Yes, seek time is no longer the proper term for it, but people use it. IOPS is the relevant measure, and as I have said, they do not list write IOPS. Read IOPS for the recent SanDisk SSD announced was 7000; this is much better than spinning media, but still pretty bad for SSD. Write IOPS are much worse, and they won't even list that.
In any case, the difference between flash and RAM SSD is far more than 10x.
That's nothing; in terms of rotating things, flywheel batteries are much more interesting. They have achieved a velocity of 2km/s at the edge. (about Mach 6)
You do realize that the SSD you reference is based on flash, right? If you look carefully, you will find that no vendors list write seek times or write IOPS for such devices. The reason is that the performance is just plain awful.
RAM based SSD is nice, but flash based SSD won't touch a decent 15k drive for any write heavy application.
All the legislation in the world hasn't done jack-shit for China, but economics doesn't care about politics... it just works. It may take 50 years, but it's better than pissing money down a hole. So you can keep on having faith in your 'intelligent design' theory of helping third world countries, and I'll keep on believing in my 'economic evolution' theory backed by real world evidence. Do you know what 50 years is? That is more or less a persons entire life. I am appalled at the ease with which you render it insignificant with your hand waving. If that life was yours, I dare say you would feel quite differently.
Except that the NFS client on the mac is one of the worst available, so it hardly matters. It is miserably slow and buggy, does not support v4, does not support large writes, handles direct I/O *very* poorly (which the Finder does a lot of...), and of course does not do Unicode re-normalization.
Just one of the dozens of areas of MacOS X which contain serious problems and has been ignored for years.
I hope you mean that the current protection isn't a generally useful solution... right? No, I mean that it will never be useful in creating an open source player that is capable of playing all (or even most) discs. Nor for exercising your other fair use rights.
Because while new discs can be pressed, which pirate would be stupid enough to buy stuff from a store and trying to decrypt it, when s/he can just download the previously decrypted version from a favorite BitTorrent tracker? I was only referring to playback of HD discs. Piracy is a completely separate issue, and for it, AACS is completely ineffective. AACS only hurts the legitimate customers; by advertising this "crack," some people will assume that it is now safe to buy HD discs, and this is simply not true.
What I wonder most is what will happen to the software HD-DVD player industry now... I can seriously see the movie industry stopping to support all of them now to protect key extraction. This is the big question, and it is likely they will stop supporting playback on anything but fully "Trusted" systems. It seems I was wrong about the TPM providing decryption services, but once Trusted Computing arrives, this won't matter.
New disks can be pressed with new keys, and the compromised software player will have it's key revoked. As such, this is not a generally useful solution. AACS remains secure, and at best, we may see individual keys available for certain pressings of certain discs. This approach will never provide general playback as DeCSS does.
However, it is my understanding that the decryption process can be done by the TPM; once this is supported, the problem will be much more difficult. Make no mistake, the battle has only just begun. Before long, software based attacks may be rendered impossible.
YES! We know there's a small reduction in quality. No, it is not a small reduction in quality. Once you encode something in a format like mp3 or aac, it introduces numerous artifacts which render the the resulting audio very difficult to re-compress. At a similar bitrate, the resulting audio will sound awful.
Apple's current encoding quality is only barely acceptable. The only acceptable option would be to re-encode it in a lossless format, and this is rarely useful for playback on a portable device.
Hopefully any net neutrality legislation will do exactly that, but it is doubtful. QoS will never work on the Internet itself; allowing it will only further discrimination of traffic, and promote abuse of it. Traffic on the internet should be treated fairly, and QoS does not do that.
There is only one way to ensure quality VOIP and IPTV, and that is to run these services on private networks. On networks where multicast can be supported. On networks where QoS can be properly implemented. These networks should to be funded by the services that run on them, and the rest of us shouldn't have to pay for them.
This does not mean that Vonage and the like can not be run over the Internet. As long as there is no discrimination, and there is adequate capacity, it should still run fine. This goes for everything on the Internet. By trying to support QoS at that level, it only provides an excuse to degrade service of non-blessed traffic instead of increasing capacity.
On the Internet, no traffic should have priority, period. If that isn't good enough, then you need to pay for access to private networks where service can be guaranteed.
When I went to cancel my Netflix service, they happily complied. After doing so, they conveniently "lost" all the discs that I had sent back to them. (So I discovered after getting the bill...)
Of course, Netflix provided absolutely no customer service contact information. I believe there was a customer support web form, but that was only available to members with an active acount.
Why would you possibly do that? Add more (cheap!) physical RAM instead until there's no need to swap. Well, if you have a Mac you can't add more cheap physical RAM. Every Mac except the Mac Pro has two SODIMM slots, that is all. For reference, a 2GB SODIMM costs about $750US now, so the flash disk doesn't look so bad anymore. Of course I would rather add four 1GB DIMMs at 1/4th the cost, but adding memory cheaply is a privilege that only >$200 PC owners have.
That said, using a flash disk for swap really doesn't help, since write performance of flash is so awful to begin with.
To date, has Freescale completed the PWRficient and offered it for sale? P.A. Semi is designing the PWRficient, not Freescale. Freescale had chips for sale though which would have been adequate for Apple in the interim.
What do you mean by "support an extra architecture", by the way? Apple now has to support IA32 and x86-64. Unlike the well designed 32-bit PPC instruction set, IA32 offers no advantages, only baggage. It is not necessary to provide Windows virtualization, and it only increases the number of binaries that Apple much support and ship.
And it would still be slower than a Core 2 Duo. I'll take the slower chip any day, if the battery in my laptop lasts twice as long.
This of course assuming the PWRficient ever actually gets made. The PPC architecture is full of promises of amazing products that never actually see the light of day. You are talking about the old Motorola. Ever since, Freescale and IBM have done very well. IBM never promised Apple a G5 laptop. They did however offer them options before the switch, such as a Cell variant, and Apple turned them all down.
Furthermore, Apple is responsible for keeping the G4 on the MPX bus, which is why it failed to compete. Freescale finally went ahead and added on-chip memory controllers, and the 8641/8641D were great chips that Apple could have used. They would have kept Apple competitive until the P.A. Semi or Cell chips arrived. While the Core 2 Duo is not bad, the Core Duo was a lame chip, and now they have to support an extra architecture.
You mean a processor from a fabless company announced six months after Apple announced the switch to Intel, and wasn't expected to sample until nearly a year after the first Intel Macintosh shipped?
P.A. Semi approached Apple before the switch, and Apple turned them down. It is not just a "processor from a fabless company," it is a processor from the lead architect of the Alpha and StrongARM, and with a very talented team. Here are two of the finest architectures ever created, one with leading performance, and the other leading in power consumption. A little known fact is that a low-power Alpha was designed, but it was decided that there was no market at the time. Truly a shame to see that architecture go.
In any case, we are at a time where the outlook of the PPC has never looked better. Apple had options, and performance per Watt is a flimsy excuse at best. Intel may be selling some decent processors now, but they will never be as cheap or efficient as the PPC. (and with the integration offered by P.A. Semi, this was only improving.) Look what happened to the price of the Mini. With this move, Apple has forsaken a huge and growing market segment.
What sort of system databases are you concerned about? Passwd? Locate? The latter could be an an issue if it indexed home directories, but it usually doesn't. In fact, I can't think of anything critical on a typical desktop machine.
Most decent OSs provide encrypted swap, so that is not an issue.
Not too long ago, a brilliant discovery was made in engine design. It is now possible to build a clean and efficient two-stroke engine which runs on a variety of fuels. This engine is fully balanced, and stable enough to balance a golf ball on during operation. Furthermore, the engine is very compact, with a high power to weight ratio. (under 1lb/HP)
For more information on OPOC engines, have a look at the following:
If you really look at ObjC objectively, it is kind of disgusting. Don't get me wrong, it is usable, and will "normally" get muppets in to less trouble than C++, but it is C with bolt on bits of smalltalk syntax, and a very odd type system. And the tool support is quite, quite tragic.
I do not think you are looking at ObjC objectively; in no way is it disgusting. It is a very simple, elegant, and powerful extension of the C language. The dynamic typing system may seem odd to you, but it is extremely flexible, and largely responsible for the beauty of Cocoa and GNUStep. The Smalltalk infix syntax may feel foreign to a C programmer, but it in the end, it is compact, and expressive. Once you use it, you will never want to go back. It is easier to code, and much more readable.
As far as tool support, it is standard in gcc; what more do you need?
By the way, if ever anything had a bolted-on feel to it, that would be C++.
The electromagnetic spectrum has limits, people This is a huge myth, nothing more. Photons don't take up any space, so there is effectively no limit.
The current "limits" are mostly because we make extremely poor use of wireless transmissions. Devices have no good way of weeding out signals not for them, so the only option is to limit interference. As phased array antennas and other techniques become more common, we can make much better use of the spectrum. Low power, wide band, directional transmissions, with fine meshes will allow the network to scale to *huge* amounts of bandwidth.
The limits are practical limits, and they will improve as our technology improves. What needs to be done, is to slowly wrest the licensed spectrum (which is essentially wasted) from the broadcast corporations, and give it to the people.
Before any new law may be passed, the legal code shall be reviewed in it's entirety and thoroughly checked for existing laws serving the same purpose. If any such law shall exist, the proposed law may not be passed. If multiple laws serving the same purpose are found, they shall be reconciled into one non-self-contradictory law with the eldest law taking precedence.
This is a great idea, and it can be taken further. Aside from searching for duplicate laws, some entity should be tasked with factoring existing laws.
Laws only get more complex now, and as you point out, it simply isn't scalable. Another advantage of de-duplicating laws, and re-factoring them, is that more attention will be paid to the intent of the law, rather than the letter of the law. No doubt, this is a huge undertaking, but it needs to be done, and we are certainly not short on lawyers. (Though, we may be short on honest ones.) I wouldn't mind if they all dropped dead, but this would be a much more productive activity for them. (Ok, not all lawyers are inherently evil, but a very small fraction of them are actually improving our lives.)
How can a good filesystem be overkill? In any case, a filesystem's primary goal is to store and retrieve data correctly. HFS+ makes no guarantees about this.
Data integrity is just as important for home computing needs, perhaps more so.
The only reason ZFS has not replaced UFS in Solaris, is because it is not yet bootable. It offers many of its advantages even on a single disk, and before long, it will be the default. There is no reason to expect it won't be the default on MacOS as well by 10.6. Even on an iMac for your average user.
The only reason UFS will remain around, is for compatibility, and those certain workloads where the in-place block update model is still preferable.
Maybe there aren't any *nixes that doe exactly what they want? And there's a reasonable amount of FreeBSD and NeXT stuff flaoting around in there already. Any one of Linux, Solaris, or FreeBSD would provide a much more solid foundation than Darwin. All of these are fundamentally similar enough that porting the GUI would be less effort than maintaining Darwin. About the only thing that needs considerable work are the various Mach interfaces, and it is arguable that those should go anyways. Darwin is a bastardized FreeBSD/Mach combination, and will always be limited by the poor performance of Mach. The valuable NeXT code lives on in Cocoa, and is almost entirely independent of the underlying OS.
Whoever's fording DRM down your throat, it isn't Apple. I've a few songs I bought off iTunes, but I wasn't forced to and can use plenty of non-DRMed media on my Mac. I may have speculated too much on this, but this is not the DRM I am referring to. MacOS already takes advantage of the TPM, to keep it running on Apple hardware, and I expect things will be worse with 10.5.
As a desktop OS, most *nixes are half-baked and severely lacking. As a desktop OS, there's nothing that meets my needs better than OS X and that's what it's trying to be, not just another *nix. *nix stuff is just a nice bonus Linux as a desktop OS is lacking, but it provides a very solid foundation. The "*nix stuff" isn't just a nice bonus, it is the foundation of MacOS X, and it is very brittle and shaky.
Over 90% of the market for computers is under $599? News to me. Apple does not sell a system for less than $999. $599 is for a bare computer. Furthermore, you have to spend an extra $200 just to get a rewrite-able DVD drive that is available on far cheaper PCs. Anyways, I was wrong, the PC market for systems over $999 is actually only 4%. This post includes a link to an article, and also summarizes the current PC market demographics.
That must be why they're selling record number of computers and have a growing market share. Maybe they should just close down the company and return the money to the shockholders? No, maybe they should do their stockholders a favor, and address the changing market. The $500 PC accounts for 38% of the market, and it is growing fast. Also, the $599 Mini does not use standard hardware, and is not upgradable, making it a very poor value compared to these cheaper machines.
Anyways, there are a ton of people who have been waiting for the "xMac"; Apple could do a whole lot better.
Re:Hmmm....WMV9 on OS X?
on
VLC 0.8.6 Released
·
· Score: 5, Informative
I have a PPC mac and it is great to have a native working WMV9/VC-1 codec. While I haven't tried it in VLC, I have used it in the recent MPlayer dev builds, and it is much better than flip4mac.
Not to pick on you specifically, but just what is it people expect the Finder to do? Is it a performance issue, or just a dislike regarding the way the interface works?
I expect the Finder to allow me to quickly and efficiently manage my files, with a minimum of surprise, that is all. If you work with a lot of files, you will notice the severe deficiencies in the Finder, especially on network volumes. It can't even keep the contents of a folder straight, with items disappearing, reappearing, and resorting themselves for no apparent reason. You can't move two items from the same folder to different destinations at the same time. Folders don't remember their configuration properly or consistently. Constant hangs with the Spinning Wheel of Death. Double-clicking an item to open it causes a resort on the first click, and then you may open the wrong item. There is even a data loss bug, where you might unknowingly send something to the trash, even when nothing is selected. Really, the list goes on and on, and it is full of serious problems and inconveniences alike. Every time I report these bugs, they are closed as duplicates, and simply ignored.
I'd also question the need for Apple to embrace a more OSS-friendly dev model. They seem to be doing just fine the way they're going now, even better than they were when they released 10.1.
If they could maintain maintain a competitive *nix, this wouldn't be as much of an issue. Look how much development goes into just the Linux kernel; Apple can't even hope to compete on a technical basis, and will only fall further and further behind. This means things like scalable SMP, efficient threading, network file systems, disk drivers (NCQ anyone?), networking, and many other technical things which while not sexy have a great impact on performance. This work simply isn't getting done. Their low-level OS effort would have a much greater benefit if expended on the GUI and interface instead; these are the areas which distinguish MacOS. Microsoft can't even competing with Linux in these areas, and Apple has but a small fraction of their resources.
As it is now, there are an immense amount of bugs, not to mention very poor performance, and it is basically impossible to even contribute fixes to Apple, which is very frustrating. Apple's uncooperative attitude is simply not productive.
... Apple has rarely been one to simply sit on a good product and not try to continue to make it better/newer. This may or may not be a good thing. Apple is also one to sit on a bad product, and let it stagnate. The Finder desperately needs attention, and improvements to it could profoundly improve the MacOS experience. Darwin (the core of MacOS) is also a similar case. I think most would agree that Aqua is very good already, and as such, Apple should focus its efforts elsewhere. (In the case of Darwin, they should embrace a successful open source *nix instead of wasting their efforts. However, I think they will stick with Darwin so that they have the opportunity to force DRM down our throats.)
Too many critical things have been ignored for far too long, while Apple implements features and eye-candy which often provide no utility. Features are not necessarily a bad thing, but Apple keeps on adding new ones without ever really finishing any of them. MacOS today is a mess, and has none of the consistency or polish that it once did. As a *nix, it is also half-baked, and severely lacking.
I keep telling myself, "I have had enough" with each release, but 10.5 may be the last. (Though, I hope not, since the Cocoa/NeXT dev model is excellent.) If Apple does not embrace a more OSS friendly (read: community friendly) development model though, I feel that they will relegate themselves to irrelevance. Similar things can be said about their hardware business--currently every computer they sell is priced outside of > 90% of the market. Not that they are not competitive on price; they simply ignore almost the entire market. This is not sustainable...
Slashdot Mirror
Yes, seek time is no longer the proper term for it, but people use it. IOPS is the relevant measure, and as I have said, they do not list write IOPS. Read IOPS for the recent SanDisk SSD announced was 7000; this is much better than spinning media, but still pretty bad for SSD. Write IOPS are much worse, and they won't even list that.
In any case, the difference between flash and RAM SSD is far more than 10x.
That's nothing; in terms of rotating things, flywheel batteries are much more interesting. They have achieved a velocity of 2km/s at the edge. (about Mach 6)
Take a look at http://www.llnl.gov/str/pdfs/04_96.2.pdf
You do realize that the SSD you reference is based on flash, right? If you look carefully, you will find that no vendors list write seek times or write IOPS for such devices. The reason is that the performance is just plain awful.
RAM based SSD is nice, but flash based SSD won't touch a decent 15k drive for any write heavy application.
Except that the NFS client on the mac is one of the worst available, so it hardly matters. It is miserably slow and buggy, does not support v4, does not support large writes, handles direct I/O *very* poorly (which the Finder does a lot of...), and of course does not do Unicode re-normalization.
Just one of the dozens of areas of MacOS X which contain serious problems and has been ignored for years.
I can seriously see the movie industry stopping to support all of them now to protect key extraction. This is the big question, and it is likely they will stop supporting playback on anything but fully "Trusted" systems. It seems I was wrong about the TPM providing decryption services, but once Trusted Computing arrives, this won't matter.
New disks can be pressed with new keys, and the compromised software player will have it's key revoked. As such, this is not a generally useful solution. AACS remains secure, and at best, we may see individual keys available for certain pressings of certain discs. This approach will never provide general playback as DeCSS does.
However, it is my understanding that the decryption process can be done by the TPM; once this is supported, the problem will be much more difficult. Make no mistake, the battle has only just begun. Before long, software based attacks may be rendered impossible.
Apple's current encoding quality is only barely acceptable. The only acceptable option would be to re-encode it in a lossless format, and this is rarely useful for playback on a portable device.
Hopefully any net neutrality legislation will do exactly that, but it is doubtful. QoS will never work on the Internet itself; allowing it will only further discrimination of traffic, and promote abuse of it. Traffic on the internet should be treated fairly, and QoS does not do that.
There is only one way to ensure quality VOIP and IPTV, and that is to run these services on private networks. On networks where multicast can be supported. On networks where QoS can be properly implemented. These networks should to be funded by the services that run on them, and the rest of us shouldn't have to pay for them.
This does not mean that Vonage and the like can not be run over the Internet. As long as there is no discrimination, and there is adequate capacity, it should still run fine. This goes for everything on the Internet. By trying to support QoS at that level, it only provides an excuse to degrade service of non-blessed traffic instead of increasing capacity.
On the Internet, no traffic should have priority, period. If that isn't good enough, then you need to pay for access to private networks where service can be guaranteed.
When I went to cancel my Netflix service, they happily complied. After doing so, they conveniently "lost" all the discs that I had sent back to them. (So I discovered after getting the bill...)
Of course, Netflix provided absolutely no customer service contact information. I believe there was a customer support web form, but that was only available to members with an active acount.
That said, using a flash disk for swap really doesn't help, since write performance of flash is so awful to begin with.
Furthermore, Apple is responsible for keeping the G4 on the MPX bus, which is why it failed to compete. Freescale finally went ahead and added on-chip memory controllers, and the 8641/8641D were great chips that Apple could have used. They would have kept Apple competitive until the P.A. Semi or Cell chips arrived. While the Core 2 Duo is not bad, the Core Duo was a lame chip, and now they have to support an extra architecture.
What sort of system databases are you concerned about? Passwd? Locate? The latter could be an an issue if it indexed home directories, but it usually doesn't. In fact, I can't think of anything critical on a typical desktop machine.
Most decent OSs provide encrypted swap, so that is not an issue.
Not too long ago, a brilliant discovery was made in engine design. It is now possible to build a clean and efficient two-stroke engine which runs on a variety of fuels. This engine is fully balanced, and stable enough to balance a golf ball on during operation. Furthermore, the engine is very compact, with a high power to weight ratio. (under 1lb/HP)
o ping_.html
For more information on OPOC engines, have a look at the following:
http://www.greencarcongress.com/2005/05/fev_devel
http://www.propulsiontech.com/opocengine.html
http://home.arcor.de/hildst/EnEx99e.html
As far as tool support, it is standard in gcc; what more do you need?
By the way, if ever anything had a bolted-on feel to it, that would be C++.
The current "limits" are mostly because we make extremely poor use of wireless transmissions. Devices have no good way of weeding out signals not for them, so the only option is to limit interference. As phased array antennas and other techniques become more common, we can make much better use of the spectrum. Low power, wide band, directional transmissions, with fine meshes will allow the network to scale to *huge* amounts of bandwidth.
The limits are practical limits, and they will improve as our technology improves. What needs to be done, is to slowly wrest the licensed spectrum (which is essentially wasted) from the broadcast corporations, and give it to the people.
This is a great idea, and it can be taken further. Aside from searching for duplicate laws, some entity should be tasked with factoring existing laws.
Laws only get more complex now, and as you point out, it simply isn't scalable. Another advantage of de-duplicating laws, and re-factoring them, is that more attention will be paid to the intent of the law, rather than the letter of the law. No doubt, this is a huge undertaking, but it needs to be done, and we are certainly not short on lawyers. (Though, we may be short on honest ones.) I wouldn't mind if they all dropped dead, but this would be a much more productive activity for them. (Ok, not all lawyers are inherently evil, but a very small fraction of them are actually improving our lives.)
How can a good filesystem be overkill? In any case, a filesystem's primary goal is to store and retrieve data correctly. HFS+ makes no guarantees about this.
Data integrity is just as important for home computing needs, perhaps more so.
The only reason ZFS has not replaced UFS in Solaris, is because it is not yet bootable. It offers many of its advantages even on a single disk, and before long, it will be the default. There is no reason to expect it won't be the default on MacOS as well by 10.6. Even on an iMac for your average user. The only reason UFS will remain around, is for compatibility, and those certain workloads where the in-place block update model is still preferable.
Anyways, there are a ton of people who have been waiting for the "xMac"; Apple could do a whole lot better.
I have a PPC mac and it is great to have a native working WMV9/VC-1 codec. While I haven't tried it in VLC, I have used it in the recent MPlayer dev builds, and it is much better than flip4mac.
If they could maintain maintain a competitive *nix, this wouldn't be as much of an issue. Look how much development goes into just the Linux kernel; Apple can't even hope to compete on a technical basis, and will only fall further and further behind. This means things like scalable SMP, efficient threading, network file systems, disk drivers (NCQ anyone?), networking, and many other technical things which while not sexy have a great impact on performance. This work simply isn't getting done. Their low-level OS effort would have a much greater benefit if expended on the GUI and interface instead; these are the areas which distinguish MacOS. Microsoft can't even competing with Linux in these areas, and Apple has but a small fraction of their resources.
As it is now, there are an immense amount of bugs, not to mention very poor performance, and it is basically impossible to even contribute fixes to Apple, which is very frustrating. Apple's uncooperative attitude is simply not productive.
... Apple has rarely been one to simply sit on a good product and not try to continue to make it better/newer. This may or may not be a good thing. Apple is also one to sit on a bad product, and let it stagnate. The Finder desperately needs attention, and improvements to it could profoundly improve the MacOS experience. Darwin (the core of MacOS) is also a similar case. I think most would agree that Aqua is very good already, and as such, Apple should focus its efforts elsewhere. (In the case of Darwin, they should embrace a successful open source *nix instead of wasting their efforts. However, I think they will stick with Darwin so that they have the opportunity to force DRM down our throats.)Too many critical things have been ignored for far too long, while Apple implements features and eye-candy which often provide no utility. Features are not necessarily a bad thing, but Apple keeps on adding new ones without ever really finishing any of them. MacOS today is a mess, and has none of the consistency or polish that it once did. As a *nix, it is also half-baked, and severely lacking.
I keep telling myself, "I have had enough" with each release, but 10.5 may be the last. (Though, I hope not, since the Cocoa/NeXT dev model is excellent.) If Apple does not embrace a more OSS friendly (read: community friendly) development model though, I feel that they will relegate themselves to irrelevance. Similar things can be said about their hardware business--currently every computer they sell is priced outside of > 90% of the market. Not that they are not competitive on price; they simply ignore almost the entire market. This is not sustainable...