It is very hard to make a lossless codec for this kind of signal. The least significant bits look more or less like noise if you have 16 bit samples.
That's assuming scheme similar to run-length encoding isn't it? A lifetime ago (when I was 16) I actually wrote a small sound-compression app for Amiga, that was able to losslessly compress (8-bit) samples by 20-50% (on average), using a very simple delta-encoding (using varying number of bits representing deltas between consequtive samples, changing number of significant bits etc). Same could be easily adapted to 16-bit samples, but I'm sure current lossless codecs use other (or additional) methods. A while ago I toyed with the idea of trying something along the lines of the old method, but after writing piece of simple java code to see maximal compression (assuming no compression overhead) I decide dit wouldn't be worth it, without additional methods (probably combining cos transforms a la MP3 or such with compressing 'leftover' bits with something else)... and didn't really have time to get back to that project.:-)
Re:Ripping from the source a disadvantage? Huh?
on
AAC vs. OGG vs. MP3
·
· Score: 1
To answer your rhetorical question: No, I wouldn't want the least-filtered signal. I'd want the one the codec was intended for. And the common ones today are intended to compress 44.1KHz or 48Khz.
It really depends on your goals. For highest quality output (or at least compromise aiming more at quality than ultra-high compression), there's no denyning one would prefer least-filtered signal. That's just a basic fact; to minimize quantization errors start with signal that has least to begin with. Compression can only add to loss of (physical) quality, not recover.
For getting best compression it may be a different story... and I'd like to hear specific optimizations that are done based on couple of "well-known" sampiling frequencies that you imply there are? (honestly it'd be interesting, and I'm not claiming there can't be any... I'm just not familiar with any).
You may need to work on your analogy though... it's bit hard to see any relevant relationship with your example and signal processing in question.
Yes and no. For individual people, what matters is what they like the most. Not what's physically closest to original, but what's most enjoyable (regarding perceived quality). Usually there is reasonable correlation between the two, but there are exceptions.
For objective testing, sure, about the only method is to try to quantify difference to known reference point, original recording (which, in case of CD, has small distortion already in compared to ideal recording due to quantization).
What that means is that actually there are 2 basic ways to approach the problem; one based on physical properties only, and the other to try to measure psychoacoustic quality (ie. perceived quality). Not suprisingly, most development in compression methods does basic validation with measuring physical qualities, and later on optimize based on psychoacoustic preferences (ie. where is the loss of accuracy least significant for average listener etc).
Anyway, back to the original point; the problem really is just defining what is actually "closer to original", as no compressed sounds is ever identical. And to answer that, physics are not enough; human hearing and sound processing brains do is what defines it eventually, and (unfortunately) there is variation in there.
I think it was just misspelling. It was meant to be "scam", not "scandal". The whole rebate thing is stupid. Don't do mail-in rebate, don't do instant rebate. Lower the price and be done with it. They are trying to buy not only by contact info, but also my (fairly precious) time.
Now; if and when they really want the information, they should
do what grocery store chains do; offer preferred customer cards, which require personal info. That way contact info for advertising only needs to be given once, and it's easier and more reliable to profile purchasing behaviour as well.
However, many people in USA (and most people outside) would not include "right to own firearms without special permits" as gengeral civil rights.
It is usually thought to belong to the category of special privileges only granted based on proven need, same as for handling other dangerous things such as explosives (need permits, builders have them etc), some bacteria (researchers) and so forth.
I definitely agree with that. I have met people with degrees that are just clowns. Usually not that many in med. school (I hope?), but definitely too many in tech field. I think degree just usually sort of follows from being talented and interested; when you got the ingredients, you may as well go make the juice. Reverse is not true; degree in itself doesn't prove talent or motivation. There is correlation, but not all that strong one.
But like the other poster said, impeachment has to be based on some legality thing, not on actual judgements as far as they are properly processed, ie. process is not compromised. So what I meant was that congress can not just kick out judges if they outraged at decisions, as long as nothing illegal was done that lead to decisions.
Ditto for president, AFAIU. Last try was based on pointing out mostly procedural things (lying under oath); having extramarital relationship in itself would not have been enough, not even if it's with a person working for him. But lying as part of investigation was enough to try to impeach Clinton.
Right, that makes sense and is along the lines I was thinking. So that judges are not above law, but they are now at the mercy of politicians regarding decisions they make (but they may be if process of decision making is corrupt, like when bribery occurs).
Right. In general easiest summary of differences between left and right side (in USA at least) is that republicans are liberal in regards to economy (usually), and much less liberal, more control-oriented regarding individual rights (esp. "moral"(ity) things). Democrats are the opposite; prefer to regulate economical issues more, but try to avoid regulating other issues (don't try to outlaw different sexual orientation, or suppress freedom of/from religion, or illegalize abortion etc. etc. etc).
There are things between "code monkey" (or any other derogative term meaning "non-skilled worker") and "engineer". Not that I really care greatly if my profession is or is not consider to qualify as "engineering", but this black-and-white stereotypic imagery is just childish and immature.
Philosophically question is, in my opinion, related to the question of whether architects (I mean architects that design buildings) could be considered engineers? Should they? I find programmers to be more similar to architects and skilled craftsmen in many ways than engineers, but there are lots of similarities to engineers too.
The very best software programmers that I have personally met were all Electrical Engineers
I have met good programmers from various undergrad fields (usually CS, EE and physics), but not one (so far?) good non-CS-degree people who hadn't either gotten masters in CS, or spent considerable time studying CS courses (often as part of their main degree). Not that those can't exist, just haven't met one yet.
Anyway I guess this is obvious, but most of all I've noticed that good programmers (software implementors / developers, if you prefer) really need is enough talent, and plenty of interest in the field; both of which usually lead to high motivation. Degree they have usually depends more on when exactly they figured out they are good at creating software. Those that learnt that early tend to be CS; those who "came out of close" later, EE or physics.:-)
And don't forget that appointments to all federal courts have to be cleared by a democratically elected Congress, which also has the power to remove them from their bench.
Really? I was under impression that this is exactly what Congress can not do, and for good reason. If it could, they would not be independent by any measure, but just dependant on Congress' good will? Which would be against "checks and balances" philosophy, not to mention more generally accepted principle of dividing legislative, executive and 'whatever - the -third of -Rousseaus- powers was' powers?
However, as a copy-protected cd is technically NOT compliant with the original philips specifications, I find it very suspect that the RIAA made thier own standard. Especially since this standard serves no purpose other than to replace the ageing 'Compact Disc' logo.
I wonder if CD trademark holders (Philips) could try to prevent them from using this misleading new 'standard'. Especially if they call it "Enhanced Compact Disc", without having much to do with "Compact Disc" it seems like this could be a trademark violation? (they probably couldn't do the same just for acronym CD). Especially since it's clearly aimed to give the impression they are "CDs but better".
In reality, I would think RIAA et al have done their homework and have either gotten OK from parties that could take legal action, or make sure to carefully avoid legal problems.
Designing software is an art. Programming is not an art, it is a craft.
Oh, I take exception with this!
Yup, me too. There is no such thing as programming without design; and the opposite is meaningless (ie. design without implementation is possible, but useless)
But more to the point; I don't really consider combination of design and implementation to be arts, as much as craft(manship). And that's not derogatory, or bad; I don't consider great masters that make violins artists, nor samurai sword makers. I consider them highly skilled craftsmen.
Above is not to say there isn't anything artistic in the process, nor that end result can not (or should not) be aesthetically pleasing. Far from it. But whereas art does not even try to create anything of practical value (but of aesthetic pleasure), these crafts do.
Note that I wasn't commenting on "die" part, just the part about firing half of employees. "To live or die" is anther, longer term open question that one can certainly argue either way.
In fact, my biggest concern about company's viability is that all the weasely stock analysts are praising Sun's "new strategies" and "new visions", and ignoring longer term strategic issues as well as actual products, product lines and plans (Sun trying to become a software company instead of hw etc. etc). Not a big shock, most analysts wouldn't recognized a clue if they saw one, but still. So, analysts claiming "it'll be just hunky dory now that these great men have spectacular big vision" is what worries me. When they praise companies, prepare to leave; when they criticize a company it's either "run as fast as you can it's all over next week" (Enron) or "business as usual, analysts without a clue". The trick is to determine which one it is.:-)
Best case scenario is if analysts just don't have much to say!
they will announce to fire half of their employees within a year.
That's absurd. Company has 3 billion dollars in cash, is in break-even situation (nominal profits from last quarter, excluding writeoffs), and is investing heavily in R&D. Why on earth would they fire half the employees?
Furthermore, unilateral my ass! America + Great Britain + 40 other countries != unilateral. I could go on and criticize you Europeans as a whole, but I won't stoop down to your level.
I agree in that criticizing americans as whole is unfair, but claiming that "coalition" is much more than just US is ignoring the facts. Only UK is sending significant amount of troops (Australia some), and rest are either good-weather friends that think it's beneficial (ie. they need US favours), or have right-wing leaders that are just pushing their ideologic solutions in direct opposition of majority opinion (as is the case for, say, Spain).
Calling it "master and 41 puppets" may be bit exaggerating, but not much more so than calling it a real global coalition (heh, "Ali Baba and 41 bandits" would be an alternative?)
Ok, VNC is good, and probably does similar job to SunRay (servers). Good point. But original poster seemed to indicate normal X-servers could do the same without any additions?
Please do not assume that "does not work for me" (as a developer, hobbyist or free thinker) equals "does not work for anyone". Like others have pointed out, SunRay concept is fairly nice example of how it can work out. I don't know extensive list of corporations (outside Sun) that have big number of SunRays, but I do know that Sun itself is using them heavily, and with good results.
The aim of the whole thin client concept is NOT individual end user, it's the business world; especially non-tech businesses. They do not care whether they have full-featured work station or just "dumb" simple terminal that is enough for browsing the web, sending/receiving emails and doing office document work. And thin clients can be competitively priced... but more importantly, very very easily (and economically) administered.
I'm honestly trying to think of what advantages Sun could bring to a desktop, and I can't think of any.
Perhaps you were only thinking about technical aspects? What Sun can bring to table is the good old "warm fuzzy feeling" for companies, just like what IBM is doing with Linux. That is, sense of trust and safety, that there is a big company that stands behind the product. I'm the first one to agree that that's usually a fallacy, but for many (esp. higher ranking) decision makers, having enterprise-level support with contracts is tempting.
That is, even if Sun can not bring much above and beyond what other Open Source integrators can, they can add their brand, their corporate weight and visibility. And that is another combination of price, standards-compliancy and support. Not everyone is just looking for cheapest thing, nor most fun to use. Many bigger corps have pretty different goals.
As to writing software, I don't see why Sun should try to rewrite Mozilla, or why OpenOffice couldn't become (if it's not already) worthy competitor to MS Office. Those are dead end paths in terms of trying to create something clearly superior. In many ways, browsers and office packages have 'only' evolution left, not revolution (you could come up with new ways of browsing net, or creating documents, but those would belong to new genres of apps most likely).
Right. The extra features over and beyond 'standard' diskless x terminals mostly reside on server side (seamless session migration). Combination of SunRays and SunRay servers thus are somewhat extended/enhanced version of stock X-server / X-term combo.
Anyhow, SunRays are about the only existing somewhat succesful implementations of... what was it called by Larry Ellison et al... network computers? Anyway, thin client, fat server idea that was supposed to revolutionaze computing landscape 5 years ago or so.
Hmmh. Pardon my ignorance here, but are you talking about something above and beyond just normal X-windows server and desktop like Gnome/KDE? I do know how extensively SunRays work (completely transparently for apps, it's complete snapshot as far as user is concerned), but I haven't seen anything close to that for Linux yet. Desktops generally remember approximately which apps you had running, and in some cases try to get back to about where you were (assuming apps play nice and help). But that's not really much of a session, it's more like having configurable starting point; reduces start up time, not having to open number of xterms (et al) you usually do.
Nice thing usually (and pain in other times when you forgot to uncheck that 'restore desktop' and get n+1 unnecessary apps).
As to smart card, smart card's whole idea is that you do not even have to logout or login (that is, inserting/removing card constitutes login/logout action). So, it's what connects you to your session. Nothing extraordinary, but added convenience.
Interesting... and while I think it's unfortunate that promising new employees (I assume you were one) don't always get a chance to get to the "upper caste" (not because of their lack of skills but because of bad timing, bit of bad luck, not enough time to show their skills and learning capabilities), it's probably reasonably efficient way to retain talent. As to my comment, I was thinking of "bad companies" during bad times. I think company you worked for wasn't quite like that, and that's why they tried to make their top talent to feel safe, exactly to prevent them from looking elsewhere "just in case".
I have also noticed that it takes between year and year and a half as a developer to really get recognized for my skills (experience from 2 of my last employers). YMMV as usual, but during these times that feels like a fairly long time. After that happens, though, things do get more interesting, and there's some more job security I think. Nothing like bullet-proof, but perhaps moving to a new tier in a way. So I guess my current employer may have somewhat similar practices.
Unix: No, not really. There is/was IE 5.5 for Solaris, but it's so badly outdated (and didn't include all or even most of Windows version's features) that it's more a curiosity (or, "hack") than a viable version.
Or have they come up with actual up-to-date version on a unix platform?
And actually it's not just "little help", motivated people get much more done, offer good ideas they have, participate, help retain talent and so on. I have worked on both motivated and de-motivated teams (fortunately, currently on a team that's mostly motivated), and I know the difference from employee side.
On the flip side, de-motivated people are there just to collect their paycheck and sporadically get work done, either to avoid getting fired, or if they would otherwise get too bored.
Whip-n-row techniques also most likely increase attrition rates a lot, so the only case where they make sense to use is where one wants to get headcount down fast, independent of consequences. In really short-term you may get something done day or two earlier, but cause burn-out for some, and get end product that has lower quality than one produced under more sane leadership.
Slashdot Mirror
That's assuming scheme similar to run-length encoding isn't it? A lifetime ago (when I was 16) I actually wrote a small sound-compression app for Amiga, that was able to losslessly compress (8-bit) samples by 20-50% (on average), using a very simple delta-encoding (using varying number of bits representing deltas between consequtive samples, changing number of significant bits etc). Same could be easily adapted to 16-bit samples, but I'm sure current lossless codecs use other (or additional) methods. A while ago I toyed with the idea of trying something along the lines of the old method, but after writing piece of simple java code to see maximal compression (assuming no compression overhead) I decide dit wouldn't be worth it, without additional methods (probably combining cos transforms a la MP3 or such with compressing 'leftover' bits with something else)... and didn't really have time to get back to that project. :-)
It really depends on your goals. For highest quality output (or at least compromise aiming more at quality than ultra-high compression), there's no denyning one would prefer least-filtered signal. That's just a basic fact; to minimize quantization errors start with signal that has least to begin with. Compression can only add to loss of (physical) quality, not recover.
For getting best compression it may be a different story... and I'd like to hear specific optimizations that are done based on couple of "well-known" sampiling frequencies that you imply there are? (honestly it'd be interesting, and I'm not claiming there can't be any... I'm just not familiar with any).
You may need to work on your analogy though... it's bit hard to see any relevant relationship with your example and signal processing in question.
For objective testing, sure, about the only method is to try to quantify difference to known reference point, original recording (which, in case of CD, has small distortion already in compared to ideal recording due to quantization).
What that means is that actually there are 2 basic ways to approach the problem; one based on physical properties only, and the other to try to measure psychoacoustic quality (ie. perceived quality). Not suprisingly, most development in compression methods does basic validation with measuring physical qualities, and later on optimize based on psychoacoustic preferences (ie. where is the loss of accuracy least significant for average listener etc).
Anyway, back to the original point; the problem really is just defining what is actually "closer to original", as no compressed sounds is ever identical. And to answer that, physics are not enough; human hearing and sound processing brains do is what defines it eventually, and (unfortunately) there is variation in there.
I think it was just misspelling. It was meant to be "scam", not "scandal". The whole rebate thing is stupid. Don't do mail-in rebate, don't do instant rebate. Lower the price and be done with it. They are trying to buy not only by contact info, but also my (fairly precious) time.
Now; if and when they really want the information, they should do what grocery store chains do; offer preferred customer cards, which require personal info. That way contact info for advertising only needs to be given once, and it's easier and more reliable to profile purchasing behaviour as well.
However, many people in USA (and most people outside) would not include "right to own firearms without special permits" as gengeral civil rights. It is usually thought to belong to the category of special privileges only granted based on proven need, same as for handling other dangerous things such as explosives (need permits, builders have them etc), some bacteria (researchers) and so forth.
I definitely agree with that. I have met people with degrees that are just clowns. Usually not that many in med. school (I hope?), but definitely too many in tech field. I think degree just usually sort of follows from being talented and interested; when you got the ingredients, you may as well go make the juice. Reverse is not true; degree in itself doesn't prove talent or motivation. There is correlation, but not all that strong one.
Ditto for president, AFAIU. Last try was based on pointing out mostly procedural things (lying under oath); having extramarital relationship in itself would not have been enough, not even if it's with a person working for him. But lying as part of investigation was enough to try to impeach Clinton.
Right, that makes sense and is along the lines I was thinking. So that judges are not above law, but they are now at the mercy of politicians regarding decisions they make (but they may be if process of decision making is corrupt, like when bribery occurs).
Right. In general easiest summary of differences between left and right side (in USA at least) is that republicans are liberal in regards to economy (usually), and much less liberal, more control-oriented regarding individual rights (esp. "moral"(ity) things). Democrats are the opposite; prefer to regulate economical issues more, but try to avoid regulating other issues (don't try to outlaw different sexual orientation, or suppress freedom of/from religion, or illegalize abortion etc. etc. etc).
There are things between "code monkey" (or any other derogative term meaning "non-skilled worker") and "engineer". Not that I really care greatly if my profession is or is not consider to qualify as "engineering", but this black-and-white stereotypic imagery is just childish and immature.
Philosophically question is, in my opinion, related to the question of whether architects (I mean architects that design buildings) could be considered engineers? Should they? I find programmers to be more similar to architects and skilled craftsmen in many ways than engineers, but there are lots of similarities to engineers too.
I have met good programmers from various undergrad fields (usually CS, EE and physics), but not one (so far?) good non-CS-degree people who hadn't either gotten masters in CS, or spent considerable time studying CS courses (often as part of their main degree). Not that those can't exist, just haven't met one yet.
Anyway I guess this is obvious, but most of all I've noticed that good programmers (software implementors / developers, if you prefer) really need is enough talent, and plenty of interest in the field; both of which usually lead to high motivation. Degree they have usually depends more on when exactly they figured out they are good at creating software. Those that learnt that early tend to be CS; those who "came out of close" later, EE or physics. :-)
Really? I was under impression that this is exactly what Congress can not do, and for good reason. If it could, they would not be independent by any measure, but just dependant on Congress' good will? Which would be against "checks and balances" philosophy, not to mention more generally accepted principle of dividing legislative, executive and 'whatever - the -third of -Rousseaus- powers was' powers?
I wonder if CD trademark holders (Philips) could try to prevent them from using this misleading new 'standard'. Especially if they call it "Enhanced Compact Disc", without having much to do with "Compact Disc" it seems like this could be a trademark violation? (they probably couldn't do the same just for acronym CD). Especially since it's clearly aimed to give the impression they are "CDs but better".
In reality, I would think RIAA et al have done their homework and have either gotten OK from parties that could take legal action, or make sure to carefully avoid legal problems.
-
Designing software is an art. Programming is not an art, it is a craft.
Oh, I take exception with this!Yup, me too. There is no such thing as programming without design; and the opposite is meaningless (ie. design without implementation is possible, but useless)
But more to the point; I don't really consider combination of design and implementation to be arts, as much as craft(manship). And that's not derogatory, or bad; I don't consider great masters that make violins artists, nor samurai sword makers. I consider them highly skilled craftsmen.
Above is not to say there isn't anything artistic in the process, nor that end result can not (or should not) be aesthetically pleasing. Far from it. But whereas art does not even try to create anything of practical value (but of aesthetic pleasure), these crafts do.
In fact, my biggest concern about company's viability is that all the weasely stock analysts are praising Sun's "new strategies" and "new visions", and ignoring longer term strategic issues as well as actual products, product lines and plans (Sun trying to become a software company instead of hw etc. etc). Not a big shock, most analysts wouldn't recognized a clue if they saw one, but still. So, analysts claiming "it'll be just hunky dory now that these great men have spectacular big vision" is what worries me. When they praise companies, prepare to leave; when they criticize a company it's either "run as fast as you can it's all over next week" (Enron) or "business as usual, analysts without a clue". The trick is to determine which one it is. :-)
Best case scenario is if analysts just don't have much to say!
That's absurd. Company has 3 billion dollars in cash, is in break-even situation (nominal profits from last quarter, excluding writeoffs), and is investing heavily in R&D. Why on earth would they fire half the employees?
I agree in that criticizing americans as whole is unfair, but claiming that "coalition" is much more than just US is ignoring the facts. Only UK is sending significant amount of troops (Australia some), and rest are either good-weather friends that think it's beneficial (ie. they need US favours), or have right-wing leaders that are just pushing their ideologic solutions in direct opposition of majority opinion (as is the case for, say, Spain).
Calling it "master and 41 puppets" may be bit exaggerating, but not much more so than calling it a real global coalition (heh, "Ali Baba and 41 bandits" would be an alternative?)
Ok, VNC is good, and probably does similar job to SunRay (servers). Good point. But original poster seemed to indicate normal X-servers could do the same without any additions?
The aim of the whole thin client concept is NOT individual end user, it's the business world; especially non-tech businesses. They do not care whether they have full-featured work station or just "dumb" simple terminal that is enough for browsing the web, sending/receiving emails and doing office document work. And thin clients can be competitively priced... but more importantly, very very easily (and economically) administered.
Perhaps you were only thinking about technical aspects? What Sun can bring to table is the good old "warm fuzzy feeling" for companies, just like what IBM is doing with Linux. That is, sense of trust and safety, that there is a big company that stands behind the product. I'm the first one to agree that that's usually a fallacy, but for many (esp. higher ranking) decision makers, having enterprise-level support with contracts is tempting.
That is, even if Sun can not bring much above and beyond what other Open Source integrators can, they can add their brand, their corporate weight and visibility. And that is another combination of price, standards-compliancy and support. Not everyone is just looking for cheapest thing, nor most fun to use. Many bigger corps have pretty different goals.
As to writing software, I don't see why Sun should try to rewrite Mozilla, or why OpenOffice couldn't become (if it's not already) worthy competitor to MS Office. Those are dead end paths in terms of trying to create something clearly superior. In many ways, browsers and office packages have 'only' evolution left, not revolution (you could come up with new ways of browsing net, or creating documents, but those would belong to new genres of apps most likely).
Anyhow, SunRays are about the only existing somewhat succesful implementations of... what was it called by Larry Ellison et al... network computers? Anyway, thin client, fat server idea that was supposed to revolutionaze computing landscape 5 years ago or so.
As to smart card, smart card's whole idea is that you do not even have to logout or login (that is, inserting/removing card constitutes login/logout action). So, it's what connects you to your session. Nothing extraordinary, but added convenience.
Am I missing something here?
I have also noticed that it takes between year and year and a half as a developer to really get recognized for my skills (experience from 2 of my last employers). YMMV as usual, but during these times that feels like a fairly long time. After that happens, though, things do get more interesting, and there's some more job security I think. Nothing like bullet-proof, but perhaps moving to a new tier in a way. So I guess my current employer may have somewhat similar practices.
Or have they come up with actual up-to-date version on a unix platform?
On the flip side, de-motivated people are there just to collect their paycheck and sporadically get work done, either to avoid getting fired, or if they would otherwise get too bored.
Whip-n-row techniques also most likely increase attrition rates a lot, so the only case where they make sense to use is where one wants to get headcount down fast, independent of consequences. In really short-term you may get something done day or two earlier, but cause burn-out for some, and get end product that has lower quality than one produced under more sane leadership.