Yeah, Verizon's coverage map is a joke. Just paint the whole country red and call it a day. Even their "detailed" maps just show their coverage in an all or nothing manner. It's either covered, or it's not, which I've found to be very misleading. T-Mobile's maps give much better detail and are usually pretty accurate.
I switched from T-Mobile to Verizon when I moved to Illinois to get their In Network calling. What I found was that I had less coverage (couldn't walk two feet inside my local grocery store without losing signal), more dropped calls, worse voice quality, and more expensive monthly bills. I gladly paid the 200 or so dollars to break my contract and go back to T-Mobile. After switching, I've had excellent coverage where I live (Chicago burbs).
Now that's not to say that Verizon isn't good for some people. I'm sure they have some very happy customers. But to categorically say that Verizon has better coverage than T-Mobile (or that any carrier has better coverage than another) is a crock. Yes, Verizon may have better coverage in rural areas. Yes, T-Mobile's network is heavily based around cities and major highways. But in the areas where I live work and travel, Verizon just didn't cut it. I'll readily admit that there are areas where T-Mobile's coverage is poor. My parents' house in Indiana is one of them. But in that area, Verizon is poor too and if you want good coverage, you have to go with Centennial (the regional 850MHz carrier there).
It really depends on how the carrier has implemented their network in your specific area. Where the towers are located, what frequency they're using, terrain, etc. My advice, when you find a carrier that's good in your area, stick to them.
Get a GSM mobile then. I use T-Mobile and find the quality to be just as good as landline under good signal conditions (granted, if you are on the fringe of the network, you are going to get some digital static).
GSM's EFR and AMR vocoders sound much better to me than CDMA's EVRC or SMV. They don't make the S and C sounds sound dulled and they tend to have a better tonal range. And well they should. The GSM codecs don't compress as much because after a certain point, there isn't much reason to for a GSM carrier (unlike a CDMA system where there are definite benefits to the carrier to compress voice as much as possible.) Also the CDMA vocoders have questionable background noise supression built in which, to me, just tends to make the user's sentences sounds choppy at the beginning and end of their phrases.
CDMA is great and all from a whizbang technology perspective but IMO, the benefit is really for the carrier and not the consumer. On a GSM system (which uses TDMA for the air interface), each user is allocated a definite amount of bandwidth that is locked to that user until the termination of the call. The caveat is that if there aren't open timeslots on a tower handoff, your call gets dropped. Also, if there aren't any open timeslots at all, you get a "network busy" error when trying to initiate a call.
CDMA systems conversely do soft handoff so there aren't as many dropped calls. (I rarely have dropped calls with GSM in Chicagoland though.) However, the caveat with CDMA is that everyone uses a pool of ether. Bandwidth can be allocated dynamically and as more users come on to the system, voice quality goes down as less and less ether is available to each user (read: the system's ability to accurately recover each user's signal is reduced). So the carrier gets a capacity gain (more users for less towers and spectrum) but the voice quality tends to suck in the process.
Well, unless you're paying for Premier or something like that, I think most "consumers" are pretty much willing to pay for either open source tools (free) or Windows Movie Maker (free).
I'd say for most consumer editing, the iLife Suite (free or $49 with iDVD) is a pretty compelling package. And power users can get their paws on Final Cut Express if they want to shell out some more bones for something a little more full featured.
Actually, it originally stood for N-Ten which was the codename for the Intel i860 RISC processor. Back in the day, Intel thought it was going to have a winner on it's hands. NT stuck but was changed to New Technology by Microsoft's marketing department.
Great, you've just posted some meaningless numbers that have no configuration or compiler information behind them. Let's go back and have the G5 vs. P4 Apple Benchmark redux.
Benchmarks are benchmarks. Take a look, get a general ballpark feel, and then make an intelligent decision based on more than just the benchmarks. You'd have to be crazy to make any purchasing decision on simply SPECint and SPECfp. (Though benchmarks like TCP tend to hold a little more water.)
To do something accurately, you would need to normalize most factors to test computer (read CPU) performance. You would need to normalize the compiler and all subsystems to the greatest amount possible. And as we learned in Apple's G5 vs. P4 brouhaha, this doesn't even satisfy. "Use ICC" the Wintel zealots said. "Normalize the compiler" the Mac zealots said. Either way, someone is never going to be happy. ICC will never be available for PowerPC and IBM XL C++ and Fortran will never be available for x86. The internals of GCC are different so no one is happy with that either. So where are we back to...square one...benchmarks like this are pretty useless.
The benchmarks posted in the parent tell us nothing. Who the hell knows what is really being tested or how we tested it.
In the PCWorld case, computer performance was not being tested. Application performance specific to a certain platform was. Granted, PCWorld shot themselves in the foot by picking some suspect applications. (Premier is a discontinued product that you can't get anymore so why test it? Any Mac zealot is going to be suspicious of a MS app test.) So why didn't they test some apps with some common code bases? (As a previous poster mentioned, maybe Mathematica.)
In any case, all this PCWorld article proves is, yeah, the AMD64 machines are pretty fast, and yeah, so are the PPC970 machines. However, it also proves that PCWorld shouldn't try to get into the benchmarking business any time soon.
-S
P.S. Look at how the Polystation (which beat almost every other computer in most benchmarks, came up last in the Quake benchmark. Is it due to the Opterons? Seems strange. Wish PCWorld would have explained that one.
LMAO. Open source doesn't pay enough for me to crap at night. I'm not going to trust the future of computing to OSS. Can you imagine if all IT workers suddenly didn't get paid. Wow, we'd have real innovation there. A bunch of unemployed workers churning out free code. I don't think anyone forced any open source developers to accept an offer with Apple now did they? Think before you post.
Slashdot Mirror
Yeah, Verizon's coverage map is a joke. Just paint the whole country red and call it a day. Even their "detailed" maps just show their coverage in an all or nothing manner. It's either covered, or it's not, which I've found to be very misleading. T-Mobile's maps give much better detail and are usually pretty accurate. I switched from T-Mobile to Verizon when I moved to Illinois to get their In Network calling. What I found was that I had less coverage (couldn't walk two feet inside my local grocery store without losing signal), more dropped calls, worse voice quality, and more expensive monthly bills. I gladly paid the 200 or so dollars to break my contract and go back to T-Mobile. After switching, I've had excellent coverage where I live (Chicago burbs). Now that's not to say that Verizon isn't good for some people. I'm sure they have some very happy customers. But to categorically say that Verizon has better coverage than T-Mobile (or that any carrier has better coverage than another) is a crock. Yes, Verizon may have better coverage in rural areas. Yes, T-Mobile's network is heavily based around cities and major highways. But in the areas where I live work and travel, Verizon just didn't cut it. I'll readily admit that there are areas where T-Mobile's coverage is poor. My parents' house in Indiana is one of them. But in that area, Verizon is poor too and if you want good coverage, you have to go with Centennial (the regional 850MHz carrier there). It really depends on how the carrier has implemented their network in your specific area. Where the towers are located, what frequency they're using, terrain, etc. My advice, when you find a carrier that's good in your area, stick to them.
Get a GSM mobile then. I use T-Mobile and find the quality to be just as good as landline under good signal conditions (granted, if you are on the fringe of the network, you are going to get some digital static).
GSM's EFR and AMR vocoders sound much better to me than CDMA's EVRC or SMV. They don't make the S and C sounds sound dulled and they tend to have a better tonal range. And well they should. The GSM codecs don't compress as much because after a certain point, there isn't much reason to for a GSM carrier (unlike a CDMA system where there are definite benefits to the carrier to compress voice as much as possible.) Also the CDMA vocoders have questionable background noise supression built in which, to me, just tends to make the user's sentences sounds choppy at the beginning and end of their phrases.
CDMA is great and all from a whizbang technology perspective but IMO, the benefit is really for the carrier and not the consumer. On a GSM system (which uses TDMA for the air interface), each user is allocated a definite amount of bandwidth that is locked to that user until the termination of the call. The caveat is that if there aren't open timeslots on a tower handoff, your call gets dropped. Also, if there aren't any open timeslots at all, you get a "network busy" error when trying to initiate a call.
CDMA systems conversely do soft handoff so there aren't as many dropped calls. (I rarely have dropped calls with GSM in Chicagoland though.) However, the caveat with CDMA is that everyone uses a pool of ether. Bandwidth can be allocated dynamically and as more users come on to the system, voice quality goes down as less and less ether is available to each user (read: the system's ability to accurately recover each user's signal is reduced). So the carrier gets a capacity gain (more users for less towers and spectrum) but the voice quality tends to suck in the process.
Well, unless you're paying for Premier or something like that, I think most "consumers" are pretty much willing to pay for either open source tools (free) or Windows Movie Maker (free).
I'd say for most consumer editing, the iLife Suite (free or $49 with iDVD) is a pretty compelling package. And power users can get their paws on Final Cut Express if they want to shell out some more bones for something a little more full featured.
Actually, it originally stood for N-Ten which was the codename for the Intel i860 RISC processor. Back in the day, Intel thought it was going to have a winner on it's hands. NT stuck but was changed to New Technology by Microsoft's marketing department.
Great, you've just posted some meaningless numbers that have no configuration or compiler information behind them. Let's go back and have the G5 vs. P4 Apple Benchmark redux.
Benchmarks are benchmarks. Take a look, get a general ballpark feel, and then make an intelligent decision based on more than just the benchmarks. You'd have to be crazy to make any purchasing decision on simply SPECint and SPECfp. (Though benchmarks like TCP tend to hold a little more water.)
To do something accurately, you would need to normalize most factors to test computer (read CPU) performance. You would need to normalize the compiler and all subsystems to the greatest amount possible. And as we learned in Apple's G5 vs. P4 brouhaha, this doesn't even satisfy. "Use ICC" the Wintel zealots said. "Normalize the compiler" the Mac zealots said. Either way, someone is never going to be happy. ICC will never be available for PowerPC and IBM XL C++ and Fortran will never be available for x86. The internals of GCC are different so no one is happy with that either. So where are we back to...square one...benchmarks like this are pretty useless.
The benchmarks posted in the parent tell us nothing. Who the hell knows what is really being tested or how we tested it.
In the PCWorld case, computer performance was not being tested. Application performance specific to a certain platform was. Granted, PCWorld shot themselves in the foot by picking some suspect applications. (Premier is a discontinued product that you can't get anymore so why test it? Any Mac zealot is going to be suspicious of a MS app test.) So why didn't they test some apps with some common code bases? (As a previous poster mentioned, maybe Mathematica.)
In any case, all this PCWorld article proves is, yeah, the AMD64 machines are pretty fast, and yeah, so are the PPC970 machines. However, it also proves that PCWorld shouldn't try to get into the benchmarking business any time soon.
-S
P.S. Look at how the Polystation (which beat almost every other computer in most benchmarks, came up last in the Quake benchmark. Is it due to the Opterons? Seems strange. Wish PCWorld would have explained that one.
LMAO. Open source doesn't pay enough for me to crap at night. I'm not going to trust the future of computing to OSS. Can you imagine if all IT workers suddenly didn't get paid. Wow, we'd have real innovation there. A bunch of unemployed workers churning out free code. I don't think anyone forced any open source developers to accept an offer with Apple now did they? Think before you post.