For a backup of $HOME? Not so much depends how you value your time and physical space i guess (and of course on how much data you have in/home).
my guess though is that dual layer DVDs have remained expensive because they have limited use (other than copying dvd based games). Most movies can be recompressed to fit on a single layer while keeping the quality loss to a level most people won't notice. For backing up a media collection or similar they aren't really big enough (imo the only people who are going to try and back up a large hard drive to DVD are those who already have too much time on thier hands and so are going to go for the cheaper per gigabyte single layer) and for backing up really important data they are generally bigger than nessacery.
is compliance with EMC standards, running high speed interfaces that are designed for internal use externally will almost certainly make the system non compliant (the same applies running a PC with the lid off and to many windowed cases)
does this matter? it depends! It doesn't if its for your own use and you don't have any sensitive gear arround or if you are a small fly by night firm or are just selling parts (generally standards are applied to the end product as a whole not individual parts) but any big vendors would not want to sell non compliant systems.
remember that metal case (even plastic cases have metal linings) on your PC isn't to keep RFI out, its to keep it in!
imo XP sp2 was really a new version of windows but they made it a service pack because the main area of improvement (security/malware resistance) was not something that most users would pay for but was something the lack of was getting them a lot of bad press for.
Folks are smart. If something is an improvement, adoption happens. folks are dumb, if something takes extra time in the short term then they are unlikely to adopt it regardless of its long term advantages.
most programmers (and i know i'm guilty of this myself) work on the assumption that fixes are easy and so quality doesn't matter much. This is true up to a point but trouble is people don't count the costs the bugs cause in the time between initially becoming apparent and being fixed. Nor the cost of tracing a bug that happens apparently randomly in production use or only on servers that run for months or whatever. How much does a MS security hotfix cost MS and thier customers? I don't have figures (probablly noone has figures for what it costs thier customers in agregate) but I bet its a VERY large sum of money.
In other engineering disciplines the costs are more obvious and the methodologies better developed.
Umm, Dapper isn't even the stable version, is it? edgy is not a long term support release, support for it will be discontinued in a relatively short timeframe and its reported to be ubuntus buggiest release to date.
You speak as though all the support options rested on the shoulders of Canonical, but that is simply not true do any of those companies have the rescources and inclination to do thier own tracking on what security issues pertain to what ubuntu versions and backport those security fixes themselves if ubuntus long term support promises turn out to be hollow?
it effectively exists on the whim of one rich guy who could lose interest and kill it whenever he wanted.
apparently the creaks are already showing in dapper (one person above mentioned downloading an update and his gui suddenly breaking with no availible soloution) and its less than a third of the way through its supposed support cycle. Support for the less major software in universe is even worse (much of the stuff in there has simply been imported straight from debian with no testing if it actually works in the ubuntu environment).
iirc long term support releases are supposed to come every two years and are finally obsoleted completely after 5 years. Maybe i'll be convinced when they have 3 of them in support at once (two for the desktop, 3 for the server) and are doing a good job for all of them i'll be convinced, until then ubuntu must be regarded as a young and vulnerable distro.
they could do that and they could also use the nasty flaw in IE that allows owners of versign certs to MITM ssl connections made by it provided you own a verisign issued certificate.
but people who share thier connections with broadband routers who plug in a PC with firefox already installed will not have any reason to run the ISPs software and hence will notice this and there isn't really anything they can do about it.
and it would only require one sufficiantly tech savy user using a browser that didn't have thier root cert and wasn't vulnerable to the exploit to spot what they were doing and out them and it could mean very bad press especially if the banks found out and started telling thier customers not to use said isp.
that sounds like a risky strategy, not having the software before the hardware is frozen could leave you easilly running out of some hardware rescource that you didn't realise the need for before you wrote the software.
Remember 1995? Dial up modems? 1 megapixel cameras? Back when file size mattered? yes speeds are increasing but here in the uk the bottom end of "broadband" is only a few times faster than the fastest dialup which in turn is less than twice as fast as the modems that were current in that era.
many people are still using dialup either because they are light users, simply can't get broadband (at prices that are acceptable to normal customers, i am aware of the existance of leased lines and satalite links. Satalite links also have high latency which throws away much if not all of the percieved speed gain when browsing) or are on the move and can't rely on being able to connect to a network at the places they visit. Mobile data is also generally similar speeds to dialup and often has high per megabyte costs as well.
and consumer cameras are now at 5 megapixel or so
and hosting providers still charge significant prices per gigabyte so there is a definite incentive for webauthors to keep filesize down.
So what has changed in the last decade (here in the uk at least). 1: average net speeds for heavy users have increased significantly to the point where a web page loads instantly. 2: consumers have an increased taste for larger media, audio and video have become much more common and image sizes have grown with the capabilties of cameras and scanner handling software (handling an image of more than a megapixel or so on a 486 with 8 megs of ram is pretty impractical) 2: the gap between users has also increased a lot and is much harder for the user to close, in the old days all you needed to increase your speed was a one off payment for a better modem but to get beyond the 56K barrier you have to take out a subscription for every location you wan't that speed out on top of the normal phone line cost, and the time for getting it at a new site may be as high as a fortnight. Youd better check your site works ok over a crappy modem link because that is how it may have to be acccessed in a pinch and you don't want to lose a good customer because of that. (remember while the poor are on dialup so are those who are fairly rich but stuck in an awkward postition) 3: lots more people are on the net overall 4: p2p has come in and has been retailiated against with massive traffic limiting and dishonesty on the part of ISPs. Speeds on consumer internet connections have become very unpredictable and nowhere near what is being advertised.
Also, bzip2 is good with large amounts of information. And surprisingly PNG isn't the case: not the whole image is compressed but group of adjacent lines. That's done to allow network usage: user can see parts of image before it is completely downloaded. totally incorrect, a png contains one long compressed stream of data (this may be split between multiple chunks to allow iterative generation but thats a minor overhead). Remember gzip is a stream compressor so you can feed it more input and get more output while keeping the compressor state intact.
What, are they going to go on a 4-day bender after the DST upgrades? no they will be cleaning up after all the breakage and fixing things that are still using the old rules and therefore causing problems.
that depends: on a traditional software ramdrive it is as you say pointless.
however some machines have ram that is not really usable as normal ram, for example some chipsets could support more ram than they could cache (you don't want uncachable ram as part of the main pool but as swap space its fine) and graphics card ram.
and you can buy (though they are pricey) hardware ramdrives that you can load up with more ram than most motherboards can take directly.
this is what that lump with lots of urls in was meant to say:
1: http://en.wikipedia.org/wiki/Terpsichori_Chryssoul aki-Vlachou <-- no citations, two external links only one of which is in english. The english one doesn't look very authoritive the greek one looks like some kind of newspaper but a local expert would be needed to determine its quality.
2: http://en.wikipedia.org/wiki/Kentucky_Route_11 <-- no citations, one external link to a cite that directly claims to be unofficial.
3: http://en.wikipedia.org/wiki/K%C4%93len <-- no citations, one external link to a person who claims to be the creator, no way of validating that claim.
4: http://en.wikipedia.org/wiki/James_Dixon_Murray <-- we actually have what appears to be an authoritive source mentioned (not cited but the article is so short thats forgivable). I can't check the book itself without quite some effort and there is no talk page or other indication that anyone else has done so.
5: http://en.wikipedia.org/wiki/Peruvian_national_ele ction%2C_2006 <-- no citations and no print references, looks like it may have some decent links to official sites but again only someone familiar with the locality could tell for sure.
6: http://en.wikipedia.org/wiki/Nayif_Abdallah_Ibrahi m_Al_Nukhaylan <-- essentially a duplicate of information from US government websites and more general information about gitmo (whose citation if any belongs in the articles about that in general which i have not checked). Overall pretty bare bones but what is said about this particular person is adequately cited.
7: http://en.wikipedia.org/wiki/Villa_Muti <-- no citations or external links or assertions of notablity whatsoever!
8: http://en.wikipedia.org/wiki/Stevenson_and_Higgins <-- no citations or external links whatsoever, use of the term queen elizibeth looks rather dubious, the would tend to reffer to a ship but the link points to a person whos life is in totally the wrong timeframe.
9: http://en.wikipedia.org/wiki/Est%C3%A1dio_do_Reste lo <-- once again no citations or external links
10: http://en.wikipedia.org/wiki/Wrestling_at_the_1952 _Summer_Olympics <-- once again no citations or external links
Is this a tacit admission that the "content over credentials" philosophy doesn't actually work in the real world? There is certainly an argument to be made. anyone who has edited wikipedia for a while and claims it works is either deluded in denial or lying.
FACT: on the vast majority of subjects the authoritive sources are not freely availible on the public web. Acessing them takes significant time and/or money. FACT: wikipedia very rarely (if ever) cites non web based sources, sometimes they are mentioned in a general references section but not cited. Even in the unlikely event that they are cited then almost noone can check up on them without either a long wait and/or payment neither of which most wikipedia editors would be likely to do for wikipedia alone. FACT: while there has been a push for more citation on wikipedia in recent times all this results in is citations of websites that are not authoritive and often contain misinformation (have a look at howstuffworks.com for a while and see how much misinformation you can find). Even if a website is authoritive how do you know this without subject area knowlage? FACT: even if you were to cite books and journals without knowlage of the field you can't know if the journals and articles cited are respectable or not. The effort of publishing reduces the ammount of junk but it doesn't eliminate it.
lets try 10 clicks on random page
1: http://en.wikipedia.org/wiki/Terpsichori_Chryssoul aki-Vlachou http://en.wikipedia.org/wiki/Kentucky_Route_11 http://en.wikipedia.org/wiki/K%C4%93len http://en.wikipedia.org/wiki/James_Dixon_Murray http://en.wikipedia.org/wiki/Peruvian_national_ele ction%2C_2006 http://en.wikipedia.org/wiki/Nayif_Abdallah_Ibrahi m_Al_Nukhaylan http://en.wikipedia.org/wiki/Villa_Muti http://en.wikipedia.org/wiki/Stevenson_and_Higgins http://en.wikipedia.org/wiki/Est%C3%A1dio_do_Reste lo http://en.wikipedia.org/wiki/Wrestling_at_the_1952 _Summer_Olympics -- once again no citations or external links
i bet most of theese articles could be edited by anyone but especially a self proffessed expert in the field to say anything that person wanted with little resistance.
wikipedia is great for certain things (computer geek subjects are VERY well covered) but you really need to take care if using information from it for anything important, especially once you go outside of its core subject area or when you delve into anything controversial.
You don't seem to understand the difference between lossy (as in JPG) and lossless (as in PNG) I do, png can losslessly store a 24 or 48 bit RGB image JPG can't.
it is the raw (as in raw data from the sensor)-rgb conversion that always involves loss.
And, surprise, there are even sensors that capture directly in RGB. i think there are a few layered sensors but i belive they are the exception not the rule. Afaict most sensors involve a monochrome sensor with a color pattern filter on top.
and even if they do the color balancing and gamma of the sensor is unlikely to match what most people want in thier RGB image.
so conversion is needed and that conversion almost inevitablly involves loss (either rounding, saturation or both)
while it is true that you can't directly see the information in the raw image you can see different parts of it by using different converter settings. Or maybe you might want to run some kind of denoising algorithm before adding in the artifacts of the conversion to rgb.
The RAW format is not a simple RGB bitmap. It contains additional data about the image, let alone about other crap. The whole point of RAW is that you can create the TIFF (or whatever) you want, not the TIFF that the camera thinks you want. my point exactly, to put it another way you can decide what data to throw away when making your TIFF from your RAW rather than letting the camera decide.
or use JPEG2000 and be safe s/safe/at risk of an unknown patent holder poping up and causing hell/
the older a format gets the less likely it is that will happen, patent holders only have so much patiance and they have a lot less leverage once the patent expires (they can still go after past damages but they won't be able to get injunctions and doctrines like latches will probablly also come into play limiting the damages they can get) and a patent newer than the reference encoder/decoder should not affect straight implementations of those because of prior-art (I know the US system currently sucks at enforcing this properly but its still better to have prior art than not to have it).
the forgent claim was a bombshell but i belive in the end the uspto invalidated the patent and it also expired meaning they can only claim against past use of jpeg not current use;) and the ongoing lawsuits were settled for relatively small sums. In summary forgent did manage to extort some money but the issue is over.
and that is why noone but graphics proffesionals uses it.
if you want to reasonablly be able to advertise the ability to read tiff files you have to drag in a huge number of different decoder libraries just to support most tiffs and you will still have people bitching about why thier tiff opens in photoshop but not in your app.
but there are still people who belive that png has poor browser support or that it is larger than gif (the fault of bad image editing tools that don't provide a clear enough seperation between changing color depth and changing format and of bad png endoders in adobe photoshop)
yes png has slowly crept onto the web but even with reasonablly tech savy people like web designers it has taken a LONG time to get acceptance.
lots of people don't have the equipment, the ears or both to tell the difference
i've ripped mp3 at 64k mono using the xing encoder (which i belive is regarded as pretty shitty compared to some modern encoders, it is FAST though) and most people i've met can't hear anything wrong with it. For most people FM radio is perfectly acceptable quality for pop music.
turning on drm by default is kind of shooting your own format in the foot too, lots of people may rip to it but its not going to be going very far if its drm'd
However, because there's no loss on the contary there is a lot of loss (and a lot of added bloat as well) going from raw CCD data to a RGB bitmap (which tiff then losslessly encodes) that is the whole point of having raw formats in the first place.
PNG with bzip2 is not worth it. PNG restarts the compression on each row. Compressing with PNG is like chopping a 1000000 byte file into 1000 files of 1000 bytes each, and then compressing the pieces separately. pure bullshit as another reply has stated
note that because deflate is a stream compressor though you can decompress as you download before receiving all the data.
At 1000 bytes per chunk, bzip2 and gzip compress about equally well. At even smaller sizes, gzip is often better. this may or may not be true but your previous error makes it irrelevent.
bzip2 on the raw image data actually competes very well with PNG. Sometimes PNG is better, and sometimes "raw.bz2" is better. but i bet a format based on png but using bzip2 in place of gzip would be better than either
it would be interesting to save a test image as a png, extract the IDAT chunk from the wrapper, decompress it and recompress it with bzip2
And yes, transforming the data before applying a compression algorithm is best of all. PNG actually does do that. A real simple transform (which PNG uses) that can help a lot is just feeding the difference between adjacent pixels to the compressor, rather than the pixel values themselves. png offers four filter types (plus the option of no filter at all which is often used for indexed color images) three of which do perform vertical comparisons. Filter type can be selected on a per-line basis.
Slashdot Mirror
For a backup of $HOME? Not so much /home).
depends how you value your time and physical space i guess (and of course on how much data you have in
my guess though is that dual layer DVDs have remained expensive because they have limited use (other than copying dvd based games). Most movies can be recompressed to fit on a single layer while keeping the quality loss to a level most people won't notice. For backing up a media collection or similar they aren't really big enough (imo the only people who are going to try and back up a large hard drive to DVD are those who already have too much time on thier hands and so are going to go for the cheaper per gigabyte single layer) and for backing up really important data they are generally bigger than nessacery.
is compliance with EMC standards, running high speed interfaces that are designed for internal use externally will almost certainly make the system non compliant (the same applies running a PC with the lid off and to many windowed cases)
does this matter? it depends! It doesn't if its for your own use and you don't have any sensitive gear arround or if you are a small fly by night firm or are just selling parts (generally standards are applied to the end product as a whole not individual parts) but any big vendors would not want to sell non compliant systems.
remember that metal case (even plastic cases have metal linings) on your PC isn't to keep RFI out, its to keep it in!
imo XP sp2 was really a new version of windows but they made it a service pack because the main area of improvement (security/malware resistance) was not something that most users would pay for but was something the lack of was getting them a lot of bad press for.
afaict there are three ways for a squatter (whether a typo-squatter or a domain snatcher) to make money
one way is advertising, some people will presumablly generate at least clicks and possiblly leads on the adverts.
another is selling the domains back to thier rightfull owners (you said you wouldn't buy but i bet many do)
the final way is drive-by installations of scumware (for users who use browsers with suitable security holes)
http://www.mini-itx.com/store/?c=16#p2620
Folks are smart. If something is an improvement, adoption happens.
folks are dumb, if something takes extra time in the short term then they are unlikely to adopt it regardless of its long term advantages.
most programmers (and i know i'm guilty of this myself) work on the assumption that fixes are easy and so quality doesn't matter much. This is true up to a point but trouble is people don't count the costs the bugs cause in the time between initially becoming apparent and being fixed. Nor the cost of tracing a bug that happens apparently randomly in production use or only on servers that run for months or whatever. How much does a MS security hotfix cost MS and thier customers? I don't have figures (probablly noone has figures for what it costs thier customers in agregate) but I bet its a VERY large sum of money.
In other engineering disciplines the costs are more obvious and the methodologies better developed.
Umm, Dapper isn't even the stable version, is it?
edgy is not a long term support release, support for it will be discontinued in a relatively short timeframe and its reported to be ubuntus buggiest release to date.
You speak as though all the support options rested on the shoulders of Canonical, but that is simply not true
do any of those companies have the rescources and inclination to do thier own tracking on what security issues pertain to what ubuntu versions and backport those security fixes themselves if ubuntus long term support promises turn out to be hollow?
the main downsides of ubuntu
it effectively exists on the whim of one rich guy who could lose interest and kill it whenever he wanted.
apparently the creaks are already showing in dapper (one person above mentioned downloading an update and his gui suddenly breaking with no availible soloution) and its less than a third of the way through its supposed support cycle. Support for the less major software in universe is even worse (much of the stuff in there has simply been imported straight from debian with no testing if it actually works in the ubuntu environment).
iirc long term support releases are supposed to come every two years and are finally obsoleted completely after 5 years. Maybe i'll be convinced when they have 3 of them in support at once (two for the desktop, 3 for the server) and are doing a good job for all of them i'll be convinced, until then ubuntu must be regarded as a young and vulnerable distro.
they could do that and they could also use the nasty flaw in IE that allows owners of versign certs to MITM ssl connections made by it provided you own a verisign issued certificate.
but people who share thier connections with broadband routers who plug in a PC with firefox already installed will not have any reason to run the ISPs software and hence will notice this and there isn't really anything they can do about it.
and it would only require one sufficiantly tech savy user using a browser that didn't have thier root cert and wasn't vulnerable to the exploit to spot what they were doing and out them and it could mean very bad press especially if the banks found out and started telling thier customers not to use said isp.
that sounds like a risky strategy, not having the software before the hardware is frozen could leave you easilly running out of some hardware rescource that you didn't realise the need for before you wrote the software.
Remember 1995? Dial up modems? 1 megapixel cameras? Back when file size mattered?
yes speeds are increasing but here in the uk the bottom end of "broadband" is only a few times faster than the fastest dialup which in turn is less than twice as fast as the modems that were current in that era.
many people are still using dialup either because they are light users, simply can't get broadband (at prices that are acceptable to normal customers, i am aware of the existance of leased lines and satalite links. Satalite links also have high latency which throws away much if not all of the percieved speed gain when browsing) or are on the move and can't rely on being able to connect to a network at the places they visit. Mobile data is also generally similar speeds to dialup and often has high per megabyte costs as well.
and consumer cameras are now at 5 megapixel or so
and hosting providers still charge significant prices per gigabyte so there is a definite incentive for webauthors to keep filesize down.
So what has changed in the last decade (here in the uk at least).
1: average net speeds for heavy users have increased significantly to the point where a web page loads instantly.
2: consumers have an increased taste for larger media, audio and video have become much more common and image sizes have grown with the capabilties of cameras and scanner handling software (handling an image of more than a megapixel or so on a 486 with 8 megs of ram is pretty impractical)
2: the gap between users has also increased a lot and is much harder for the user to close, in the old days all you needed to increase your speed was a one off payment for a better modem but to get beyond the 56K barrier you have to take out a subscription for every location you wan't that speed out on top of the normal phone line cost, and the time for getting it at a new site may be as high as a fortnight. Youd better check your site works ok over a crappy modem link because that is how it may have to be acccessed in a pinch and you don't want to lose a good customer because of that. (remember while the poor are on dialup so are those who are fairly rich but stuck in an awkward postition)
3: lots more people are on the net overall
4: p2p has come in and has been retailiated against with massive traffic limiting and dishonesty on the part of ISPs. Speeds on consumer internet connections have become very unpredictable and nowhere near what is being advertised.
Also, bzip2 is good with large amounts of information. And surprisingly PNG isn't the case: not the whole image is compressed but group of adjacent lines. That's done to allow network usage: user can see parts of image before it is completely downloaded.
totally incorrect, a png contains one long compressed stream of data (this may be split between multiple chunks to allow iterative generation but thats a minor overhead). Remember gzip is a stream compressor so you can feed it more input and get more output while keeping the compressor state intact.
What, are they going to go on a 4-day bender after the DST upgrades?
no they will be cleaning up after all the breakage and fixing things that are still using the old rules and therefore causing problems.
(I've rarely heard "Get me the best Dell, whatever the price!", I've heard that regularly about Macs.) ;)
o a/6924042/wo/H84TLj1rESUy25S5bl21WopohZG/2.?p=0 and see just how much a top spec mac pro would cost
they must be pretty rich then
for fun i decided to go to http://store.apple.com/Apple/WebObjects/ukstore.w
Price: £11,717.47
VAT: £2,050.56
Subtotal: £13,768.03
ouch!
that depends: on a traditional software ramdrive it is as you say pointless.
however some machines have ram that is not really usable as normal ram, for example some chipsets could support more ram than they could cache (you don't want uncachable ram as part of the main pool but as swap space its fine) and graphics card ram.
and you can buy (though they are pricey) hardware ramdrives that you can load up with more ram than most motherboards can take directly.
this is what that lump with lots of urls in was meant to say:
l aki-Vlachou <-- no citations, two external links only one of which is in english. The english one doesn't look very authoritive the greek one looks like some kind of newspaper but a local expert would be needed to determine its quality.
e ction%2C_2006 <-- no citations and no print references, looks like it may have some decent links to official sites but again only someone familiar with the locality could tell for sure.
i m_Al_Nukhaylan <-- essentially a duplicate of information from US government websites and more general information about gitmo (whose citation if any belongs in the articles about that in general which i have not checked). Overall pretty bare bones but what is said about this particular person is adequately cited.
s <-- no citations or external links whatsoever, use of the term queen elizibeth looks rather dubious, the would tend to reffer to a ship but the link points to a person whos life is in totally the wrong timeframe.
e lo <-- once again no citations or external links
2 _Summer_Olympics <-- once again no citations or external links
1: http://en.wikipedia.org/wiki/Terpsichori_Chryssou
2: http://en.wikipedia.org/wiki/Kentucky_Route_11 <-- no citations, one external link to a cite that directly claims to be unofficial.
3: http://en.wikipedia.org/wiki/K%C4%93len <-- no citations, one external link to a person who claims to be the creator, no way of validating that claim.
4: http://en.wikipedia.org/wiki/James_Dixon_Murray <-- we actually have what appears to be an authoritive source mentioned (not cited but the article is so short thats forgivable). I can't check the book itself without quite some effort and there is no talk page or other indication that anyone else has done so.
5: http://en.wikipedia.org/wiki/Peruvian_national_el
6: http://en.wikipedia.org/wiki/Nayif_Abdallah_Ibrah
7: http://en.wikipedia.org/wiki/Villa_Muti <-- no citations or external links or assertions of notablity whatsoever!
8: http://en.wikipedia.org/wiki/Stevenson_and_Higgin
9: http://en.wikipedia.org/wiki/Est%C3%A1dio_do_Rest
10: http://en.wikipedia.org/wiki/Wrestling_at_the_195
Is this a tacit admission that the "content over credentials" philosophy doesn't actually work in the real world? There is certainly an argument to be made.
l aki-Vlachou http://en.wikipedia.org/wiki/Kentucky_Route_11 http://en.wikipedia.org/wiki/K%C4%93len http://en.wikipedia.org/wiki/James_Dixon_Murray http://en.wikipedia.org/wiki/Peruvian_national_ele ction%2C_2006 http://en.wikipedia.org/wiki/Nayif_Abdallah_Ibrahi m_Al_Nukhaylan http://en.wikipedia.org/wiki/Villa_Muti http://en.wikipedia.org/wiki/Stevenson_and_Higgins http://en.wikipedia.org/wiki/Est%C3%A1dio_do_Reste lo http://en.wikipedia.org/wiki/Wrestling_at_the_1952 _Summer_Olympics -- once again no citations or external links
anyone who has edited wikipedia for a while and claims it works is either deluded in denial or lying.
FACT: on the vast majority of subjects the authoritive sources are not freely availible on the public web. Acessing them takes significant time and/or money.
FACT: wikipedia very rarely (if ever) cites non web based sources, sometimes they are mentioned in a general references section but not cited. Even in the unlikely event that they are cited then almost noone can check up on them without either a long wait and/or payment neither of which most wikipedia editors would be likely to do for wikipedia alone.
FACT: while there has been a push for more citation on wikipedia in recent times all this results in is citations of websites that are not authoritive and often contain misinformation (have a look at howstuffworks.com for a while and see how much misinformation you can find). Even if a website is authoritive how do you know this without subject area knowlage?
FACT: even if you were to cite books and journals without knowlage of the field you can't know if the journals and articles cited are respectable or not. The effort of publishing reduces the ammount of junk but it doesn't eliminate it.
lets try 10 clicks on random page
1: http://en.wikipedia.org/wiki/Terpsichori_Chryssou
i bet most of theese articles could be edited by anyone but especially a self proffessed expert in the field to say anything that person wanted with little resistance.
wikipedia is great for certain things (computer geek subjects are VERY well covered) but you really need to take care if using information from it for anything important, especially once you go outside of its core subject area or when you delve into anything controversial.
You don't seem to understand the difference between lossy (as in JPG) and lossless (as in PNG)
I do, png can losslessly store a 24 or 48 bit RGB image JPG can't.
it is the raw (as in raw data from the sensor)-rgb conversion that always involves loss.
And, surprise, there are even sensors that capture directly in RGB.
i think there are a few layered sensors but i belive they are the exception not the rule. Afaict most sensors involve a monochrome sensor with a color pattern filter on top.
and even if they do the color balancing and gamma of the sensor is unlikely to match what most people want in thier RGB image.
so conversion is needed and that conversion almost inevitablly involves loss (either rounding, saturation or both)
while it is true that you can't directly see the information in the raw image you can see different parts of it by using different converter settings. Or maybe you might want to run some kind of denoising algorithm before adding in the artifacts of the conversion to rgb.
The RAW format is not a simple RGB bitmap. It contains additional data about the image, let alone about other crap. The whole point of RAW is that you can create the TIFF (or whatever) you want, not the TIFF that the camera thinks you want.
my point exactly, to put it another way you can decide what data to throw away when making your TIFF from your RAW rather than letting the camera decide.
or use JPEG2000 and be safe
;) and the ongoing lawsuits were settled for relatively small sums. In summary forgent did manage to extort some money but the issue is over.
s/safe/at risk of an unknown patent holder poping up and causing hell/
the older a format gets the less likely it is that will happen, patent holders only have so much patiance and they have a lot less leverage once the patent expires (they can still go after past damages but they won't be able to get injunctions and doctrines like latches will probablly also come into play limiting the damages they can get) and a patent newer than the reference encoder/decoder should not affect straight implementations of those because of prior-art (I know the US system currently sucks at enforcing this properly but its still better to have prior art than not to have it).
the forgent claim was a bombshell but i belive in the end the uspto invalidated the patent and it also expired meaning they can only claim against past use of jpeg not current use
and that is why noone but graphics proffesionals uses it.
if you want to reasonablly be able to advertise the ability to read tiff files you have to drag in a huge number of different decoder libraries just to support most tiffs and you will still have people bitching about why thier tiff opens in photoshop but not in your app.
but there are still people who belive that png has poor browser support or that it is larger than gif (the fault of bad image editing tools that don't provide a clear enough seperation between changing color depth and changing format and of bad png endoders in adobe photoshop)
yes png has slowly crept onto the web but even with reasonablly tech savy people like web designers it has taken a LONG time to get acceptance.
lots of people don't have the equipment, the ears or both to tell the difference
i've ripped mp3 at 64k mono using the xing encoder (which i belive is regarded as pretty shitty compared to some modern encoders, it is FAST though) and most people i've met can't hear anything wrong with it. For most people FM radio is perfectly acceptable quality for pop music.
turning on drm by default is kind of shooting your own format in the foot too, lots of people may rip to it but its not going to be going very far if its drm'd
However, because there's no loss
on the contary there is a lot of loss (and a lot of added bloat as well) going from raw CCD data to a RGB bitmap (which tiff then losslessly encodes) that is the whole point of having raw formats in the first place.
PNG with bzip2 is not worth it. PNG restarts the compression on each row. Compressing with PNG is like chopping a 1000000 byte file into 1000 files of 1000 bytes each, and then compressing the pieces separately.
pure bullshit as another reply has stated
note that because deflate is a stream compressor though you can decompress as you download before receiving all the data.
At 1000 bytes per chunk, bzip2 and gzip compress about equally well. At even smaller sizes, gzip is often better.
this may or may not be true but your previous error makes it irrelevent.
bzip2 on the raw image data actually competes very well with PNG. Sometimes PNG is better, and sometimes "raw.bz2" is better.
but i bet a format based on png but using bzip2 in place of gzip would be better than either
it would be interesting to save a test image as a png, extract the IDAT chunk from the wrapper, decompress it and recompress it with bzip2
And yes, transforming the data before applying a compression algorithm is best of all. PNG actually does do that. A real simple transform (which PNG uses) that can help a lot is just feeding the difference between adjacent pixels to the compressor, rather than the pixel values themselves.
png offers four filter types (plus the option of no filter at all which is often used for indexed color images) three of which do perform vertical comparisons. Filter type can be selected on a per-line basis.