has been any extensive testing of Synroc, so I don't consider it a proven solution. Perhaps it would work out. (And could you use it for process heat?)

The thing is, this is deciding that we're going to throw away most of the recoverable energy. I think hot breeders are a more desirable approach, with something like Synroc used only on the final result. But it's also quite possible that really counting all the costs would leave the decision as "This costs more than the alternatives in almost all situations."

Clearly the current approach is bad policy, poorly and unsafely implemented. It's not clear to me what good policy would be, particularly when one can pretty much guarantee that over time some people are going to be lazy, stuipd, and selfish. This seems to mean that you need to ensure that it's difficult to profit from unethical behavior. Preferably both difficult and dangerous, but at least difficult. So fast breeders are problematic. It's possible to stop them and extract weapons grade materials. One of the arguments for pebble bed reactors is that it makes this more difficult. But pebble bed reactors have lots of high level waste. What's really needed is a way to make that useful, not just a better way to throw it away.

A fail-safe circuit will destroy others. -- Klipstein