I have one password that I've used for the past 15 years or so. It's 8
characters (9 if I need to mix numbers with it), and it appears completely random.
A coworker told me the exact same story when wondering why his password had expired. He was mildly pissed. I understood his frustration completely. In turn, he understood why I couldn't just make an exception for him.
At that point I thought of something, which I've never implemented but I think would be worthwhile. I think that when you change your password, the password strength checker should assign an expiration date based on entropy. If you want to use a password like "Cindy" - fine, but it will expire in 24 hours. If you use "a79xoibf", it will never expire. I'm assuming cracklib has a reasonable way to estimate password entropy.
Has this ever been implemented? I think, over time, such a system would encourage people to use good passwords. Having to remember a new password every week is a drag compared to keeping the same hard password for a year.
The problem is the caching architectures i've seen have the exit router communicating to the cache
through a gre tunnel, so the cache has no way to actually let it "pass through" as the cache isnt the
actual box doing the routing.
That doesn't absolve the proxy. They could figure out a way. Current behavior is broken, full stop. Perhaps the proxy's internal architecture makes it difficult to find a solution, but that is no excuse. If they have to modify the protocol between frontend and backend of the proxy... so be it.
On of my (many) concerns with this legislation has been that an adopted standard would be
unimplementable in an open-source OS. This seems to address that. Hmm.
No, it doesn't, unfortunately. Two reasons:
1. "open source code" does not necessarily mean source code covered by the Open Source Definition. Their trademark application was dropped, remember. Al Gore's last political campaign made much of having an "open source web site" - sounds wonderful, right? - but as far as I could tell it didn't meet the OSD. I emailed the webmaster asking if they had cleared it with the Open Source Initiative so that they could use the (then-pending) trademark. No reply, of course.
2. The OSD does not cover patents, only copyrights. You can release code under (say) the GNU GPL which infringes someone else's patent. People will have the right to download, copy, and redistribute your software, but not to actually use it! Strange but true.
This is an important point. Law-abiding citizens, and organizations such as Software in the Public Interest (umbrella corp for the Debian Project) will not distribute software that you can't freely use. This is also why OpenSSH does not include support for the IDEA encryption algorithm - there is certainly an open-source implementation of it available, but the patent license is considered too restrictive. If you want to use IDEA with OpenSSH, you have to download an "unauthorised" third-party patch or distribution.
And that's what I'm guessing would happen here. The code may be freely redistributable, but would be encumbered by enough patents with restrictive patent licenses that as a practical matter, you won't be allowed to use it without agreeing to onerous licensing restrictions such as "Thou shalt not disable any DRM provisions" and such.
Many would say - who cares? As long as we can crack the DRM with help from the source? And I reply - I care. Much as I disagree with the provisions of the law, I won't just break it willy-nilly. I see a big difference in being able to do something and being legally allowed to. If something is SSSCA-verboten, I will be restricted in my actions, no matter how easy it is to hack the literal mechanisms. And assuming enough open-source developers hold a similar view, perhaps the necessary software won't be written.
DISNEY IP COP: We know what you've been doing here little girl! You cannot look at that picture without
paying a licensing fee! You are violating our intellectual property!
Actually it's more like
DISNEY IP COP: Look, you can now buy this poster to put on your wall, for only $2/month. Great deal, eh? We used to charge $5/month.
They're not stupid. The carrot works better than the stick. If they can lower expectations to the point where Americans believe the pay-per-use model is normal and reasonable, that's what they'll do. Instead of people saying "ridiculous - I've already paid for this work, WTF would I want to pay per-listen or per-view or per-month?", I believe the media companies want us to eventually start thinking "Duh, of course you pay per-listen or per-view or per-month, how else would they make any money? And at only $.25 per-view this is a really good deal."
Remember, the vast majority of DVD player owners put up with region coding. "What, I can't buy that Japanese DVD because it won't play on my player? Oh well, guess I'll have to wait for the US release then." Pay-per-use might seem ridiculous now, but I believe people are complacent enough that it can be phased in. Depressing, yes....
This is for basic apartment and basic food, and not in Silicon Valley. Let's say monthly expenses are
$2000 (which buys a fairly modest to low standard of living).
Again - guess I'm glad I don't live there, wherever there is. Maybe Wichita, Kansas has a lower relative COL than I thought. In any case I typically spend less than $600 per month - and that buys a decent apartment, more than enough food, and miscellany.
Certainly I don't have a lot of "luxury" line items like cable tv, inflated energy bill from air conditioning, wireless, car payment, broadband, etc - but that's exactly my point: such things should be considered non-essentials. I think a lot of Americans believe they are essentials. In my opinion their standards are unreasonably high.
The contrapositive: many people would say I maintain a rather low standard of living and have unreasonably low expectations. Whatever. I don't feel poor.
How can software/software tell
the difference between a movie that I made myself or one who's encryption has been broken??
Because the one with broken encryption will have a watermark on it.
Remember Professor Ed Felton? The movie industry tried to produce a watermarking scheme that would identify copy-protected (read: use-protected) content so that conformant hardware/software would refuse to violate its terms. They thought, and apparently some people still think, that it would be possible to produce a watermark that would survive various transforms such as decoding / re-encoding in lossy formats (JPEG, MP3, DivX) and even AD/DA conversion ("camcorder [movie] rips"). Perhaps it is possible. Not my area of expertise (if indeed I can be said to have one).
However, Felton and others have shown that given a little computer science, it is not all that hard to crack watermarking schemes - thus making your pirate movie look once again like a home recording, as far as your SSSCA-compliant player is concerned. Given an arms race between watermark creators and Felton, I'd put my money on Felton.
Did you know that the movie industry tried this DVD format, that was pay-per-use (essentially), and that
it flopped horribly? People that bought it didn't like it, and everyone who knew how bad it was stayed
away. And now it is everyone's favorite movie codec, Divx!
Actually the Circuit City format was known as "DivX" and the popular codec is called "DivX:-)". The codec's name is making fun of the Circuit City product. Beyond that the two have little in common. DivX:-) is actually a family of codecs, with the latest version being based largely on MPEG 4. And that's all I know about it....
I will bring up the SSSCA, "Fair Use" and how expect him to vote "Nay" on this or any similar bill.
Usually "Fair Use!" is the right rallying cry, but in this case I doubt it will be effective. They are already promising to uphold fair use of the content you purchase. (Yes I know, I don't trust them either, but that's what they'll say -- it's what Hollings already said -- RTFA.)
In this case the right thing to point out is that they are tying the hands of hardware and software designers. Most specifically, open source software will no longer be legal... but neither will open hardware specs. Innovation will be constrained to follow whatever rules industry / the FCC comes up with.
Perhaps the right way to protest this bill is to point out that it will stifle innovation and erect an unreasonable barrier to entry into all computing markets.
If you want to argue fair use, point out that in developing this industry standard we will all be forced to comply with, there is absolutely no incentive to protect any fair use rights - essentially the individual consumer doesn't hold any cards. So unless we can trust the FCC to be a watchdog for consumer rights (har har har), fair use will be an almost certain casualty - whatever the bill says about protecting it.
Other flaws that come to mind:
1. The DMCA was supposed to bring us exciting new media content. It didn't. If Senator Hollings wants to tell us that that's the whole purpose of the SSSCA^H^H^H^H^HAICNQR (Acronym I Can Never Quite Remember), he's either lying to us or to himself, or just deluded by the unofficial sponsors of the bill (you know who I'm talking about).
2. One of Sen. Hollings's points is that broadband is offered to something like 60-80% of the American public but only 10-15% subscribe. (I could have messed up the figures - but I did read the article. (: ) He basically blames this national calamity on the assertion that Hollywood just hasn't made the 'net cool enough yet. Therefore, to make sure everyone gets wired, we need to create a safe world for Hollywood.
Stuff and nonsense! Whoever said that universal broadband was necessary or even desirable? If the consumer doesn't want it, well, maybe that's OK! Maybe broadband is just overpriced and the various cable monopolies / cartels need to be addressed, or something. But since when was it the role of the federal government to ensure the mass adoption of specific new technology? (OK, so the FCC tries to do this cf. HDTV, but honestly, cable modems?!?!?)
My point here is that Hollings is again being either deliberately deceptive or is himself deluded. If the stated goal of the SSSCA is to increase broadband use, the federal government is definitely stepping way out of its mandate. If not, it's a smoke screen and someone should probe Sen. Hollings's real movitavation. (Yes, you and I already know the real motivation. But anyway.)
3. Point out that the technology isn't feasible. From an electronics point of view, a camcorder pirated movie looks a great deal like a home movie. Watermarking is supposed to differentiate between the two, but all watermarking can be defeated so far, according to Professor Felton, who should know. Sen. Hollings seems to have a lot of faith in technology - but can he really guarantee that the industry will be able to produce consumer-grade technology which will successfully prevent camcorder movie piracy while still allowing camcorders? I wouldn't bet on it, and neither should Congress.
4. This bill is being billed as good for consumers. Shouldn't they get the opinions of a few consumers? Ask anyone: "If Congress passes a bill that makes MP3 players illegal, and many consumer electronics will go up $100 per unit to cover new features whose purpose is to prevent you from breaking the law, on the assumption that you would otherwise break the law, and you probably will not be able to burn your own `mix' compilation CDs as is perfectly legal today, and third-party plugins for Windows Media Player to do fancy oscilloscope displays and such will most likely disappear due to the new restrictions... but in return, you will be able to buy movies over the Internet and download them instead of having a DVD shipped to you... given that set of tradeoffs, would you be in favor of that?" This of course ignores many issues such as making free software illegal, but I'd still bet the average consumer would look at you funny and say "They'd never pass that sort of law!"
Semi-offtopic: my 13-year-old sister was in town yesterday. I happened to be wearing my copyleft DeCSS t-shirt, and she asked about it. I explained that the back of the shirt was an illegal computer program, a guy had gone to jail for it. She was puzzled: "Isn't that free speech?" And she's not (yet) even a geek! I was so proud! (: Just goes to show that even the "average American" (not that my sister is average, mind!) can "get it", given a few facts.
But in a world where a $15/hour job is insufficient to pay for adequate shelter, food, etc. and where
such a job can be yanked away at the slightest whim of "management," people have to be over-concerned
with money, and I think that is at least part of the problem.
<shrug> Glad I don't live in that world. I've been making less than $15/hr for less than 30 hr/wk for many years. People often ask me "what are you doing working here, you could be making $BIGBUCKS at $COMPANY" but frankly, I make a lot more than I spend so what'd be the point?
OK so I ride a bicycle and drive a PIIX3, but my point is that I think expectations for "reasonable standard of living" are grossly inflated in this country. Maybe I'm just insensitive to the reality of living in (say) Silicon Valley where the whole market is grossly inflated.
If it really uses any more than 1GB ram with file system cache, binary cache and compiler results I
would be surprised. That said, it's been a very long time since I've compiled it. Anyway, 6gb, 60gb, or
600gb isn't any different for this operation (aside from increased addressing time, possible transfer
delays, bank switching and other silly stuff).
"6 GB ought to be enough for anybody." (:
In this case, yes. But Anton's setup is a bit more memory-involved than one might think. First you have the 160 MB of kernel source code, which you certainly want in cache. Second, if the box is NUMA (and I don't know if it is or not), you really want a "local" file cache for each CPU set. For the 32 CPUs that means 4 or 8 copies of the 160 MB source - I think IBM is putting either 4 or 8 POWER4 cores on each die. (Think HyperThreading, only better and much more expensive.) Then you need to be able to run two or three dozen instances of gcc, itself a bit of a memory hog (and remember, this is a 64-bit architecture, where any program is a memory hog compared to its 32-bit equivalent), in parallel. The intermediate and output files will chew up another 100 MB or so.
I could see using 1.5 GB for this job. Not counting the GB you need for that Emacs process you are running on the side to browse the web, since Mozilla is just too lean 'n' mean.
Use executable software residing on the Workstation? Does that bar the running of any server type or P2P
type software that can respond to remote commands?
That hits on an interesting point. Back in the early days of NT, Microsoft started heavily investing in technology known as DCE/RPC, aka the Distributed Computing Environment specification for Remote Procedure Call. DCE/RPC is the framework for the whole NT domain system, MS Exchange, and other software from Microsoft, as well as a lot of third-party software.
For the less technically minded, RPC is a network paradigm similar to client/server, except that from the programmer's standpoint the client is not "making a request to a server" but rather "executing a procedure on the server". Thus an application protocol is defined in terms of functions, with passed parameters and return values. Exactly how to define these functions, and how to compile them into the client and server ends of a client/server app, depend on the RPC specification - in this case, DCE/RPC.
My point. Yes, I was getting to a point. Would this license clause then state that for all RPC traffic occurring with XP as the RPC server, the client must also be running XP?
(Furthermore, with DCE/RPC there is a "callback" provision so that the server can in some cases turn around and call a procedure on the client. If your application uses this facility, the above boldfaced statement gets even broader.)
If so, you can forget running remote regedit, among many other things, in heterogeneous networks....
Why didn't they just ban Netscape Navigator in the lisence agreement. Then they wouldn't have had to
write IE. What other competing software don't they want people to use? Put that into the lisence too.
You laugh, but they did put that into the NT 4 license (I think just NT Server, can't remember for sure). If you want to run a web server with more than a given number of hits per second, you must use IIS. This license was written before Apache was ported to Win32, but it must have sucked to be Netscape, who was selling two web server products at the time.
I haven't read the Win2k or newer server licenses, so I don't know if that clause is still in there.
So if you are running XP on the machine you are runnning VNC you do not violate the license
Well, if you are using XP Home Edition on your machine and you VNC over to a machine running XP Professional... you are probably in violation. Ditto if you are running XP Pro and you VNC over to a machine running XP Server. The latter is easy enough to imagine wanting to do.
(I say "imagine wanting to do" because I personally don't need to - we're refusing to deploy any XP until someone points out any reason to do so that would outweigh its various drawbacks. So far nobody has.)
Ya, heaven forbid a server OS has CAL's... just like any other commercial OS out there.
Except in this case the CAL is the same thing as the original L. In other words you're buying an entire copy of XP just to access it remotely. Other CALs are typically an order of magnitude cheaper than the OS. See the difference?
So if you want to use a VNC server on an XP server (what is the "NT Server" product name now?) you have to buy multiple copies of the server license. For several hundreds of dollars each. That, to me, is well to the north of unreasonable.
(Oh, and you can forget using the web-based VNC client to dial back to your server to check or change something while outside the machine room. Unless you want to buy XP Server licenses for all your desktops, even those that already run XP!)
As Microsoft keeps ratcheting license costs up and adding restrictions, an org has two basic choices: get an MS site license, or start migrating to other platforms. Depressingly often, the powers that be choose the former....
Codecs generally contain a lot of platform specific optimizations, and heavily leverage SIMD (AltiVec, SSE).
Granted.
Not only is it a lot cheaper to encode on a G4, P4, or Athlon, one would expect it to be a heck of a lot
faster as well.
Contrary to what you may have read in the glossies, Intel didn't invent SIMD. Cray Research had them beat by a couple decades at least.[*] I understand SPARC or at least SPARCv9 has an SSE-like instruction set. Come to think of it, Intel also didn't invent RISC, pipelining, CISC, branch prediction, 32-bit CPUs, i-caches, virtual memory, or even floating point.
[*] Actually I'm not sure on this point - AFAIK Cray's vector processing is basically SIMD, but in the likely event that I'm wrong, I welcome forcible enlightenment.
then it sure won't be GNU/Girlfriend, that is something that won't be shared
Sigh, another misunderstanding of the GPL. You don't have to distribute your GNU/Girlfriend at all. It's just that if you do, you also have to share a complete sample of her DNA.
The DNA sample can either be provided with the GNU/Girlfriend, or you can keep it on file and run off copies as people request them, for a reasonable media (test tube? DVD-ROM of sequence? not sure) duplication fee.
But Mozilla handles bad HTML GREAT. I have
seen JS that works great in mozilla, but crashes IE. The switch to gecko will cause less problems not
more !!
Ummm, I know about what the RFC's say about being liberal in what you accept. But personally I would rather that the most popular browser in the world be the most picky parser. Why? Because that way I don't get screwed by web designers who make silly typos and don't notice because they user $BROWSER_OF_THE_MONTH and I don't.[*]
Just like I'd prefer that everyone use standards-compliant browers without a lot of extensions à la Netscape 1.1 (for its day) or any flavor of MSIE. Because I don't want designers using those bleepin' extensions.
[*] ObBrowserPlug: For most uses I'm
very happy with links, a sort-of clone of Lynx with more features, saner defaults and fewer bugs.
"Cerberus was the watchdog of hell. There he lay, chained to the gates of Acheron, harassing the
spirits entering Hades and devouring those who tried to escape."
You sure it wasn't just because it ran Kerberos? (:
Actually this reminds me of Osiris. That was the general-purpose Unix server at Harvey Mudd College for a few years. It ran Solaris 2.2, back in the day, and anyone who's BTDT will remember exactly how stable Solaris 2.2 wasn't. The motd file went something like:
"Osiris was the ancient Egyptian god of the dead. He was killed many times, but each time rose again, more powerful than before."
then I SHOULD be able to take code released under the GPL, and use it any way I wish.
And so you can. Quoting the GPL:
Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.
In other words: unlike a lot of proprietary software, there are no restrictions on what you can use a GPL program for. You can load up as many copies onto as many computers as you wish, for example.
The only thing that is restricted in any way is redistribution, with or without modification.
And even distribution is allowed, subject to the restriction of distributing (or promising to distribute, if requested) source code to people to whom you give binary code. Oh, and you are not allowed to change the license terms. As you have received, so must you give, if you give at all. Which sounds fair to me....
Many of the comments on this thread might be summarized as follows: why is Microsoft doing this? The
answer is that we really want the ECMA standard to succeed (and that includes success for non-Microsoft
CLI implementations!) and we also want to seed the use of the CLI over the long haul.
It's been said before, but I just have to say it again.. The bigwigs at Microsoft have recently been singing the praises of the BSD license in contrast to the GNU General Public License.
So there you go. Look up "X Window System" some time for an example of a large software project promoted by large companies as a standard, whose reference implementation was distributed under a BSD-like license. X is similar to your CLI in that the proponents wanted widespread acceptance.
For a counterexample, do a web search on "Sun Community Source License" - specifically the reaction to same by the tech community. In short: nobody in the free software camp would have anything to do with it. If you want to promote something as a cross-platform standard, the SCCS is an excellent template for what not to put in your license.
GNU people do not get made when you download their software for free. Movie people do. Movie people want
to make you pay. GNU people don't. How is this a contradiction?
My giving you the right to copy my copyrighted work does not imply that I should have the right to copy your work without your permission. By and large we seem to be law-abiding people, even if some of us disagree with some particulars of the law (i.e. the insanely long copyright term).
GPL people don't mind if you copy our software and obey the constraints of the GPL. We do mind if you copy our software and do not abide by the GPL. You are then doing something we have not explicitly authorized.
So the contradiction here would be if I were to become indignant when you refuse to follow my copyright license, yet I blatantly ignore someone else's copyright license. That would be hypocrisy of the worst sort (well, all hypocrisy is of the worst sort, right?) and I would be surprised to find out that any serious GPL users (and by "users" I mean "using it for one's own work") have that flaw.
Slashdot Mirror
A coworker told me the exact same story when wondering why his password had expired. He was mildly pissed. I understood his frustration completely. In turn, he understood why I couldn't just make an exception for him.
At that point I thought of something, which I've never implemented but I think would be worthwhile. I think that when you change your password, the password strength checker should assign an expiration date based on entropy. If you want to use a password like "Cindy" - fine, but it will expire in 24 hours. If you use "a79xoibf", it will never expire. I'm assuming cracklib has a reasonable way to estimate password entropy.
Has this ever been implemented? I think, over time, such a system would encourage people to use good passwords. Having to remember a new password every week is a drag compared to keeping the same hard password for a year.
That doesn't absolve the proxy. They could figure out a way. Current behavior is broken, full stop. Perhaps the proxy's internal architecture makes it difficult to find a solution, but that is no excuse. If they have to modify the protocol between frontend and backend of the proxy ... so be it.
Toronto? Oh, I meant US$600, not CN$600.
(:
No, it doesn't, unfortunately. Two reasons:
1. "open source code" does not necessarily mean source code covered by the Open Source Definition. Their trademark application was dropped, remember. Al Gore's last political campaign made much of having an "open source web site" - sounds wonderful, right? - but as far as I could tell it didn't meet the OSD. I emailed the webmaster asking if they had cleared it with the Open Source Initiative so that they could use the (then-pending) trademark. No reply, of course.
2. The OSD does not cover patents, only copyrights. You can release code under (say) the GNU GPL which infringes someone else's patent. People will have the right to download, copy, and redistribute your software, but not to actually use it! Strange but true.
This is an important point. Law-abiding citizens, and organizations such as Software in the Public Interest (umbrella corp for the Debian Project) will not distribute software that you can't freely use. This is also why OpenSSH does not include support for the IDEA encryption algorithm - there is certainly an open-source implementation of it available, but the patent license is considered too restrictive. If you want to use IDEA with OpenSSH, you have to download an "unauthorised" third-party patch or distribution.
And that's what I'm guessing would happen here. The code may be freely redistributable, but would be encumbered by enough patents with restrictive patent licenses that as a practical matter, you won't be allowed to use it without agreeing to onerous licensing restrictions such as "Thou shalt not disable any DRM provisions" and such.
Many would say - who cares? As long as we can crack the DRM with help from the source? And I reply - I care. Much as I disagree with the provisions of the law, I won't just break it willy-nilly. I see a big difference in being able to do something and being legally allowed to. If something is SSSCA-verboten, I will be restricted in my actions, no matter how easy it is to hack the literal mechanisms. And assuming enough open-source developers hold a similar view, perhaps the necessary software won't be written.
Actually it's more like
DISNEY IP COP: Look, you can now buy this poster to put on your wall, for only $2/month. Great deal, eh? We used to charge $5/month.
They're not stupid. The carrot works better than the stick. If they can lower expectations to the point where Americans believe the pay-per-use model is normal and reasonable, that's what they'll do. Instead of people saying "ridiculous - I've already paid for this work, WTF would I want to pay per-listen or per-view or per-month?", I believe the media companies want us to eventually start thinking "Duh, of course you pay per-listen or per-view or per-month, how else would they make any money? And at only $.25 per-view this is a really good deal."
Remember, the vast majority of DVD player owners put up with region coding. "What, I can't buy that Japanese DVD because it won't play on my player? Oh well, guess I'll have to wait for the US release then." Pay-per-use might seem ridiculous now, but I believe people are complacent enough that it can be phased in. Depressing, yes....
Again - guess I'm glad I don't live there, wherever there is. Maybe Wichita, Kansas has a lower relative COL than I thought. In any case I typically spend less than $600 per month - and that buys a decent apartment, more than enough food, and miscellany.
Certainly I don't have a lot of "luxury" line items like cable tv, inflated energy bill from air conditioning, wireless, car payment, broadband, etc - but that's exactly my point: such things should be considered non-essentials. I think a lot of Americans believe they are essentials. In my opinion their standards are unreasonably high.
The contrapositive: many people would say I maintain a rather low standard of living and have unreasonably low expectations. Whatever. I don't feel poor.
Because the one with broken encryption will have a watermark on it.
Remember Professor Ed Felton? The movie industry tried to produce a watermarking scheme that would identify copy-protected (read: use-protected) content so that conformant hardware/software would refuse to violate its terms. They thought, and apparently some people still think, that it would be possible to produce a watermark that would survive various transforms such as decoding / re-encoding in lossy formats (JPEG, MP3, DivX) and even AD/DA conversion ("camcorder [movie] rips"). Perhaps it is possible. Not my area of expertise (if indeed I can be said to have one).
However, Felton and others have shown that given a little computer science, it is not all that hard to crack watermarking schemes - thus making your pirate movie look once again like a home recording, as far as your SSSCA-compliant player is concerned. Given an arms race between watermark creators and Felton, I'd put my money on Felton.
Actually the Circuit City format was known as "DivX" and the popular codec is called "DivX:-)". The codec's name is making fun of the Circuit City product. Beyond that the two have little in common. DivX:-) is actually a family of codecs, with the latest version being based largely on MPEG 4. And that's all I know about it....
Usually "Fair Use!" is the right rallying cry, but in this case I doubt it will be effective. They are already promising to uphold fair use of the content you purchase. (Yes I know, I don't trust them either, but that's what they'll say -- it's what Hollings already said -- RTFA.)
In this case the right thing to point out is that they are tying the hands of hardware and software designers. Most specifically, open source software will no longer be legal ... but neither will open hardware specs. Innovation will be constrained to follow whatever rules industry / the FCC comes up with.
Perhaps the right way to protest this bill is to point out that it will stifle innovation and erect an unreasonable barrier to entry into all computing markets.
If you want to argue fair use, point out that in developing this industry standard we will all be forced to comply with, there is absolutely no incentive to protect any fair use rights - essentially the individual consumer doesn't hold any cards. So unless we can trust the FCC to be a watchdog for consumer rights (har har har), fair use will be an almost certain casualty - whatever the bill says about protecting it.
Other flaws that come to mind:
1. The DMCA was supposed to bring us exciting new media content. It didn't. If Senator Hollings wants to tell us that that's the whole purpose of the SSSCA^H^H^H^H^HAICNQR (Acronym I Can Never Quite Remember), he's either lying to us or to himself, or just deluded by the unofficial sponsors of the bill (you know who I'm talking about).
2. One of Sen. Hollings's points is that broadband is offered to something like 60-80% of the American public but only 10-15% subscribe. (I could have messed up the figures - but I did read the article. (: ) He basically blames this national calamity on the assertion that Hollywood just hasn't made the 'net cool enough yet. Therefore, to make sure everyone gets wired, we need to create a safe world for Hollywood.
Stuff and nonsense! Whoever said that universal broadband was necessary or even desirable? If the consumer doesn't want it, well, maybe that's OK! Maybe broadband is just overpriced and the various cable monopolies / cartels need to be addressed, or something. But since when was it the role of the federal government to ensure the mass adoption of specific new technology? (OK, so the FCC tries to do this cf. HDTV, but honestly, cable modems?!?!?)
My point here is that Hollings is again being either deliberately deceptive or is himself deluded. If the stated goal of the SSSCA is to increase broadband use, the federal government is definitely stepping way out of its mandate. If not, it's a smoke screen and someone should probe Sen. Hollings's real movitavation. (Yes, you and I already know the real motivation. But anyway.)
3. Point out that the technology isn't feasible. From an electronics point of view, a camcorder pirated movie looks a great deal like a home movie. Watermarking is supposed to differentiate between the two, but all watermarking can be defeated so far, according to Professor Felton, who should know. Sen. Hollings seems to have a lot of faith in technology - but can he really guarantee that the industry will be able to produce consumer-grade technology which will successfully prevent camcorder movie piracy while still allowing camcorders? I wouldn't bet on it, and neither should Congress.
4. This bill is being billed as good for consumers. Shouldn't they get the opinions of a few consumers? Ask anyone: "If Congress passes a bill that makes MP3 players illegal, and many consumer electronics will go up $100 per unit to cover new features whose purpose is to prevent you from breaking the law, on the assumption that you would otherwise break the law, and you probably will not be able to burn your own `mix' compilation CDs as is perfectly legal today, and third-party plugins for Windows Media Player to do fancy oscilloscope displays and such will most likely disappear due to the new restrictions ... but in return, you will be able to buy movies over the Internet and download them instead of having a DVD shipped to you ... given that set of tradeoffs, would you be in favor of that?" This of course ignores many issues such as making free software illegal, but I'd still bet the average consumer would look at you funny and say "They'd never pass that sort of law!"
Semi-offtopic: my 13-year-old sister was in town yesterday. I happened to be wearing my copyleft DeCSS t-shirt, and she asked about it. I explained that the back of the shirt was an illegal computer program, a guy had gone to jail for it. She was puzzled: "Isn't that free speech?" And she's not (yet) even a geek! I was so proud! (: Just goes to show that even the "average American" (not that my sister is average, mind!) can "get it", given a few facts.
<shrug> Glad I don't live in that world. I've been making less than $15/hr for less than 30 hr/wk for many years. People often ask me "what are you doing working here, you could be making $BIGBUCKS at $COMPANY" but frankly, I make a lot more than I spend so what'd be the point?
OK so I ride a bicycle and drive a PIIX3, but my point is that I think expectations for "reasonable standard of living" are grossly inflated in this country. Maybe I'm just insensitive to the reality of living in (say) Silicon Valley where the whole market is grossly inflated.
"6 GB ought to be enough for anybody." (:
In this case, yes. But Anton's setup is a bit more memory-involved than one might think. First you have the 160 MB of kernel source code, which you certainly want in cache. Second, if the box is NUMA (and I don't know if it is or not), you really want a "local" file cache for each CPU set. For the 32 CPUs that means 4 or 8 copies of the 160 MB source - I think IBM is putting either 4 or 8 POWER4 cores on each die. (Think HyperThreading, only better and much more expensive.) Then you need to be able to run two or three dozen instances of gcc, itself a bit of a memory hog (and remember, this is a 64-bit architecture, where any program is a memory hog compared to its 32-bit equivalent), in parallel. The intermediate and output files will chew up another 100 MB or so.
I could see using 1.5 GB for this job. Not counting the GB you need for that Emacs process you are running on the side to browse the web, since Mozilla is just too lean 'n' mean.
That hits on an interesting point. Back in the early days of NT, Microsoft started heavily investing in technology known as DCE/RPC, aka the Distributed Computing Environment specification for Remote Procedure Call. DCE/RPC is the framework for the whole NT domain system, MS Exchange, and other software from Microsoft, as well as a lot of third-party software.
For the less technically minded, RPC is a network paradigm similar to client/server, except that from the programmer's standpoint the client is not "making a request to a server" but rather "executing a procedure on the server". Thus an application protocol is defined in terms of functions, with passed parameters and return values. Exactly how to define these functions, and how to compile them into the client and server ends of a client/server app, depend on the RPC specification - in this case, DCE/RPC.
My point. Yes, I was getting to a point. Would this license clause then state that for all RPC traffic occurring with XP as the RPC server, the client must also be running XP?
(Furthermore, with DCE/RPC there is a "callback" provision so that the server can in some cases turn around and call a procedure on the client. If your application uses this facility, the above boldfaced statement gets even broader.)
If so, you can forget running remote regedit, among many other things, in heterogeneous networks....
You laugh, but they did put that into the NT 4 license (I think just NT Server, can't remember for sure). If you want to run a web server with more than a given number of hits per second, you must use IIS. This license was written before Apache was ported to Win32, but it must have sucked to be Netscape, who was selling two web server products at the time.
I haven't read the Win2k or newer server licenses, so I don't know if that clause is still in there.
Well, if you are using XP Home Edition on your machine and you VNC over to a machine running XP Professional ... you are probably in violation. Ditto if you are running XP Pro and you VNC over to a machine running XP Server. The latter is easy enough to imagine wanting to do.
(I say "imagine wanting to do" because I personally don't need to - we're refusing to deploy any XP until someone points out any reason to do so that would outweigh its various drawbacks. So far nobody has.)
Except in this case the CAL is the same thing as the original L. In other words you're buying an entire copy of XP just to access it remotely. Other CALs are typically an order of magnitude cheaper than the OS. See the difference?
So if you want to use a VNC server on an XP server (what is the "NT Server" product name now?) you have to buy multiple copies of the server license. For several hundreds of dollars each. That, to me, is well to the north of unreasonable.
(Oh, and you can forget using the web-based VNC client to dial back to your server to check or change something while outside the machine room. Unless you want to buy XP Server licenses for all your desktops, even those that already run XP!)
As Microsoft keeps ratcheting license costs up and adding restrictions, an org has two basic choices: get an MS site license, or start migrating to other platforms. Depressingly often, the powers that be choose the former....
What is the "nether"? Sounds like the antimatter equivalent to the ether.[*] Which would make for an interesting pronunciation.
Granted.
Contrary to what you may have read in the glossies, Intel didn't invent SIMD. Cray Research had them beat by a couple decades at least.[*] I understand SPARC or at least SPARCv9 has an SSE-like instruction set. Come to think of it, Intel also didn't invent RISC, pipelining, CISC, branch prediction, 32-bit CPUs, i-caches, virtual memory, or even floating point.
Sigh, another misunderstanding of the GPL. You don't have to distribute your GNU/Girlfriend at all. It's just that if you do, you also have to share a complete sample of her DNA.
The DNA sample can either be provided with the GNU/Girlfriend, or you can keep it on file and run off copies as people request them, for a reasonable media (test tube? DVD-ROM of sequence? not sure) duplication fee.
I just think it would be hella cool to see a Microsoft CAB file containing "COPYING" and/or "COPYING.LIB". (:
Ummm, I know about what the RFC's say about being liberal in what you accept. But personally I would rather that the most popular browser in the world be the most picky parser. Why? Because that way I don't get screwed by web designers who make silly typos and don't notice because they user $BROWSER_OF_THE_MONTH and I don't.[*]
Just like I'd prefer that everyone use standards-compliant browers without a lot of extensions à la Netscape 1.1 (for its day) or any flavor of MSIE. Because I don't want designers using those bleepin' extensions.
Yes, 15 chars. NetBIOS names are 16 chars long but the last char is actually a byte representing service type.
You sure it wasn't just because it ran Kerberos? (:
Actually this reminds me of Osiris. That was the general-purpose Unix server at Harvey Mudd College for a few years. It ran Solaris 2.2, back in the day, and anyone who's BTDT will remember exactly how stable Solaris 2.2 wasn't. The motd file went something like:
"Osiris was the ancient Egyptian god of the dead. He was killed many times, but each time rose again, more powerful than before."
And so you can. Quoting the GPL:
In other words: unlike a lot of proprietary software, there are no restrictions on what you can use a GPL program for. You can load up as many copies onto as many computers as you wish, for example. The only thing that is restricted in any way is redistribution, with or without modification.
And even distribution is allowed, subject to the restriction of distributing (or promising to distribute, if requested) source code to people to whom you give binary code. Oh, and you are not allowed to change the license terms. As you have received, so must you give, if you give at all. Which sounds fair to me....
It's been said before, but I just have to say it again.. The bigwigs at Microsoft have recently been singing the praises of the BSD license in contrast to the GNU General Public License.
So there you go. Look up "X Window System" some time for an example of a large software project promoted by large companies as a standard, whose reference implementation was distributed under a BSD-like license. X is similar to your CLI in that the proponents wanted widespread acceptance.
For a counterexample, do a web search on "Sun Community Source License" - specifically the reaction to same by the tech community. In short: nobody in the free software camp would have anything to do with it. If you want to promote something as a cross-platform standard, the SCCS is an excellent template for what not to put in your license.
My giving you the right to copy my copyrighted work does not imply that I should have the right to copy your work without your permission. By and large we seem to be law-abiding people, even if some of us disagree with some particulars of the law (i.e. the insanely long copyright term).
GPL people don't mind if you copy our software and obey the constraints of the GPL. We do mind if you copy our software and do not abide by the GPL. You are then doing something we have not explicitly authorized.
So the contradiction here would be if I were to become indignant when you refuse to follow my copyright license, yet I blatantly ignore someone else's copyright license. That would be hypocrisy of the worst sort (well, all hypocrisy is of the worst sort, right?) and I would be surprised to find out that any serious GPL users (and by "users" I mean "using it for one's own work") have that flaw.