"100dB SNR" is a number they toss out because it's something that the unit can benchmarket quite well. SNR means very, very little to good sound.
60dB is a reasonable power range for truly high-fidelity audio. That's a 1,000,000:1 power ratio - or a milliwatt to a kilowatt of power, to produce normal program levels. Very few amps produce a kilowatt of power, and those that do aren't real good at doing a milliwatt. And even if the amp can pull it off, most speakers other than high-quality horns will crap out (or blow out) long before then. So a *really good* system might do 40dB dynamic range (10,000:1 ratio) decently. Brutally clip or compress 10dB of peak, and lose 10dB of low-level detail. Your typical "home theatre" stereo crap is in the 20-30dB range, which basically means it is constantly clipping or recovering from clipping, and much of the low-level detail is lost.
The reason for the distinct "sound" of most audio equipment isn't SNR or THD or bandwidth, but rather how it recovers from overload conditions and how much low-level signal it loses. This stuff isn't easy to measure (it's much easier to analyze static signals than dynamic ones), so they produce the numbers that look good instead and claim that this somehow represents quality. When is the last time you saw THD measurements for a speaker? They don't do that, because 10% THD is really good behavior for a speaker, and that looks bad next to the.0001% THD $150 home theatre amp. Marketing is propaganda, and never forget that.
According to their FAQ, there is support for USB audio in the 2.4 kernel, but they haven't tried it themselves. I'd be interested in seeing if anyone has hacked it into working order. It'd be really nice for my laptop.
Ask yourself this... did Ukraine's refusal to agree to these standards come from principled opposition to the suppression of free speech by multinational corporations... or from internal pressure by Ukraine's homegrown media piracy industry?
I suspect the latter. Which sucks, because it's exactly what the corporate thieves are saying, and i'm not used to them telling the truth about anything.
Either way, ordinary Ukrainians lose, not to mention Americans.
...was the day i managed to actually crash it, while doing slightly dumb things with preferences. Up popped a dialog box informing me that Windowmaker had suffered a segmentation fault, and asking me if i would like it to restart itself. (and that is the ONLY time i ever managed to crash it, unlike, say, Enlightenment or Sawfish)
That is the most graceful crash handling i have ever encountered. Beautiful! I've been a loyal user ever since.
1. Stop auto-execution of content within Outlook. Ideally, make it impossible to execute content from a mail reader.
2. Stop designing operating systems where the default user account has write access to system binaries. Make it easy enough to do basic administration without formal administrator access that users don't run with administrator access by default (NT, W2K, XP desktop use).
3. Build bounds checking into Visual C++, at least as an option. Require programs under development to be tested with bounds checking on in order to detect buffer overflows.
I could go on, but you get the picture. No, you can't stop all security problems completely. However, you can make a very good dent in them. Just because a burglar can break your door down or pick the locks doesn't mean you shouldn't lock the doors to keep out the less skilled or ambitious.
Actually, ELF executables running under a normal user account CANNOT do the most interesting part, namely run their own SMTP server. Root access is required to open a low-numbered port.
Geez, don't people know at least the rudiments here?
Re:Is this slashdot or a Windows bug tracker?
on
Clever New Windows Worm
·
· Score: 5, Insightful
The XP exploit, at least, is an entirely new class of security hole, not seen before, and every last one of the 10M+ XP boxes shipped is vulnerable to total control from the outside.
If that ain't news, what is?
As for the worm... well, it's mildly technically interesting. But if Microsoft worms have become so common that they are no longer news... well, i think that's news, too!
Apache has a veto-proof majority of the web servers out there. Where are the Apache worms? Why is IIS, with far less market share, getting them? It's because Apache is secure and IIS is not, period.
Linux and OSX are both based on the Unix security model, a fundamentally sound design refined by two decades of real-world practice (dating back to the RTM worm in the early 1980s). It's not a matter of the virus writers aren't looking... it's a matter of a lack of exploitable holes. Name ONE Unix email client stupid enough to auto-execute code. Just one!
Yes, there are still exploitable holes here and there in Unix/Linux. But they generally require real mastery to find. Windows macro viruses can be written by 14 year old boys. My wife, a technical writer, doesn't know enough programming to write heapsort (do you?), but she knows enough to write a macro virus in VBA.
Get it through your head... the number of viruses and worms today is not a function of popularity or attention. It is a function of poor design and poor implementation, combined with security by obscurity (a technique discredited everywhere but Microsoft).
Really, learn about it. Don't just whine because Microsoft is getting a richly deserved spanking, and you don't want to hear how bad your favorite OS sucks.
It seems to me (as a guitarist, computer programmer, and amp builder) that part of the purpose, if not the MAIN purpose, of the guitar amp is to color the sound of the guitar in pleasing ways. So if tubes produce better colorations than "technically-superior digital solid state amps", then the tubes are technically superior, n'est pas?
The only thing "technically superior" about digital amps is that they are cheaper to manufacture.
And no, i won't be putting ethernet on my Gibson. Experience and simple physics dictates that the cord itself from the high-impedance guitar electronics to the amplifier input also colors the tone, and i'm not going to give up that coloration. Digitizing at 16bit/44.1khz "CD quality" commits absolute horrors on the subtleties of good tone (this can be mostly defeated with sufficient bandwidth, ie 24bit/96khz, but the Philips/Sony "Perfect Sound Forever" format is a crime against music).
Then again, my main guitar is an acoustic with no electronics at all, so i suppose it won't be needing ethernet.:}
Re:Why keep re-inventing SCSI?
on
Firewire and Linux?
·
· Score: 4, Informative
iirc, the 1394 spec started life as the SCSI-3 committee. In other words, FireWire is what SCSI was supposed to evolve into - including among other things, much cheaper chipsets and cabling.
Come see the violence inherent in the System!
on
Globalization
·
· Score: 3, Insightful
The battle isn't changing - only the battleground is.
The real fight is the ongoing friction between ever-larger units of society - the individual, the tribe, the nation, and now global society. Individuals chafe against the constraints of their own culture. Then as representatives of their own culture, they struggle against the crush of nationalism. Beyond that, the nations are fighting the coming globalism. This is not a fight that will ever be clearly resolved.
I think by nature humans are individualist and tribalist. However, the lines of those tribes are becoming more and more fluid. I belong to several tribes - SF fandom, Open Source programming, Unitarian Universalism, etc - that overlap some, but are really separate groups, each with their own struggle. As an Open Source advocate, i'm fighting against globalist corporatism on one level. As a Unitarian, i'm fighting against it on another. And against my own tribes, i'm fighting to protect my own identity.
Our tribes give us our connection to society. That connection is what gives us meaning and purpose, beyond mere survival. Nationalism and globalism simplify the survival question by improving our standard of living, but they don't give us much to feed our spirit. And both nationalism and globalism work to crush our tribes, which get in the way of convenient homogeneity.
As for the Middle East, look at what they're getting. They see the worst of globalism - Coca-Cola and Britney Spears - while getting nothing of the best of it, like freedom of speech and a growing economy. And we're crushing the strong and beautiful tribe of Arab and Islamic culture. No wonder they are fighting back! However, i don't think the medievalists like bin Laden can win in the long run, either, because they don't offer anything BUT tribalism.
There's a key... globalist culture provides huge economic incentives to participation, but you pay with your soul. It's great to have a Starbuck's everywhere so you can always get good coffee, but it sucks that Starbuck's is putting the funky individualistic cafes out of business. T-shirts are wiping out tribal dress because they're cheaper (unless you're a geek like me, where the t-shirt and its logo IS your tribal dress. I'm wearing a Klingon Kultural Ekchange shirt under my business casual).
Test-First Programming (TFP) is a key part of the Extreme Programming methodology. The JUnit unit testing library has been ported from Java to pretty much every widely used language. So the tools are there to produce robust code.
Here's how it works... BEFORE you write the body of a method or function, you write a unit test(s) for that function, to make sure it provides correct results for whatever inputs you might encounter. All of those tests should fail. THEN you write the body of the method/function. All the tests then should pass. If the tests don't pass, fix until they do. If bugs are encountered later that aren't caught by the unit tests, use test-first for the repairs - that way, you know your fix actually works. Just keep adding tests as you learn more.
Now put calling those unit tests into a framework and call it from your makefile. Unit test every time you compile.
Here are some of the benefits...
1. If new code breaks old code, the unit tests catch the error, and you can fix it appropriately right away.
2. You code with far greater confidence.
3. You keep your APIs very clean, because you have to test them right away.
4. Your APIs are thoroughly documented by the unit tests themselves.
5. Maintenance, especially by other programmers, is far easier, because they have the unit tests for reference and can easily narrow down where any bugs occur.
6. Refactoring is much easier, as any errors caused by refactoring are caught by the tests.
TRY THIS. It will change your whole approach to programming!
Okay, we know that MSN is excluding certain (often standards-compliant) browsers, ostensibly in the name of standards compliance, while allowing other browsers that are clearly not standards compliant, but have little market share. There are two possible explanations for this. The first is that Microsoft is being malicious and deliberately undermining competing browsers. The second is that the Microsoft programmers are too dumb to know the difference between "exclude all but..." and "include all but...".
These explanations are not contradictory. The odd MSN.com behavior may be an example of malicious anticompetitive behavior, very badly implemented.
Yes, Ballmer said Microsoft stock was overpriced. Why?
At that time, Microsoft made a significant portion of its earnings (like 10%) buying and selling Microsoft stock. When Ballmer made his announcement, the stock immediately dropped several points. Microsoft buys. A few days later, the market forgets, and its back where it was before. Microsoft sells, pocketing a hefty pile of money. Follow vaporware product announcements and other market-manipulating crap they do, and watch Microsoft make billions because they KNOW which direction their stock will go on any given day.
ISCABBS (telnet bbs.isca.uiowa.edu), one of the first Internet-based bbs systems, is over ten years old, still alive, and fairly healthy. At its peak in the mid-90s, it often had close to 2000 users online at any given time. It can still pull a few hundred when it's busy. It was the original DOC (Dave's Own Citadel) version of Citadel, and is fairly robust. The Citadel system of "X" messages (private messaging) and public forums is still the best online *community* form i've seen.
My wife works as a tech writer for a major airline. Although there were some minor layoffs before, she felt pretty secure in her job.
Until two weeks ago.
The airline industry had a disaster, and they need to lay off 20% of their staff in order to stay afloat. And they need to lay them off NOW. And you know what? That's okay. It'll suck for us if she loses her job (especially the benefits, since i'm a contractor), but we understand that the company is in serious trouble, through no fault of its own.
Luckily, i'm contracting at a well-funded nonprofit, so my position is secure, and our income is good.
FETs as followers only, with something else for voltage gain, *might* work. But i've never heard a FET amp that really does it for me. However, some people have had really good luck using FETs on the input of phono stages, cascoded with a triode, for gain. Supposedly, the FET is sonically invisible in this location. But that's a whole different game than power amps.
FETs also suffer from thermal overload problems, which i didn't go into earlier. When a transistor or FET is overloaded, the die overheats, and its gate capacitance changes. If the device is in a feedback loop to linearize it (as they nearly always are), changing the gate capacitance changes the time constants in the feedback loop, in a nonlinear way! And ALL amps clip ALL the time, unless you have 30-40dB of headroom. This is one of the reasons amps can blow up dealing with a speaker load at even normal volumes.
I'm not sure i believe in objective measures of sonic performance in amplifiers. Power, THD, and noise measurements can tell you something about it at least working correctly, but the amp/speaker/room combination is so sensitive and complex that i think the only real value call is your ears. Beyond that, it's largely a matter of your taste in music and hi-fi. The best overall system i've ever heard was optimized for small jazz groups, and couldn't play much of my own music well (3/4 of a watt can only do so much, even with super-efficient Lowther speakers).
What i can tell you is what measures are NOT really relevant... THD and watts. With even moderately efficient speakers, 8-10 watts from an amp that is optimized for good clipping behavior will more than fill a room. With more efficient speakers, only a watt or so is necessary. Think about it... if your speakers are 96dB sensitive, and you never listen more than 85dB in your living room, how much power do you really need? Enough for headroom, that's all. Now, you won't prevent it from clipping on peaks, unless you have a kilowatt of power into 100dB speakers (i know guys who use big PA rigs at home for just that reason), but if your amp clips gracefully, you're okay. --
An output signal closer to the input signal? First, what signals are you looking at? Sine waves (the preferred source of transistor amp manufacturers everywhere), or musical signals spread across a 10 octave bandwidth and 60dB dynamic range (that's a milliwatt to a kilowatt)? I suspect the former.
Second, what is the predominant shape of whatever inaccuracy you see at the end result? The human ear is perfectly willing to ignore broad, gross distortions, but can pick up nearly unmeasurable changes in phase (the curse of two-way and three-way speakers everywhere... phase distortion).
I'd argue that the problem isn't "warmth", but rather "coldness". "Warm" tube amps are generally sluggish and lack detail and deserve their criticism. But the best tube amps i've heard (all DIY homemade jobs) don't sound warm... they sound transparent, which is completely different. The cold sound of solid-state is the layer of electronic-sounding haze they spread over everything... intermodulation distortion, high-order harmonics, inaccurate error correction from relying on feedback for their reward (hence using sine waves rather than dynamic, wide-bandwidth signals for measurement... feedback loves the simple model), etc.
If you don't think tubes can make a real difference, you've either never heard a properly set up tube audio system, or you've simply ignored the evidence of your senses in favor of the pseudo-scientific snake oil of the mainstream audio world. Well-designed tube amplifiers with proper speakers sound far better than virtually any solid-state amplifiers. There are sound electrical engineering and psychoacoustic reasons for this, but they're lost in the simpleminded marketing bullshit of watts and THD specs, often to the point where listeners make up nonsense like "euphonic distortion" to explain why the tube system sounds so much better than their "audiophile" solid-state systems with gobs of power and low THD.
The basic problem is that the human ear is very good about filtering out signal-following distortions even at high levels (which is why a subwoofer with 10% distortion, like an excellent $2000 Velodyne, is even tolerable), but is extremely sensitive to distortions that don't modulate along with the signal. What we find in the electrical real world of the amp-speaker interface is a lot of mechanically stored energy messing with the signal. Say the amp spits out a big pulse of energy. The speaker stores MOST of that energy as mechanical crud, then spits it back into the amp as an electrical signal, maybe tens or hundreds of milliseconds later (speakers are maybe 1% efficient... where does the 99% go, except to mechanical and thermal dissipation? And a speaker motor also works as a generator for mechanically stored energy in the suspension. This is high school physics). That energy coming back from the speaker in a highly distorted, nonlinear way gets fed back to the beginning of the amplifier via the global negative feedback loop (required to make transistors even remotely linear) as *error correction*, and thus modulates totally unrelated signals occuring well after the original pulse.
And that's just the *beginning* of the problem. This is a very, very difficult problem, despite the simpleton math the marketing departments of the audio makers feed you.
Triode vacuum tubes are the ONLY amplifying devices linear enough to follow a voltage signal accurately without a feedback loop. The better tubes have distortion so low it is difficult to measure. And, as they approach their output limits, the distortion in a properly designed amplification stage is mostly second harmonic - so benign is is inaudible at less than 5% levels, and is perceived as an increase in loudness beyond that. A triode output stage can absorb reflected energy from the speaker via its own impedance, without feeding "error correction" back into previous stages. Thus, a triode with no negative feedback is MUCH better behaved into complex reactive loads with mechanical energy storage - i.e. speakers. The relatively high measured THD of such amps just shows the stupidity of THD measurements.
I could go on into numerous other shortcomings of transistors as amplifying devices for music, but this is a start.
Real Geeks build their own amplifiers from scratch, using the highest quality components for audio reproduction - triode vacuum tubes, the most linear amplifying devices ever made.
But for a cheap and easy system, pick up a pair of old Dynaco Stereo 70 amplifiers for the front and rear channels. Triode-wire them and remove the global feedback (ten minutes with a soldering iron). Add good mini-monitor speakers like the NHT Superzero, and a Paradigm powered sub (i'm assuming you're on a budget), and you're in a land of sound quality quite beyond the comprehension of the average home theatre geek for not too much money.
Hell, just get the Paradigm sub, a single ST-70 amp, and a pair of speakers, and run in stereo rather than Dolby Digital. You'd be surprised at just how good it can be, and that setup could be built for under $1000.
PostGreSQL only allows *one* database per server, and doesn't have good inter-server communication. When asked about this, the PostGreSQL developer community responds that we should all just put everything in one big database, rather than splitting our databases up along functional lines and doing cross-database queries when needed. This goes against the way those of us raised on Sybase and other commercial databases were trained - and against good data modelling, imho.
That is the #1 gripe we have against PostGreSQL, and one of the main reasons we're using MySQL instead. While feature-rich, the design of the engine interferes with what we understand as good data model architecture.
The gripes against MySQL are missing features - better locking, transactions, stored procedures, and triggers. We learn to cope with them. And PostGreSQL doesn't offer stored procs at all, and requires triggers to be written in C - ugly, ugly, ugly. It ain't that much better, and in some ways it's worse.
Slashdot Mirror
"100dB SNR" is a number they toss out because it's something that the unit can benchmarket quite well. SNR means very, very little to good sound.
.0001% THD $150 home theatre amp. Marketing is propaganda, and never forget that.
60dB is a reasonable power range for truly high-fidelity audio. That's a 1,000,000:1 power ratio - or a milliwatt to a kilowatt of power, to produce normal program levels. Very few amps produce a kilowatt of power, and those that do aren't real good at doing a milliwatt. And even if the amp can pull it off, most speakers other than high-quality horns will crap out (or blow out) long before then. So a *really good* system might do 40dB dynamic range (10,000:1 ratio) decently. Brutally clip or compress 10dB of peak, and lose 10dB of low-level detail. Your typical "home theatre" stereo crap is in the 20-30dB range, which basically means it is constantly clipping or recovering from clipping, and much of the low-level detail is lost.
The reason for the distinct "sound" of most audio equipment isn't SNR or THD or bandwidth, but rather how it recovers from overload conditions and how much low-level signal it loses. This stuff isn't easy to measure (it's much easier to analyze static signals than dynamic ones), so they produce the numbers that look good instead and claim that this somehow represents quality. When is the last time you saw THD measurements for a speaker? They don't do that, because 10% THD is really good behavior for a speaker, and that looks bad next to the
According to their FAQ, there is support for USB audio in the 2.4 kernel, but they haven't tried it themselves. I'd be interested in seeing if anyone has hacked it into working order. It'd be really nice for my laptop.
Ask yourself this... did Ukraine's refusal to agree to these standards come from principled opposition to the suppression of free speech by multinational corporations... or from internal pressure by Ukraine's homegrown media piracy industry?
I suspect the latter. Which sucks, because it's exactly what the corporate thieves are saying, and i'm not used to them telling the truth about anything.
Either way, ordinary Ukrainians lose, not to mention Americans.
...was the day i managed to actually crash it, while doing slightly dumb things with preferences. Up popped a dialog box informing me that Windowmaker had suffered a segmentation fault, and asking me if i would like it to restart itself. (and that is the ONLY time i ever managed to crash it, unlike, say, Enlightenment or Sawfish)
That is the most graceful crash handling i have ever encountered. Beautiful! I've been a loyal user ever since.
1. Stop auto-execution of content within Outlook. Ideally, make it impossible to execute content from a mail reader.
2. Stop designing operating systems where the default user account has write access to system binaries. Make it easy enough to do basic administration without formal administrator access that users don't run with administrator access by default (NT, W2K, XP desktop use).
3. Build bounds checking into Visual C++, at least as an option. Require programs under development to be tested with bounds checking on in order to detect buffer overflows.
I could go on, but you get the picture. No, you can't stop all security problems completely. However, you can make a very good dent in them. Just because a burglar can break your door down or pick the locks doesn't mean you shouldn't lock the doors to keep out the less skilled or ambitious.
Actually, ELF executables running under a normal user account CANNOT do the most interesting part, namely run their own SMTP server. Root access is required to open a low-numbered port.
Geez, don't people know at least the rudiments here?
The XP exploit, at least, is an entirely new class of security hole, not seen before, and every last one of the 10M+ XP boxes shipped is vulnerable to total control from the outside.
If that ain't news, what is?
As for the worm... well, it's mildly technically interesting. But if Microsoft worms have become so common that they are no longer news... well, i think that's news, too!
Apache has a veto-proof majority of the web servers out there. Where are the Apache worms? Why is IIS, with far less market share, getting them? It's because Apache is secure and IIS is not, period.
Linux and OSX are both based on the Unix security model, a fundamentally sound design refined by two decades of real-world practice (dating back to the RTM worm in the early 1980s). It's not a matter of the virus writers aren't looking... it's a matter of a lack of exploitable holes. Name ONE Unix email client stupid enough to auto-execute code. Just one!
Yes, there are still exploitable holes here and there in Unix/Linux. But they generally require real mastery to find. Windows macro viruses can be written by 14 year old boys. My wife, a technical writer, doesn't know enough programming to write heapsort (do you?), but she knows enough to write a macro virus in VBA.
Get it through your head... the number of viruses and worms today is not a function of popularity or attention. It is a function of poor design and poor implementation, combined with security by obscurity (a technique discredited everywhere but Microsoft).
Really, learn about it. Don't just whine because Microsoft is getting a richly deserved spanking, and you don't want to hear how bad your favorite OS sucks.
"no evidence of malicious code in the operating system has been reported".
:}
Never attribute to malice that which can be adequately explained by stupidity.
It seems to me (as a guitarist, computer programmer, and amp builder) that part of the purpose, if not the MAIN purpose, of the guitar amp is to color the sound of the guitar in pleasing ways. So if tubes produce better colorations than "technically-superior digital solid state amps", then the tubes are technically superior, n'est pas?
:}
The only thing "technically superior" about digital amps is that they are cheaper to manufacture.
And no, i won't be putting ethernet on my Gibson. Experience and simple physics dictates that the cord itself from the high-impedance guitar electronics to the amplifier input also colors the tone, and i'm not going to give up that coloration. Digitizing at 16bit/44.1khz "CD quality" commits absolute horrors on the subtleties of good tone (this can be mostly defeated with sufficient bandwidth, ie 24bit/96khz, but the Philips/Sony "Perfect Sound Forever" format is a crime against music).
Then again, my main guitar is an acoustic with no electronics at all, so i suppose it won't be needing ethernet.
iirc, the 1394 spec started life as the SCSI-3 committee. In other words, FireWire is what SCSI was supposed to evolve into - including among other things, much cheaper chipsets and cabling.
The battle isn't changing - only the battleground is.
The real fight is the ongoing friction between ever-larger units of society - the individual, the tribe, the nation, and now global society. Individuals chafe against the constraints of their own culture. Then as representatives of their own culture, they struggle against the crush of nationalism. Beyond that, the nations are fighting the coming globalism. This is not a fight that will ever be clearly resolved.
I think by nature humans are individualist and tribalist. However, the lines of those tribes are becoming more and more fluid. I belong to several tribes - SF fandom, Open Source programming, Unitarian Universalism, etc - that overlap some, but are really separate groups, each with their own struggle. As an Open Source advocate, i'm fighting against globalist corporatism on one level. As a Unitarian, i'm fighting against it on another. And against my own tribes, i'm fighting to protect my own identity.
Our tribes give us our connection to society. That connection is what gives us meaning and purpose, beyond mere survival. Nationalism and globalism simplify the survival question by improving our standard of living, but they don't give us much to feed our spirit. And both nationalism and globalism work to crush our tribes, which get in the way of convenient homogeneity.
As for the Middle East, look at what they're getting. They see the worst of globalism - Coca-Cola and Britney Spears - while getting nothing of the best of it, like freedom of speech and a growing economy. And we're crushing the strong and beautiful tribe of Arab and Islamic culture. No wonder they are fighting back! However, i don't think the medievalists like bin Laden can win in the long run, either, because they don't offer anything BUT tribalism.
There's a key... globalist culture provides huge economic incentives to participation, but you pay with your soul. It's great to have a Starbuck's everywhere so you can always get good coffee, but it sucks that Starbuck's is putting the funky individualistic cafes out of business. T-shirts are wiping out tribal dress because they're cheaper (unless you're a geek like me, where the t-shirt and its logo IS your tribal dress. I'm wearing a Klingon Kultural Ekchange shirt under my business casual).
I could go on. Does any of this make sense?
Test-First Programming (TFP) is a key part of the Extreme Programming methodology. The JUnit unit testing library has been ported from Java to pretty much every widely used language. So the tools are there to produce robust code.
Here's how it works... BEFORE you write the body of a method or function, you write a unit test(s) for that function, to make sure it provides correct results for whatever inputs you might encounter. All of those tests should fail. THEN you write the body of the method/function. All the tests then should pass. If the tests don't pass, fix until they do. If bugs are encountered later that aren't caught by the unit tests, use test-first for the repairs - that way, you know your fix actually works. Just keep adding tests as you learn more.
Now put calling those unit tests into a framework and call it from your makefile. Unit test every time you compile.
Here are some of the benefits...
1. If new code breaks old code, the unit tests catch the error, and you can fix it appropriately right away.
2. You code with far greater confidence.
3. You keep your APIs very clean, because you have to test them right away.
4. Your APIs are thoroughly documented by the unit tests themselves.
5. Maintenance, especially by other programmers, is far easier, because they have the unit tests for reference and can easily narrow down where any bugs occur.
6. Refactoring is much easier, as any errors caused by refactoring are caught by the tests.
TRY THIS. It will change your whole approach to programming!
Okay, we know that MSN is excluding certain (often standards-compliant) browsers, ostensibly in the name of standards compliance, while allowing other browsers that are clearly not standards compliant, but have little market share. There are two possible explanations for this. The first is that Microsoft is being malicious and deliberately undermining competing browsers. The second is that the Microsoft programmers are too dumb to know the difference between "exclude all but..." and "include all but...".
These explanations are not contradictory. The odd MSN.com behavior may be an example of malicious anticompetitive behavior, very badly implemented.
"There is more freedom in any moderately deStalinized dictatorship than there is in the American workplace."
-Bob Black, "The Abolition of Work"
Look it up on Google. It's instructive.
As an Internet discussion grows larger, the probability of a comparison involving terrorism or bin Laden approaches one.
i n's-Law.html)
(see http://www.tuxedo.org/~esr/jargon/html/entry/Godw
Sigh.
At that time, Microsoft made a significant portion of its earnings (like 10%) buying and selling Microsoft stock. When Ballmer made his announcement, the stock immediately dropped several points. Microsoft buys. A few days later, the market forgets, and its back where it was before. Microsoft sells, pocketing a hefty pile of money. Follow vaporware product announcements and other market-manipulating crap they do, and watch Microsoft make billions because they KNOW which direction their stock will go on any given day.
You were suckered, my friend.
Think God will listen to their case?
ISCABBS (telnet bbs.isca.uiowa.edu), one of the first Internet-based bbs systems, is over ten years old, still alive, and fairly healthy. At its peak in the mid-90s, it often had close to 2000 users online at any given time. It can still pull a few hundred when it's busy. It was the original DOC (Dave's Own Citadel) version of Citadel, and is fairly robust. The Citadel system of "X" messages (private messaging) and public forums is still the best online *community* form i've seen.
My wife works as a tech writer for a major airline. Although there were some minor layoffs before, she felt pretty secure in her job.
Until two weeks ago.
The airline industry had a disaster, and they need to lay off 20% of their staff in order to stay afloat. And they need to lay them off NOW. And you know what? That's okay. It'll suck for us if she loses her job (especially the benefits, since i'm a contractor), but we understand that the company is in serious trouble, through no fault of its own.
Luckily, i'm contracting at a well-funded nonprofit, so my position is secure, and our income is good.
FETs as followers only, with something else for voltage gain, *might* work. But i've never heard a FET amp that really does it for me. However, some people have had really good luck using FETs on the input of phono stages, cascoded with a triode, for gain. Supposedly, the FET is sonically invisible in this location. But that's a whole different game than power amps.
FETs also suffer from thermal overload problems, which i didn't go into earlier. When a transistor or FET is overloaded, the die overheats, and its gate capacitance changes. If the device is in a feedback loop to linearize it (as they nearly always are), changing the gate capacitance changes the time constants in the feedback loop, in a nonlinear way! And ALL amps clip ALL the time, unless you have 30-40dB of headroom. This is one of the reasons amps can blow up dealing with a speaker load at even normal volumes.
I'm not sure i believe in objective measures of sonic performance in amplifiers. Power, THD, and noise measurements can tell you something about it at least working correctly, but the amp/speaker/room combination is so sensitive and complex that i think the only real value call is your ears. Beyond that, it's largely a matter of your taste in music and hi-fi. The best overall system i've ever heard was optimized for small jazz groups, and couldn't play much of my own music well (3/4 of a watt can only do so much, even with super-efficient Lowther speakers).
What i can tell you is what measures are NOT really relevant... THD and watts. With even moderately efficient speakers, 8-10 watts from an amp that is optimized for good clipping behavior will more than fill a room. With more efficient speakers, only a watt or so is necessary. Think about it... if your speakers are 96dB sensitive, and you never listen more than 85dB in your living room, how much power do you really need? Enough for headroom, that's all. Now, you won't prevent it from clipping on peaks, unless you have a kilowatt of power into 100dB speakers (i know guys who use big PA rigs at home for just that reason), but if your amp clips gracefully, you're okay.
--
An output signal closer to the input signal? First, what signals are you looking at? Sine waves (the preferred source of transistor amp manufacturers everywhere), or musical signals spread across a 10 octave bandwidth and 60dB dynamic range (that's a milliwatt to a kilowatt)? I suspect the former.
Second, what is the predominant shape of whatever inaccuracy you see at the end result? The human ear is perfectly willing to ignore broad, gross distortions, but can pick up nearly unmeasurable changes in phase (the curse of two-way and three-way speakers everywhere... phase distortion).
I'd argue that the problem isn't "warmth", but rather "coldness". "Warm" tube amps are generally sluggish and lack detail and deserve their criticism. But the best tube amps i've heard (all DIY homemade jobs) don't sound warm... they sound transparent, which is completely different. The cold sound of solid-state is the layer of electronic-sounding haze they spread over everything... intermodulation distortion, high-order harmonics, inaccurate error correction from relying on feedback for their reward (hence using sine waves rather than dynamic, wide-bandwidth signals for measurement... feedback loves the simple model), etc.
--
If you don't think tubes can make a real difference, you've either never heard a properly set up tube audio system, or you've simply ignored the evidence of your senses in favor of the pseudo-scientific snake oil of the mainstream audio world. Well-designed tube amplifiers with proper speakers sound far better than virtually any solid-state amplifiers. There are sound electrical engineering and psychoacoustic reasons for this, but they're lost in the simpleminded marketing bullshit of watts and THD specs, often to the point where listeners make up nonsense like "euphonic distortion" to explain why the tube system sounds so much better than their "audiophile" solid-state systems with gobs of power and low THD.
The basic problem is that the human ear is very good about filtering out signal-following distortions even at high levels (which is why a subwoofer with 10% distortion, like an excellent $2000 Velodyne, is even tolerable), but is extremely sensitive to distortions that don't modulate along with the signal. What we find in the electrical real world of the amp-speaker interface is a lot of mechanically stored energy messing with the signal. Say the amp spits out a big pulse of energy. The speaker stores MOST of that energy as mechanical crud, then spits it back into the amp as an electrical signal, maybe tens or hundreds of milliseconds later (speakers are maybe 1% efficient... where does the 99% go, except to mechanical and thermal dissipation? And a speaker motor also works as a generator for mechanically stored energy in the suspension. This is high school physics). That energy coming back from the speaker in a highly distorted, nonlinear way gets fed back to the beginning of the amplifier via the global negative feedback loop (required to make transistors even remotely linear) as *error correction*, and thus modulates totally unrelated signals occuring well after the original pulse.
And that's just the *beginning* of the problem. This is a very, very difficult problem, despite the simpleton math the marketing departments of the audio makers feed you.
Triode vacuum tubes are the ONLY amplifying devices linear enough to follow a voltage signal accurately without a feedback loop. The better tubes have distortion so low it is difficult to measure. And, as they approach their output limits, the distortion in a properly designed amplification stage is mostly second harmonic - so benign is is inaudible at less than 5% levels, and is perceived as an increase in loudness beyond that. A triode output stage can absorb reflected energy from the speaker via its own impedance, without feeding "error correction" back into previous stages. Thus, a triode with no negative feedback is MUCH better behaved into complex reactive loads with mechanical energy storage - i.e. speakers. The relatively high measured THD of such amps just shows the stupidity of THD measurements.
I could go on into numerous other shortcomings of transistors as amplifying devices for music, but this is a start.
--
Real Geeks build their own amplifiers from scratch, using the highest quality components for audio reproduction - triode vacuum tubes, the most linear amplifying devices ever made.
But for a cheap and easy system, pick up a pair of old Dynaco Stereo 70 amplifiers for the front and rear channels. Triode-wire them and remove the global feedback (ten minutes with a soldering iron). Add good mini-monitor speakers like the NHT Superzero, and a Paradigm powered sub (i'm assuming you're on a budget), and you're in a land of sound quality quite beyond the comprehension of the average home theatre geek for not too much money.
Hell, just get the Paradigm sub, a single ST-70 amp, and a pair of speakers, and run in stereo rather than Dolby Digital. You'd be surprised at just how good it can be, and that setup could be built for under $1000.
--
PostGreSQL only allows *one* database per server, and doesn't have good inter-server communication. When asked about this, the PostGreSQL developer community responds that we should all just put everything in one big database, rather than splitting our databases up along functional lines and doing cross-database queries when needed. This goes against the way those of us raised on Sybase and other commercial databases were trained - and against good data modelling, imho.
That is the #1 gripe we have against PostGreSQL, and one of the main reasons we're using MySQL instead. While feature-rich, the design of the engine interferes with what we understand as good data model architecture.
The gripes against MySQL are missing features - better locking, transactions, stored procedures, and triggers. We learn to cope with them. And PostGreSQL doesn't offer stored procs at all, and requires triggers to be written in C - ugly, ugly, ugly. It ain't that much better, and in some ways it's worse.
--