I have two friends with the public beta of OSX, and neither of them have X servers nor X libraries or anything of the sort. There's a big difference between an OS which runs on a BSD kernel and provides some standard Unix-style services, and an open-source OS that provides a large array of open-source apps, desktops, GUIs, etc.
OSX will be OSX, it isn't a Linux killer. I welcome it. If it provides a good, usable, fast, robust desktop to the masses, it's about time.
My snake oil warning meter is fairly high on this one. Typically companies which claim "the most secure communication system there is" are full of hot air. Their site doesn't give any description of their cryptosystem that I can find, and there is this disturbing quote:
The bottom line is that there is no straightforward and concise answer to your question. We at AFTI have analyzed a number of encryption systems, and we believe SafeMessage to be more secure than any of the competition. But we can't provide a simple bit-count, for example, because our system encrypts the same data with several different ciphers and keys, some symmetric, some asymmetric from large fields, complicating the math of arriving at said bitcount.
This doesn't sound promising. The previous paragraphs leading up to this quote discuss the various bit strengths of well-known algorithms, so they appear to be trying to set themselves apart from well-known good crypto. Which usually means bad crypto.
The annoying thing is that the press pick these press releases up and write an article without any serious investigation of the claims made by the company.
> There's a bunch out there that are already much better established and > much better thought-through
Such as? There are no deployed systems of this type with the exception of ZKS's Freedom system, and it requires that 1) you pay and 2) you use windoze. Further they've never released protocol details or source code, so the system is untrustworthy (regardless of the fact that a number of 'big-name' cypherpunks are part of the company).
The best implementation of this kind of system is an onion-routing based approach, which Fling is. I will agree that in its current state some of the protocol design is naive and subject to large-scale traffic analysis, but the approach is correct.
What systems were you referring to that are better?
My read on this brief is that there simply isn't any copyright violation going on. According to the brief, the AHRA provides for noncommercial copying as a perfectly legitimate and non-infringing activity. No?
Many thanks to David Boies for taking this case... he seems to have his act together.
It's not a matter of frightened, it's a matter of just not liking it. I've written Python, Perl, Tcl, Pike, Lisp, sh, C, C++, etc, etc. Whitespace shouldn't be syntax, IMO. If you disagree, no problem, use Python all you want. I won't. I don't believe the look of source code should be enforced in the language; it's like legislating morality, it just doesn't work. Good programmers will write good code and most of the time it will be readable and reasonably indented and formatted. Bad programmers will write bad code and no language will make that go away.
The point of this thread was not that Python is lame or great or anything else, it was that Pike is at least worth a look. In passing I mentioned that I much preferred its syntax to Python's. Cope.
How is this on topic? The point of my post had nothing to do with Python. The fact that I dislike Python due to its use of whitespace as syntax is completely irrelevant to Pike. It was an illustration. Don't get so bent out of shape.
Now with regards to Python and whitespace, the indentation issue is one of the _first_ things that keeps me from writing code in Python. Therefore I mention it. Is it the only thing I don't like? No. It is the most important? Probably not. It just happens to be the first one I think of. My post wasn't a litany of gripes about Python, it was an exhortation to people to actually explore new languages before tossing them aside. The point stands, regardless of Python gripes.
How will you know until you investigate? It used to be that only M$ was "bankable" in terms of OSes. I'd hate to think what would have happened if noone just went and played with new stuff. Seems to me that an OO C-like scripting language with Gtk and lots of other bindings and modules is pretty interesting at least on the surface.
Python offers alot of OO features but has a limitation I can't deal with: it considers indentation as syntax. Pike might make a better language for OO scripting than Python given its syntax.
The point is that it's worth looking at and that innovation and new ideas, even if they turn out to be lame, shouldn't be swept under the carpet because "we've already got one of those." I would much rather see a discussion of the actual features of the language than questions like "what's the point?"
(BTW I have no interest in Pike other than it looks pretty interesting and appears to already have alot of useful features)
Pike has Gtk bindings too and my (limited) experience with them has been good. Seems like a pretty cool language. There's a good Pike resource site at http://www.pike-community.org/.
Yes, Metallica have every right under the current copyright laws to do what they're doing. However, that really isn't the point. There are two points that are highlighted by all of this:
Just because Metallica have the right to do what they are doing, that doesn't imply that they should. Legal and illegal are not the same as right and wrong. Metallica has long been a "band of rebels" and grew out of a culture which they have now, judging by their actions, turned completely against. They're welcome to file all the lawsuits they want to, but the music-listening community should drop them as far and as fast as possible. They clearly don't deserve their fans.
The copyright laws need changing. They currently infringe so far on the freedom of people in this country as to make interacting in natural ways with friends completely illegal. Granted this is a huge can of worms and the "Right (tm)" solution isn't necessarily obvious, but the laws must change.
We need to send Metallica and other 'corporate' bands a huge message about their arrogance. Under the current laws, they have the legal right to sue their fans, but under no circumstances are they justified in doing so. Frankly they're disgusting.
I've enjoyed listening to their music in the past, but I won't listen to, buy, recommend, or otherwise promote them ever again. Further, I suggest that we all destroy our Metallica CDs and mail the remains to them, including a polite but to the point note as to why. Take pictures of your pile of Metallica trash and put them on your website.
I have been living in Houston and working for a company in Santa Clara for over a year now, and it has worked very well for me. I write code though, so that makes it somewhat easier. If your job involved more hands-on people stuff it would be somewhat more difficult.
Also it helps alot to have clueful people on the other end. My company is very spread around, including offices in 5 or 6 different cities, plus a number of full-time telecommuters. So they're used to this sort of thing. You're less likely to find support for full-time telecommuting if your company or your boss don't really understand how it can work. Plus my direct boss is very technical and understands ssh, firewalls, etc., so he knows that it can work and work securely.
If you're a coder, developing software for Linux and Unixes and platforms whose development tools are available on Unixes makes it much easier to telecommute, than say developing windoze software, since the tools can all be operated remotely. While much of the time I use my development environment on my own desktop, often I will build something on a different platform that resides in the company office in California. If you have to push buttons on a GUI to build, it's much more difficult to work remotely. For example, we build for VxWorks (embedded OS) from a Solaris machine, which works great remotely. However, if the VxWorks target machine ever gets hosed running some code (and it does, it's all one big happy memory block), it stops functioning and just beeps incessantly. The only repair is to hard reset it, but I can't do that remotely.
Aside from some of the logistical issues, I miss some of the office atmosphere, interacting with colleagues, etc, but by and large I'm happier without it.
In general I highly recommend it, especially if your company can cope and the kind of work you do lends itself to it.
The premise of this article is essentially broken... the notion of "a wireless Internet" is paradoxical. The Internet has always been the Internet regardless of what physical communications technologies are used to connect it together. Whether modem, ISDN, DSL, leased line, etc., the net doesn't care. The same is true of wireless.
There will never be "a wireless Internet". The Internet will always be a rich amalgam of technologies, wireless being one. Wireless has advantages and disadvantages and it will be applied where it has substantial useful advantages, but that isn't everywhere or even a large percentage of everywhere.
In order for the Internet to function and to provide the kinds of things we're used to it providing, there will necessarily be diversity in access technology and there will be methods of connecting one to another, as there are now. So whatever portion of the Internet is wireless, now and in the future, noone will know or care. All that matters is that packets get through.
What portion of the spectrum is open to the public really doesn't matter IMO. It is important to have the spectrum open to public use, but not because we all need wireless Internet. The two ideas aren't related. (In other words, having the some portion of the spectrum open doesn't yield wireless Internet).
That said, having wireless _access_ to the net is really good and should be widely available for cheap. Ricochet provides an excellent example of this (see www.ricochet.net). But this is unrelated to either the notion of "a wireless Internet" or to public access to the EM spectrum.
I don't much buy the whole NSA thing. Bruce Schneier has made some great comments on sci.crypt regarding this, check them out. In any case, this article:
seems to shed some good light on the subject. This find may be a good thing, allowing people to insert domestic crypto CSPs in export copies of windoze. In any case, as bad as M$ is, I'd check this one out thoroughly before passing judgement.
Just a note... Eric Blossom's (he co-founded Starium) original 3DES voice encryption box displayed some sort of identifying information which, if checked, would prevent a man-in-the-middle attack. It may have been a hash of the key used or somesuch, I don't recall. Hopefully this new device will have some similar feature.
Slashdot Mirror
Proof positive that REAL emoticons don't have noses. :)
OSX will be OSX, it isn't a Linux killer. I welcome it. If it provides a good, usable, fast, robust desktop to the masses, it's about time.
The Snake Oil FAQ (http://www.interhack.ne t/people/cmcurtin/snake-oil-faq.html) has alot to say about this sort of thing.
The annoying thing is that the press pick these press releases up and write an article without any serious investigation of the claims made by the company.
> much better thought-through
Such as? There are no deployed systems of this type with the exception of ZKS's Freedom system, and it requires that 1) you pay and 2) you use windoze. Further they've never released protocol details or source code, so the system is untrustworthy (regardless of the fact that a number of 'big-name' cypherpunks are part of the company).
The best implementation of this kind of system is an onion-routing based approach, which Fling is. I will agree that in its current state some of the protocol design is naive and subject to large-scale traffic analysis, but the approach is correct.
What systems were you referring to that are better?
Many thanks to David Boies for taking this case... he seems to have his act together.
The point of this thread was not that Python is lame or great or anything else, it was that Pike is at least worth a look. In passing I mentioned that I much preferred its syntax to Python's. Cope.
Now with regards to Python and whitespace, the indentation issue is one of the _first_ things that keeps me from writing code in Python. Therefore I mention it. Is it the only thing I don't like? No. It is the most important? Probably not. It just happens to be the first one I think of. My post wasn't a litany of gripes about Python, it was an exhortation to people to actually explore new languages before tossing them aside. The point stands, regardless of Python gripes.
Get a grip.
Python offers alot of OO features but has a limitation I can't deal with: it considers indentation as syntax. Pike might make a better language for OO scripting than Python given its syntax.
The point is that it's worth looking at and that innovation and new ideas, even if they turn out to be lame, shouldn't be swept under the carpet because "we've already got one of those." I would much rather see a discussion of the actual features of the language than questions like "what's the point?"
(BTW I have no interest in Pike other than it looks pretty interesting and appears to already have alot of useful features)
Pike has Gtk bindings too and my (limited) experience with them has been good. Seems like a pretty cool language. There's a good Pike resource site at http://www.pike-community.org/.
Just because Metallica have the right to do what they are doing, that doesn't imply that they should. Legal and illegal are not the same as right and wrong. Metallica has long been a "band of rebels" and grew out of a culture which they have now, judging by their actions, turned completely against. They're welcome to file all the lawsuits they want to, but the music-listening community should drop them as far and as fast as possible. They clearly don't deserve their fans.
The copyright laws need changing. They currently infringe so far on the freedom of people in this country as to make interacting in natural ways with friends completely illegal. Granted this is a huge can of worms and the "Right (tm)" solution isn't necessarily obvious, but the laws must change.
We need to send Metallica and other 'corporate' bands a huge message about their arrogance. Under the current laws, they have the legal right to sue their fans, but under no circumstances are they justified in doing so. Frankly they're disgusting.
I've enjoyed listening to their music in the past, but I won't listen to, buy, recommend, or otherwise promote them ever again. Further, I suggest that we all destroy our Metallica CDs and mail the remains to them, including a polite but to the point note as to why. Take pictures of your pile of Metallica trash and put them on your website.
Regards, moibus.
Also it helps alot to have clueful people on the other end. My company is very spread around, including offices in 5 or 6 different cities, plus a number of full-time telecommuters. So they're used to this sort of thing. You're less likely to find support for full-time telecommuting if your company or your boss don't really understand how it can work. Plus my direct boss is very technical and understands ssh, firewalls, etc., so he knows that it can work and work securely.
If you're a coder, developing software for Linux and Unixes and platforms whose development tools are available on Unixes makes it much easier to telecommute, than say developing windoze software, since the tools can all be operated remotely. While much of the time I use my development environment on my own desktop, often I will build something on a different platform that resides in the company office in California. If you have to push buttons on a GUI to build, it's much more difficult to work remotely. For example, we build for VxWorks (embedded OS) from a Solaris machine, which works great remotely. However, if the VxWorks target machine ever gets hosed running some code (and it does, it's all one big happy memory block), it stops functioning and just beeps incessantly. The only repair is to hard reset it, but I can't do that remotely.
Aside from some of the logistical issues, I miss some of the office atmosphere, interacting with colleagues, etc, but by and large I'm happier without it.
In general I highly recommend it, especially if your company can cope and the kind of work you do lends itself to it.
Regards.
There will never be "a wireless Internet". The Internet will always be a rich amalgam of technologies, wireless being one. Wireless has advantages and disadvantages and it will be applied where it has substantial useful advantages, but that isn't everywhere or even a large percentage of everywhere.
In order for the Internet to function and to provide the kinds of things we're used to it providing, there will necessarily be diversity in access technology and there will be methods of connecting one to another, as there are now. So whatever portion of the Internet is wireless, now and in the future, noone will know or care. All that matters is that packets get through.
What portion of the spectrum is open to the public really doesn't matter IMO. It is important to have the spectrum open to public use, but not because we all need wireless Internet. The two ideas aren't related. (In other words, having the some portion of the spectrum open doesn't yield wireless Internet).
That said, having wireless _access_ to the net is really good and should be widely available for cheap. Ricochet provides an excellent example of this (see www.ricochet.net). But this is unrelated to either the notion of "a wireless Internet" or to public access to the EM spectrum.
http://ntbugtraq.ntad vice.com/default.asp?sid=1&pid=47&aid=52
seems to shed some good light on the subject. This find may be a good thing, allowing people to insert domestic crypto CSPs in export copies of windoze. In any case, as bad as M$ is, I'd check this one out thoroughly before passing judgement.
Just a note... Eric Blossom's (he co-founded Starium) original 3DES voice encryption box displayed some sort of identifying information which, if checked, would prevent a man-in-the-middle attack. It may have been a hash of the key used or somesuch, I don't recall. Hopefully this new device will have some similar feature.