If the BBC, you can buy T-shirts, memorabilia, copies of shows, etc.
If the police, I've never been in the UK, and from what's been posted it would seem I couldn't expect to see such a blue police box on the streets. Even though it might have been a trademark at some point, it lost that status because it's status as a trademark was not maintained and protected.
Had they continually protected and maintained the blue police box as a unique feature or trademark, they might have had legal ground. For example, the RCMP (Royal Canadian Mounted Police) dress uniform is a very distinctive trademark and is appropriately protected. The protection isn't so much to make money for the RCMP as it is to protect the historical importance of that image and theoretically to ensure that it isn't abused for profit by a person or business that didn't build that reputation.
As to "Damn cops", everyone has the right to their own prejudice. I've never had any problems with the police -- give respect and you get it back. Most of them are just regular people who spend an awful lot of time dealing with whackos, drunks, crazies, domestic arguments, and a lot of other situations where they have no assurance it's not going to blow up in their face. Give them attitude and they have to assume you're a problem and will react accordingly.
Don't get me wrong -- there are a few gung-ho gung-ho jarheads and corrupt/prejudiced idiots with badges. They just aren't the norm.
It's been ages since I heard about RMS doing anything but rant about GNU/Linux. I caught his interview on "The Screen Savers" on tech TV last night, and was very disappointed to see he was so "tired" he couldn't hold a coherent dialogue. I grant the host was doing his best to rattle him, but last time I saw Rick involved in such I dialogue he held his own.
As to Palladium, well of course it's a "bad thing". I sure as hell don't want my systems frelled for the sake of Hollywood or the RIAA -- I use standalone DVD and CD/SACD players for that. Hands off my computers, my software, and my livelihood!
You are entitled to misunderstanding what security servers are, how redundancy prevents the issues you raise, and how large scale systems are secured.
I can only hope that your clients listen to more knowledgeable sources than you before they make decisions. While smaller environments might not be able to justify the expense of full security servers, they should do so with full information about the risks they incur by doing so.
In the meantime, please get your brain off that "single point of failure" focus. That is what distributed security services and authentication databases are designed to prevent. Go over to IBM's web site and read some redbooks on Kerberos architectures and DCE, think about what they are saying, and I think you will eventually realize that your concerns have been addressed for well over a decade.
You might also be thinking that I am advocating a web-wide security service, such as Microsoft tried to force on everyone. If so, your assumption is incorrect -- no one company should have that kind of power and responsibility.
Can you really imagine going to the movies as a family without munchies? When I was a kid movies were a "treat" and usually included the mammoth family popcorn tub, drinks (sometimes shared), and a candy bar.
Now you can't get those family popcorn tubs, even though prices have skyrocketed. You can't get a standard size candy bar, only the oversize packs. You pay more for a 20oz watery soda with too much ice than for 2-3 large bottles at 7-11, much less a grocery store.
Despite that obscene gouging, I cannot imagine being so tight-fisted that I wouldn't buy the kids in the group the same as I got as a child. I'm not saying you need to give in to the whining for more candy when it's gone, but it just wouldn't be right to eliminate it. Even at home I usually nuke a bag of popcorn when I sit down to watch a flick on DVD!
The problem with your R&D analogy is that the only R&D occurs with companies producing SFX, film stock, cameras, etc. The movies and albums themselves are "art" or "expression."
The MPAA and RIAA don't produce anything, but they speak on behalf of their industries, and are a means of referring to their members as a whole (The "A" at the end of the acronyms refers to "Association".)
They are bad at marketing. All they market is their biggest budget efforts, with little to no regard for quality. The only exception I can recall was "The Blair Witch Project", which was done on a very low budget compared to movies or records that usually get the push.
When is the last time you saw an ad for a movie that didn't have at least one multi-million dollar star involved? When was the last time you saw a band that didn't fit a top-40 profile for a non-major genre get promoted? How many times now have we found out that a "band" was actually a fraud that was lip-synching or so heavily processed that the singer in concert sounds nothing like the album?
The problems that occur with the developere edition do not occur with the commercial or demo versions. It is a weirdity of the developer build, and I don't want to give people the idea that there is a generic problem with Sybase 12.5. This is a very special case problem and they are working on a resolution.
How can I get the importance of one-way hashing at the security services database through to you? It is a technology that does not actually store your password on the server at all! Instead, it relies on using your password as a seed value to other information unique to the server to produce a hash value. Only that hash value is stored on disk.
The algorithms used for one way hashing on security servers combine the user's password, a security server seed, and potentially other seeds to produce the value stored to disk. Even if you swipe the "password file" from the security servers, you lack the seeds necessary to even try a brute force attack to produce the key. Because multiple passwords will hash down to the same value, the only way to even know you have "the" password is if you have multiple databases being crunched (you cannot restore the same security data files to a server that is configured with a different seed.)
You also need to read up on what terms like "trusted client" means in the realm of computer security. Generally, it means that the client resides in a physically secure environment where it is guaranteed to be maintained, such as in a corporate data server subnet.
Where you are at risk with centralized authentication services is poorly implemented and maintained security servers. This is not a task managed by a typical sysadmin nor by an MCSE. Why did you think security infrastructure specialists are so damned expensive, even in the post-dot-bomb era? (No, I am not such a specialist, but I learned most of what I know about the topic from people who are security/cryptography specialists. They start talking in set/pattern/probability mathematics that make my head hurt, which is why I never went into that specialty.)
It costs about $8-9 USD for a ticket for one person. If a couple is going, that's $16-18 just to get in the door, and there are a rather large number of DVDs that sell for that price (including new releases.)
Anyone with kids is hopelessly punished by the ticket prices, not to mention the confectionary stand. (Suuuure you can convince the kids to skip that $2 medium drink and those $3.50 candies!)
And for what? To have your feet stick to the floor? To listen to the idiot with the cell phone, or the couple/group that spend more time talking than watching? Perhaps for the joy of screaming "Focus! Focus!" when the monkey upstairs in the projection booth lets everything go fuzzy?
As to "going bankrupt", maybe Hollywood's big money directors and stars will be forced to do what many of us in the tech industry did last year -- take a pay cut in order to keep working. I realize 10-15% cuts for them amount to a few million dollars a year in some cases, but they can afford it far better than "normal" people can.
And if I hear another MPAA or RIAA exec trying to justify the prices as being necessary to cover the costs of producing the "failures", I think I'm going to puke. No other industry I know of tries to justify their costs by pointing to perpetual mis-management, poor marketing, and poor salary negotiation skills. It's called "ROI" people, and if you can't grasp that basic concept and deal with it you should be out of business!
If i have a centralised account on one company and they mess up all my accounts is screwed.
Which is why redundancy is so important for the security servers. Facilities like DCE use a distributed cluster of authentication/authorization based on Kerberos for just this reason.
I would much rather have an application that held all my accounts in it encryptet on the HD like in Mozilla. Then they have to break into my computer to get them.
Client-based security implies that your clients are trusted systems, which is very, very bad as a security model. In the case of Mozilla, what it actually keeps is a copy of your passwords, which means that if that data file is stolen it can be decrypted. In the case of facilities like Kerberos, the password is only stored on disk after a one-way hash, which is much harder to crack.
Then they have to break into my computer to get them. If they do that no Passport or liberty in the world will save me anyway since they can look at what i type and then logon by themselves with my account.
If they break into your client computer and steal your Mozilla password file, they don't have to remain connected to crack your passwords, which minimizes their exposure to tracing. They also don't need to modify any files to do this, which makes it much harder to detect the intrusion in the first place.
They can only "look at what I type" if they install a trojan or virus that is going to capture and transmit your keystrokes. That is much easier to detect and prevent via anti-virus software and application protection software (e.g. Norton, NetworkICE) than a read-only file transmission.
I know Passport is irrelevant as far as the browser client sees (it's all buried in tickets/cookies), but I'm talking about "thick" client applications. You log in to Passport on a WinXX box once because it is shared by all applications through Microsoft internals. Using HTTP over SSL from a client would not allow that shared authentication behavior because the information is not shared by the clients in any standard fashion.
That also means that if you have different browsers started, you need to log in for each one seperately, which negates one of the key points of a single-login facility.
If you download the Sybase 12.5 ASE Developer edition, you'll find you can't get it to work under SuSE 8.0 or Mandrake 8.2 because it only works with a very specific glibc patch level under RedHat. Of course you can't run it if you update your RedHat installation, either.
But that is precisely the kind of problem I'm concerned about. When you are dealing with any sort of security services, you do not want server patches held off because of a package dependancy that you have no way to work around. Some of the patches that don't get along with ASE 12.5 are rather important security patches, for example.
A properly set up centralized security server does nothing but run the authentication services and possibly the authorization services. It should not be running other services such as NFS, print services, HTTP servers, etc.
That is not to say that different applications can't use secondary passwords to authorize certain facilities, or to mandate a seperate security ticket for the duration of a special session (e.g. starting an admin tool to add new users to the application's authorization set, or changing their authorization lists.)
Many authentication and authorization services also support facilities like session limits (the place I work at right now only allows each id to be used for a single client station at a time; development and support staff are a special case.)
Centralizing security also means that you only have to deal with hardening one set of authentication servers (gotta have redundant server clusters in a large environment for something this critical!) When patches are needed, you know they've been applied because you don't have to run around to all the application, data, and web server systems. Some application/web servers might break if they aren't patched to work with the updated security server, but that is a good thing -- you don't want outdated clients being authenticated when they're running software that has known security issues.
However, there are far better products out there that aren't limited to Microsoft clients. Tools like Kerberos, Verisign products, Netware, etc. I just cannot fathom why anyone would voluntarily limit their options rather than just using a non-Microsoft product.
Interesting how such a limited platform list is provided for servers. What about Mandrake, SuSE, et. al.? With canned commercial support, what of potential customers that want to use a different HTTP server, different patches, different languages/tools, etc.?
You also note that there is no mention of support for developing client software under the *nix platforms. It's yet another way to lock in the desktop as Microsoft-only, much as many of their "servers" already do.
You cannot hear the difference between good quality 10-12 guage lamp cord and $12/ft oxygen-free copper "certified digital" wires. The first key is "certified digital", which has jack shit to do with the analogue currents that run through the speaker wiring, and is just marketing buzzwords. The second is that all you usually get at $12/ft vs $0.12/ft. is a shielded jacket with painted arrows and cheap terminators (jacks/plugs/spades.)
Jump into the bottom end of audiophile cabling, such as MIT's old PC2 cabling at around $30-40/ft., and you will start hearing a difference in how clean the low frequency bass is controlled (assuming the rest of the equipment matches.)
Your comparison in a "corporate boardroom" is completely useless. I guarantee the speakers were buried in the walls or ceiling, not properly aligned drivers in tuned cabinets with high quality crossovers. Even the shittiest of speakers can have their sound improved significantly through proper placement. Typically you want them in the neighbourhood of 2-3 feet from the back wall and 2-3 feet from sidewalls in the case of typical bass reflex designs, with tweaking the positioning having a dramatic effect as you try to minimize standing bass waves.
A $1500 THX certified amp is not audiophile equipment. It is mid-fi, mass produced, and about half the price I paid for my amplifier (and mine is about as cheap as you get for what would even be considered by audiophiles.) Think about it -- that amp has at least 5 channels, which means the amplifier components themselves are roughly comparable to a $600 stereo amplifier. Ten years ago even Pioneer and Onkyo mid-fi amps cost more than that!
That $295.00 Sony DVD player you are comparing to actually has the same high-quality DACs used in their top of the line equipment. What you get with the top of the line SA series from Sony is discrete output circuitry instead of chip amps. With the amplifier and speaker setup you were using, I would be shocked if you could have heard the difference.
I picked up a Sony DVD player that happens to do SACD as well (my old player died.)
I have several Mobile Fidelity CDs in my collection, which typically cost me $5-7 more than regular CDs. The use of gold in the discs was more gimic than useful, but the care they took in doing the A/D conversion to avoid clipping led to a tremendous improvement in sound quality.
I have yet to buy even one SACD, despite having the equipment to make use of the format (and hear the difference.) Clearly it isn't the price difference, or I wouldn't have paid extra for MFSL CDs when they were in business.
Could it be the complete and utter lack of material that I want to hear? Nothing but classical, jazz, and modern "artists" who have been so over-processed that any sense of realism isn't on the master in the first place! Not one decent transfer of single-miked blues, rock, live performances, etc.
Billy Idol "VH1 Storytellers", Days of the New first album, Blondie's recent "No Exit", Eric Clapton "Unplugged" -- any of these albums would probably be worth upgrading to SACD, if they were available.
As usual, it is the record companies themselves that are killing improved technology by their own blatant stupidity. How many years did it take for their boneheaded audio engineers to realize that pushing A/D converters to the point of clipping sounded like crap compared to the soft-clipping that occured with analogue media? Even now I still hear a lot of CDs that were obviously mastered by people who just don't understand the importance of avoiding digital clipping.
I don't want to hear that the sound is better from any "audiophiles" either. Audiophiles are the same morons who bought distilled water from discWasher for $5 for 4 ounces and buy "directional" speaker cables today and who use a green magic marker on the rim of their CDs. And then claim to be able to hear phase-shift distortion in CD music.
Most self-proclaimed "audiophiles" who buy crap like DiscWasher or played with the green magic markers are actually mid-fi enthusiasts who think that multi-channel amps are "audiophile" equipment. True audiophile equipment is just basic, clean designs manufactured in small quantities with extra high quality components.
To try to make a reasonable analogy, let me try comparing "beef products." Sony, Pioneer, JVC, are the McDonalds/Burger King of audio, with Carver, Polk, Infinity, et. al. being comparable to Wendy's (i.e. the latter is a little better, but still mass market fast food.) Real audiophile equipment from Conrad-Johnson, Sonic Frontiers, et. al. are prime steak at a 4-5 star restaurant in comparison, with a corresponding increase in price and service.
If all you want is "something to eat" (listen to), don't care about ambience (soundstage, depth, air, detail), or dump ketchup all over everything (failure to properly position speakers, seating, use quality cabling, etc) you're not going to be impressed with a $30 steak compared to a $5 burger.
I and many others do read articles before commenting. I happen to agree with many of Rick's philosophies, but not the way he has been presenting them over the past 4-5 years.
His fame and founding-father status with GNU mean that he gets a lot of respect and press, even from those who disagree with his ideals. But when he ignores whether a forum is appropriate and storms on with his idealistic rants, he provides the anti-open-source community with a poster boy to paint the community as a bunch of fanatics.
Some claim that his persistant ranting is a "challenge" for us all to follow our ideals and morals regardless of the consequences. While this is a terrific sentiment for those who wish to be seen as dedicated religious icons, it is hardly a suitable approach for those who don't live in their office as RMS does.
But in a world where "reality" TV is so popular, I guess I shouldn't be surprised. Realism is the last thing the sound-bite hungry populace seems to want, and a few "sure win" discussions like this ensure that the page counts stay high enough to improve the eyeball stats for advertising.
We will never know the answer to this puzzler because he is the only person in the world to get slashdot headlines by posting flamebait to the wrong news groups. Anyone else would either be ignored, flamed for cross-posting, or deleted by the moderators.
AmigaDOS was actually created as a masters thesis project in operating systems at a university in England, IIRC (it could have been a different country.)
It did not have Unix-style commands, APIs, or underpinnings. There were a lot of add-on programs created to give it shell-like functionality, and it supported ideas like process parentage and priorities, but no one who has ever done systems programming on a *nix system would confuse it with a *nix core.
The Amiga died due to Commodore's pathetic marketing. Period.
Eventually DEC relented and came out with their own version of Unix for the Vax based mostly on BSD and called it Ultrix.
There is no "mostly" about it. Ultrix was BSD 4.1, released after we had been using BSD 4.2 on our VAX systems in University. Many of the manpages still had the BSD id strings, as did many header files. Running "strings" on libraries often reported BSD authorship as well. The final clincher was that all the bugs we had before upgrading to 4.2 returned with Ultrix (kernel panics, etc.)
Not really surprising, when you think about it. DEC was famous for providing "bug-for-bug" compatible VAX CPUs on the various models, because fixing the bugs would have broken too much code.
SVR4 added significant extensions to the data carried by signal handling, IPC, etc. when compared to SVR3. Useful concepts from the BSD family were used, as were a few from Xenix and SunOS.
The scheduling algorithms, memory management, configuration, and other features of SVR4 were sufficiently different that porting BSD code to the default APIs was a "challenging" project. Apps which used shared memory, semaphores, etc. required significant changes to run reliably.
The SVR4 environments I used (Sequent Dynix/PTX, Solaris, and another marginal player whose name I forget at the moment) each had BSD compatability libraries that would let you recompile your apps with minimal pain. I believe those compatability layers were largely vendor-specific rather than a generic feature of SVR4.
With Dynix/PTX, for example, you had a couple commands which let you switch between defaulting to BSD or SVR4 commands and libraries. You could blend the two environments in your code if you had to, but it was often tricky stuff.
Solaris did not use the same approach as Dynix/PTX, and actually seemed to make it harder to port old SunOS code than it did to recompile the same code under Dynix/PTX.
That unnamed OS had compatability libraries for recompiling code, but only supported SVR4 commands.
It would be a mistake to assume that the incorporation of good ideas from various *nix systems to be equivalent to merging them (which is what I read the "+" of the systems you mention to mean.) Remember that SunOS had a lot of proprietary code as well as the BSD base, and I seriously doubt too much of that was available for use in SVR4 (it could have happened, I just question the likelihood as that was the peak of the Unix incompatability wars.)
Although I've used HP-UX 9/10 and AIX (3.5/6?), I didn't work with them at the same low levels I did SunOS, Solaris, Dynix/PTX, "pure" SVR4, or Linux.
<OffTopicRamble>
How low-level was the work with Dynix/PTX?
By the time we were done with a two-year project, a 32 processor box was being bottlenecked by I/O capacity and the system bus, not the application code. The subsystem I wrote used X.25 data collection slaves, file writers similar to database redo logs, and shared memory under the coordination of a master/monitor process used to dynamically start/stop the processes. The first round of the functionality written using Tuxedo had proven to just be too slow, and had to be tweaked to the max.
That application almost beat out some work I did under VMS as far as maximizing system throughput. The work on VMS was a much simpler task, but I managed to get the I/O channels on three tape devices, five spindles of disk, the memory utilization, and CPU utilization on a VAX8600 (not positive about that model number) to all stay in the 95-98% load range...
The old SunOS was based on BSD. Solaris is based on an SVR4 core, which is what made the transition so painful (different APIs for signal handling, etc.)
I stuck through the first episode, but I'll be shocked if this effort survives a month. Even the anime they ripped off (Cowboy Bebop, et. al.) had more depth of character in the first episode than Firefly.
I kept thinking that somewhere Roddenberry is spinning in shame that someone took the "wagontrain to the stars" phrase literally and tried to make a show about it.
A train robbery? The whore with a heart of gold? Six guns? Shit, why didn't they tie the damned ship to a tree as an anchor to finish off the list of lame cliches!
Fox killing off a solid show like Dark Angel for this drek, and SciFi trying to kill off Farscape just proves that there truly are no "entertainment" execs who have a clue. Then we've got John Doe which looks like a cheap last-minute knock-off of The Dead Zone (I'll give it a chance, but have zero hope of it being worthwhile.)
I wish Fox would take a clue from their cable sister "FX", which stuck by The Shield despite objections from advertisers and the religious groups. While it might not be everyone's idea of a great show, it's the only time in the past decade or so that I've seen a network support a show that deserved the support.
And these morons are worrying that people might pirate their shows with digital HDTV feeds???!!! *LOLOLOLOL* Don't worry, Mr. TV Executive -- with this kind of drek you really don't need to worry about piracy...
The original DirecTV stream saved by DirecTiVO is a modified 480x480 NTSC SVCD MPEG2 VBR stream, running at above-normal bitrates. I've heard there are utilities that will readjust the data stream to use standard MPEG2 VBR formats so that they can be dumped off for later viewing.
The problem with that is that the vast majority of set-top players can't handle that high an SVCD stream rate, so playback is an issue. If you copy the SVCD data to HDD for playback, the data rate isn't an issue, but you can't do that with a set-top player.
Some people have had luck with using MPEG editors to clip commercials from the data stream, tie the two MPEG segments together, and burning them off to DVD-R. Some set-top players can deal with that and will display a full NTSC screen with such a source, but it is not a standard DVD format so there are no guarantees.
Any direct data stream work assumes you're going to muck around with various kernel patches and stuff to disable the encryption of the data during the write to the HDD.
Personally I just settled for doing S-VHS wiring to an ASUS GF2MX card with video capture. Edit the HuffYUV files with VDub, and save as DiVX. Takes hours of CPU (PIII933), but it works. I find quality based 1-pass at 2.1 or 2-pass at 1500+ is indistinguishable from the original source, provided you've got enough CPU to turn some of the playback filtering on.
Eventually I got tired of wasting my time with all the editing and stuff, so the box is back to running SuSE 8.0. It was a very educational few months, though, and I learned a lot about video processing, filtering, and formats. I ended up with a few seasons worth of series archived, but movies are easier to just buy.
If finances permit, it's a lot cheaper to just buy season box sets, even at $80/season. Even at 15 minutes to edit out commercials, 20 episodes/season, you're still looking at about 4 hours of manual labour to capture and save a season. Add some more time for burning (say an hour), a reasonable pay scale in the tech industry, and it costs less to buy DVDs that use higher resolution and a much higher data rate.
I believe that where the cooling is actually needed is on the chips that do the processing (at least with DirecTiVO.) The rather small cooler gets too hot to keep your finger on when the anti-pirate code cycles are happening.
I had been thinking in terms of either carving the case so a small fan would blow down directly on the heatsink involved, and/or replacing the puny heatsink with something larger (maybe even copper.)
One thing that is abundantly clear to me is that the so-called temperature monitoring isn't even as accurate as what is done with mobo based CPU temp monitoring.
Slashdot Mirror
If the BBC, you can buy T-shirts, memorabilia, copies of shows, etc.
If the police, I've never been in the UK, and from what's been posted it would seem I couldn't expect to see such a blue police box on the streets. Even though it might have been a trademark at some point, it lost that status because it's status as a trademark was not maintained and protected.
Had they continually protected and maintained the blue police box as a unique feature or trademark, they might have had legal ground. For example, the RCMP (Royal Canadian Mounted Police) dress uniform is a very distinctive trademark and is appropriately protected. The protection isn't so much to make money for the RCMP as it is to protect the historical importance of that image and theoretically to ensure that it isn't abused for profit by a person or business that didn't build that reputation.
As to "Damn cops", everyone has the right to their own prejudice. I've never had any problems with the police -- give respect and you get it back. Most of them are just regular people who spend an awful lot of time dealing with whackos, drunks, crazies, domestic arguments, and a lot of other situations where they have no assurance it's not going to blow up in their face. Give them attitude and they have to assume you're a problem and will react accordingly. Don't get me wrong -- there are a few gung-ho gung-ho jarheads and corrupt/prejudiced idiots with badges. They just aren't the norm.
It's been ages since I heard about RMS doing anything but rant about GNU/Linux. I caught his interview on "The Screen Savers" on tech TV last night, and was very disappointed to see he was so "tired" he couldn't hold a coherent dialogue. I grant the host was doing his best to rattle him, but last time I saw Rick involved in such I dialogue he held his own.
As to Palladium, well of course it's a "bad thing". I sure as hell don't want my systems frelled for the sake of Hollywood or the RIAA -- I use standalone DVD and CD/SACD players for that. Hands off my computers, my software, and my livelihood!
For me the munchies are a tradition, but making it a "fun" meal and a movie is a good alternative I wouldn't have thought of.
You are entitled to misunderstanding what security servers are, how redundancy prevents the issues you raise, and how large scale systems are secured.
I can only hope that your clients listen to more knowledgeable sources than you before they make decisions. While smaller environments might not be able to justify the expense of full security servers, they should do so with full information about the risks they incur by doing so.
In the meantime, please get your brain off that "single point of failure" focus. That is what distributed security services and authentication databases are designed to prevent. Go over to IBM's web site and read some redbooks on Kerberos architectures and DCE, think about what they are saying, and I think you will eventually realize that your concerns have been addressed for well over a decade.
You might also be thinking that I am advocating a web-wide security service, such as Microsoft tried to force on everyone. If so, your assumption is incorrect -- no one company should have that kind of power and responsibility.
Can you really imagine going to the movies as a family without munchies? When I was a kid movies were a "treat" and usually included the mammoth family popcorn tub, drinks (sometimes shared), and a candy bar.
Now you can't get those family popcorn tubs, even though prices have skyrocketed. You can't get a standard size candy bar, only the oversize packs. You pay more for a 20oz watery soda with too much ice than for 2-3 large bottles at 7-11, much less a grocery store.
Despite that obscene gouging, I cannot imagine being so tight-fisted that I wouldn't buy the kids in the group the same as I got as a child. I'm not saying you need to give in to the whining for more candy when it's gone, but it just wouldn't be right to eliminate it. Even at home I usually nuke a bag of popcorn when I sit down to watch a flick on DVD!
Ok, so you aren't an MPAA or RIAA exec.... *g*
The problem with your R&D analogy is that the only R&D occurs with companies producing SFX, film stock, cameras, etc. The movies and albums themselves are "art" or "expression."
The MPAA and RIAA don't produce anything, but they speak on behalf of their industries, and are a means of referring to their members as a whole (The "A" at the end of the acronyms refers to "Association".)
They are bad at marketing. All they market is their biggest budget efforts, with little to no regard for quality. The only exception I can recall was "The Blair Witch Project", which was done on a very low budget compared to movies or records that usually get the push.
When is the last time you saw an ad for a movie that didn't have at least one multi-million dollar star involved? When was the last time you saw a band that didn't fit a top-40 profile for a non-major genre get promoted? How many times now have we found out that a "band" was actually a fraud that was lip-synching or so heavily processed that the singer in concert sounds nothing like the album?
The problems that occur with the developere edition do not occur with the commercial or demo versions. It is a weirdity of the developer build, and I don't want to give people the idea that there is a generic problem with Sybase 12.5. This is a very special case problem and they are working on a resolution.
How can I get the importance of one-way hashing at the security services database through to you? It is a technology that does not actually store your password on the server at all! Instead, it relies on using your password as a seed value to other information unique to the server to produce a hash value. Only that hash value is stored on disk.
The algorithms used for one way hashing on security servers combine the user's password, a security server seed, and potentially other seeds to produce the value stored to disk. Even if you swipe the "password file" from the security servers, you lack the seeds necessary to even try a brute force attack to produce the key. Because multiple passwords will hash down to the same value, the only way to even know you have "the" password is if you have multiple databases being crunched (you cannot restore the same security data files to a server that is configured with a different seed.)
You also need to read up on what terms like "trusted client" means in the realm of computer security. Generally, it means that the client resides in a physically secure environment where it is guaranteed to be maintained, such as in a corporate data server subnet.
Where you are at risk with centralized authentication services is poorly implemented and maintained security servers. This is not a task managed by a typical sysadmin nor by an MCSE. Why did you think security infrastructure specialists are so damned expensive, even in the post-dot-bomb era? (No, I am not such a specialist, but I learned most of what I know about the topic from people who are security/cryptography specialists. They start talking in set/pattern/probability mathematics that make my head hurt, which is why I never went into that specialty.)
It costs about $8-9 USD for a ticket for one person. If a couple is going, that's $16-18 just to get in the door, and there are a rather large number of DVDs that sell for that price (including new releases.)
Anyone with kids is hopelessly punished by the ticket prices, not to mention the confectionary stand. (Suuuure you can convince the kids to skip that $2 medium drink and those $3.50 candies!)
And for what? To have your feet stick to the floor? To listen to the idiot with the cell phone, or the couple/group that spend more time talking than watching? Perhaps for the joy of screaming "Focus! Focus!" when the monkey upstairs in the projection booth lets everything go fuzzy?
As to "going bankrupt", maybe Hollywood's big money directors and stars will be forced to do what many of us in the tech industry did last year -- take a pay cut in order to keep working. I realize 10-15% cuts for them amount to a few million dollars a year in some cases, but they can afford it far better than "normal" people can.
And if I hear another MPAA or RIAA exec trying to justify the prices as being necessary to cover the costs of producing the "failures", I think I'm going to puke. No other industry I know of tries to justify their costs by pointing to perpetual mis-management, poor marketing, and poor salary negotiation skills. It's called "ROI" people, and if you can't grasp that basic concept and deal with it you should be out of business!
They can only "look at what I type" if they install a trojan or virus that is going to capture and transmit your keystrokes. That is much easier to detect and prevent via anti-virus software and application protection software (e.g. Norton, NetworkICE) than a read-only file transmission.
I know Passport is irrelevant as far as the browser client sees (it's all buried in tickets/cookies), but I'm talking about "thick" client applications. You log in to Passport on a WinXX box once because it is shared by all applications through Microsoft internals. Using HTTP over SSL from a client would not allow that shared authentication behavior because the information is not shared by the clients in any standard fashion.
That also means that if you have different browsers started, you need to log in for each one seperately, which negates one of the key points of a single-login facility.
But that is precisely the kind of problem I'm concerned about. When you are dealing with any sort of security services, you do not want server patches held off because of a package dependancy that you have no way to work around. Some of the patches that don't get along with ASE 12.5 are rather important security patches, for example.
A properly set up centralized security server does nothing but run the authentication services and possibly the authorization services. It should not be running other services such as NFS, print services, HTTP servers, etc.
That is not to say that different applications can't use secondary passwords to authorize certain facilities, or to mandate a seperate security ticket for the duration of a special session (e.g. starting an admin tool to add new users to the application's authorization set, or changing their authorization lists.)
Many authentication and authorization services also support facilities like session limits (the place I work at right now only allows each id to be used for a single client station at a time; development and support staff are a special case.)
Centralizing security also means that you only have to deal with hardening one set of authentication servers (gotta have redundant server clusters in a large environment for something this critical!) When patches are needed, you know they've been applied because you don't have to run around to all the application, data, and web server systems. Some application/web servers might break if they aren't patched to work with the updated security server, but that is a good thing -- you don't want outdated clients being authenticated when they're running software that has known security issues.
However, there are far better products out there that aren't limited to Microsoft clients. Tools like Kerberos, Verisign products, Netware, etc. I just cannot fathom why anyone would voluntarily limit their options rather than just using a non-Microsoft product.
Interesting how such a limited platform list is provided for servers. What about Mandrake, SuSE, et. al.? With canned commercial support, what of potential customers that want to use a different HTTP server, different patches, different languages/tools, etc.?
You also note that there is no mention of support for developing client software under the *nix platforms. It's yet another way to lock in the desktop as Microsoft-only, much as many of their "servers" already do.
You cannot hear the difference between good quality 10-12 guage lamp cord and $12/ft oxygen-free copper "certified digital" wires. The first key is "certified digital", which has jack shit to do with the analogue currents that run through the speaker wiring, and is just marketing buzzwords. The second is that all you usually get at $12/ft vs $0.12/ft. is a shielded jacket with painted arrows and cheap terminators (jacks/plugs/spades.)
Jump into the bottom end of audiophile cabling, such as MIT's old PC2 cabling at around $30-40/ft., and you will start hearing a difference in how clean the low frequency bass is controlled (assuming the rest of the equipment matches.)
Your comparison in a "corporate boardroom" is completely useless. I guarantee the speakers were buried in the walls or ceiling, not properly aligned drivers in tuned cabinets with high quality crossovers. Even the shittiest of speakers can have their sound improved significantly through proper placement. Typically you want them in the neighbourhood of 2-3 feet from the back wall and 2-3 feet from sidewalls in the case of typical bass reflex designs, with tweaking the positioning having a dramatic effect as you try to minimize standing bass waves.
A $1500 THX certified amp is not audiophile equipment. It is mid-fi, mass produced, and about half the price I paid for my amplifier (and mine is about as cheap as you get for what would even be considered by audiophiles.) Think about it -- that amp has at least 5 channels, which means the amplifier components themselves are roughly comparable to a $600 stereo amplifier. Ten years ago even Pioneer and Onkyo mid-fi amps cost more than that!
That $295.00 Sony DVD player you are comparing to actually has the same high-quality DACs used in their top of the line equipment. What you get with the top of the line SA series from Sony is discrete output circuitry instead of chip amps. With the amplifier and speaker setup you were using, I would be shocked if you could have heard the difference.
I picked up a Sony DVD player that happens to do SACD as well (my old player died.)
I have several Mobile Fidelity CDs in my collection, which typically cost me $5-7 more than regular CDs. The use of gold in the discs was more gimic than useful, but the care they took in doing the A/D conversion to avoid clipping led to a tremendous improvement in sound quality.
I have yet to buy even one SACD, despite having the equipment to make use of the format (and hear the difference.) Clearly it isn't the price difference, or I wouldn't have paid extra for MFSL CDs when they were in business.
Could it be the complete and utter lack of material that I want to hear? Nothing but classical, jazz, and modern "artists" who have been so over-processed that any sense of realism isn't on the master in the first place! Not one decent transfer of single-miked blues, rock, live performances, etc.
Billy Idol "VH1 Storytellers", Days of the New first album, Blondie's recent "No Exit", Eric Clapton "Unplugged" -- any of these albums would probably be worth upgrading to SACD, if they were available.
As usual, it is the record companies themselves that are killing improved technology by their own blatant stupidity. How many years did it take for their boneheaded audio engineers to realize that pushing A/D converters to the point of clipping sounded like crap compared to the soft-clipping that occured with analogue media? Even now I still hear a lot of CDs that were obviously mastered by people who just don't understand the importance of avoiding digital clipping.
Most self-proclaimed "audiophiles" who buy crap like DiscWasher or played with the green magic markers are actually mid-fi enthusiasts who think that multi-channel amps are "audiophile" equipment. True audiophile equipment is just basic, clean designs manufactured in small quantities with extra high quality components.
To try to make a reasonable analogy, let me try comparing "beef products." Sony, Pioneer, JVC, are the McDonalds/Burger King of audio, with Carver, Polk, Infinity, et. al. being comparable to Wendy's (i.e. the latter is a little better, but still mass market fast food.) Real audiophile equipment from Conrad-Johnson, Sonic Frontiers, et. al. are prime steak at a 4-5 star restaurant in comparison, with a corresponding increase in price and service.
If all you want is "something to eat" (listen to), don't care about ambience (soundstage, depth, air, detail), or dump ketchup all over everything (failure to properly position speakers, seating, use quality cabling, etc) you're not going to be impressed with a $30 steak compared to a $5 burger.
I and many others do read articles before commenting. I happen to agree with many of Rick's philosophies, but not the way he has been presenting them over the past 4-5 years.
His fame and founding-father status with GNU mean that he gets a lot of respect and press, even from those who disagree with his ideals. But when he ignores whether a forum is appropriate and storms on with his idealistic rants, he provides the anti-open-source community with a poster boy to paint the community as a bunch of fanatics.
Some claim that his persistant ranting is a "challenge" for us all to follow our ideals and morals regardless of the consequences. While this is a terrific sentiment for those who wish to be seen as dedicated religious icons, it is hardly a suitable approach for those who don't live in their office as RMS does.
But in a world where "reality" TV is so popular, I guess I shouldn't be surprised. Realism is the last thing the sound-bite hungry populace seems to want, and a few "sure win" discussions like this ensure that the page counts stay high enough to improve the eyeball stats for advertising.
We will never know the answer to this puzzler because he is the only person in the world to get slashdot headlines by posting flamebait to the wrong news groups. Anyone else would either be ignored, flamed for cross-posting, or deleted by the moderators.
AmigaDOS was actually created as a masters thesis project in operating systems at a university in England, IIRC (it could have been a different country.)
It did not have Unix-style commands, APIs, or underpinnings. There were a lot of add-on programs created to give it shell-like functionality, and it supported ideas like process parentage and priorities, but no one who has ever done systems programming on a *nix system would confuse it with a *nix core.
The Amiga died due to Commodore's pathetic marketing. Period.
There is no "mostly" about it. Ultrix was BSD 4.1, released after we had been using BSD 4.2 on our VAX systems in University. Many of the manpages still had the BSD id strings, as did many header files. Running "strings" on libraries often reported BSD authorship as well. The final clincher was that all the bugs we had before upgrading to 4.2 returned with Ultrix (kernel panics, etc.)
Not really surprising, when you think about it. DEC was famous for providing "bug-for-bug" compatible VAX CPUs on the various models, because fixing the bugs would have broken too much code.
SVR4 added significant extensions to the data carried by signal handling, IPC, etc. when compared to SVR3. Useful concepts from the BSD family were used, as were a few from Xenix and SunOS.
The scheduling algorithms, memory management, configuration, and other features of SVR4 were sufficiently different that porting BSD code to the default APIs was a "challenging" project. Apps which used shared memory, semaphores, etc. required significant changes to run reliably.
The SVR4 environments I used (Sequent Dynix/PTX, Solaris, and another marginal player whose name I forget at the moment) each had BSD compatability libraries that would let you recompile your apps with minimal pain. I believe those compatability layers were largely vendor-specific rather than a generic feature of SVR4.
With Dynix/PTX, for example, you had a couple commands which let you switch between defaulting to BSD or SVR4 commands and libraries. You could blend the two environments in your code if you had to, but it was often tricky stuff.
Solaris did not use the same approach as Dynix/PTX, and actually seemed to make it harder to port old SunOS code than it did to recompile the same code under Dynix/PTX.
That unnamed OS had compatability libraries for recompiling code, but only supported SVR4 commands.
It would be a mistake to assume that the incorporation of good ideas from various *nix systems to be equivalent to merging them (which is what I read the "+" of the systems you mention to mean.) Remember that SunOS had a lot of proprietary code as well as the BSD base, and I seriously doubt too much of that was available for use in SVR4 (it could have happened, I just question the likelihood as that was the peak of the Unix incompatability wars.)
Although I've used HP-UX 9/10 and AIX (3.5/6?), I didn't work with them at the same low levels I did SunOS, Solaris, Dynix/PTX, "pure" SVR4, or Linux.
<OffTopicRamble>
How low-level was the work with Dynix/PTX?
By the time we were done with a two-year project, a 32 processor box was being bottlenecked by I/O capacity and the system bus, not the application code. The subsystem I wrote used X.25 data collection slaves, file writers similar to database redo logs, and shared memory under the coordination of a master/monitor process used to dynamically start/stop the processes. The first round of the functionality written using Tuxedo had proven to just be too slow, and had to be tweaked to the max.
That application almost beat out some work I did under VMS as far as maximizing system throughput. The work on VMS was a much simpler task, but I managed to get the I/O channels on three tape devices, five spindles of disk, the memory utilization, and CPU utilization on a VAX8600 (not positive about that model number) to all stay in the 95-98% load range...
</OffTopicRamble>
The old SunOS was based on BSD. Solaris is based on an SVR4 core, which is what made the transition so painful (different APIs for signal handling, etc.)
I stuck through the first episode, but I'll be shocked if this effort survives a month. Even the anime they ripped off (Cowboy Bebop, et. al.) had more depth of character in the first episode than Firefly.
I kept thinking that somewhere Roddenberry is spinning in shame that someone took the "wagontrain to the stars" phrase literally and tried to make a show about it.
A train robbery? The whore with a heart of gold? Six guns? Shit, why didn't they tie the damned ship to a tree as an anchor to finish off the list of lame cliches!
Fox killing off a solid show like Dark Angel for this drek, and SciFi trying to kill off Farscape just proves that there truly are no "entertainment" execs who have a clue. Then we've got John Doe which looks like a cheap last-minute knock-off of The Dead Zone (I'll give it a chance, but have zero hope of it being worthwhile.)
I wish Fox would take a clue from their cable sister "FX", which stuck by The Shield despite objections from advertisers and the religious groups. While it might not be everyone's idea of a great show, it's the only time in the past decade or so that I've seen a network support a show that deserved the support.
And these morons are worrying that people might pirate their shows with digital HDTV feeds???!!! *LOLOLOLOL* Don't worry, Mr. TV Executive -- with this kind of drek you really don't need to worry about piracy...
The original DirecTV stream saved by DirecTiVO is a modified 480x480 NTSC SVCD MPEG2 VBR stream, running at above-normal bitrates. I've heard there are utilities that will readjust the data stream to use standard MPEG2 VBR formats so that they can be dumped off for later viewing.
The problem with that is that the vast majority of set-top players can't handle that high an SVCD stream rate, so playback is an issue. If you copy the SVCD data to HDD for playback, the data rate isn't an issue, but you can't do that with a set-top player.
Some people have had luck with using MPEG editors to clip commercials from the data stream, tie the two MPEG segments together, and burning them off to DVD-R. Some set-top players can deal with that and will display a full NTSC screen with such a source, but it is not a standard DVD format so there are no guarantees.
Any direct data stream work assumes you're going to muck around with various kernel patches and stuff to disable the encryption of the data during the write to the HDD.
Personally I just settled for doing S-VHS wiring to an ASUS GF2MX card with video capture. Edit the HuffYUV files with VDub, and save as DiVX. Takes hours of CPU (PIII933), but it works. I find quality based 1-pass at 2.1 or 2-pass at 1500+ is indistinguishable from the original source, provided you've got enough CPU to turn some of the playback filtering on.
Eventually I got tired of wasting my time with all the editing and stuff, so the box is back to running SuSE 8.0. It was a very educational few months, though, and I learned a lot about video processing, filtering, and formats. I ended up with a few seasons worth of series archived, but movies are easier to just buy.
If finances permit, it's a lot cheaper to just buy season box sets, even at $80/season. Even at 15 minutes to edit out commercials, 20 episodes/season, you're still looking at about 4 hours of manual labour to capture and save a season. Add some more time for burning (say an hour), a reasonable pay scale in the tech industry, and it costs less to buy DVDs that use higher resolution and a much higher data rate.
The bottom of what? Your drive?
I believe that where the cooling is actually needed is on the chips that do the processing (at least with DirecTiVO.) The rather small cooler gets too hot to keep your finger on when the anti-pirate code cycles are happening.
I had been thinking in terms of either carving the case so a small fan would blow down directly on the heatsink involved, and/or replacing the puny heatsink with something larger (maybe even copper.)
One thing that is abundantly clear to me is that the so-called temperature monitoring isn't even as accurate as what is done with mobo based CPU temp monitoring.