I grew up in Los Alamos and I worked there during my high school years through some of graduate school. The article completely failed to mention one of the main culprits for a lot of these problems: The Department of Energy. While I do not have knowledge beyond what is in the press for most of the incidents mentioned, the ones where I do mostly include a major role in the problem played by DOE ranging from their screwed up policies to direct involvement. Given this, a new contractor can only do so much.
The Roku rates pretty low in terms of scriptability/hackabilty IMHO. Yes, it has an API, but through numerous poor choices on Roku's part, this is not as useful as other platforms. Their Brightscript is not well designed, and the VM is buggy. I can cause the device to hard reset with one line of code (an accidental discovery trying to make the thing work). The library availability for the language is poor. Heaven help you if you want to do something that's provided in libraries on the net in nearly every language but not available on Brightscript. If you want to do something very very simple, you can pull it off, but more complicated programs are an exercise in frustration. Additionally, their provided displays are missing key functionality. For example, the Grid Display cannot be stacked without crashing the device, you cannot set the selection on it before it is displayed, and there is no callback for the display of the grid so you can set the selection then. Also, use the debugger too much, and you'll lock up the device, requiring pulling the power to correct it. Lastly, you cannot have more than one development app on the device at a time.
I went to a GoogleTV, and it's hackability is vastly superior to the Roku. It uses a language which has a large library availability (java) and the programming environment is inside an IDE (Eclipse). The remote debugger is more useful, synchronizing the execution with the location in the code and displaying local variables, in stark contrast to using telnet on a port to step through which is essentially a stripped down command line gdb (which is missing many of the more useful commands, such as breakpoints).
I cannot speak for the other devices, but if they were to provide any API for third party development, I would be surprised if they were not superior to the Roku. It has a development environment you would expect from the previous decade, not this one.
> later when they upgraded the OS (without breaking any compatibility) This caused me to laugh. I have a Roku 2 and experienced a regression on http live streaming resulting in audio dropouts, video freezing, and even the device rebooting. My dad has a Roku 1, and experienced regressions on two separate "updates." This is not isolated; just see the forums and the number of requests to revert the firmware and Roku's refusal to allow it.
It is being tested... The problem is that the changes rely on the ssl in the libcurl library that 10.1 does not have. Fire is maintaing compatibility with 10.1, proteus does not. Expect a release today.
For my senior design project, we had to make some dedicated hardware to exactly this. It was solved using an FFT inside a Vertex-II Pro FPGA using a sliding window.
Now, as far as this contest goes, this can be done quite easilly in a computer. In fact, we had made a bitwise simulation of the FPGA using matlab before doing any Verilog to make sure that our design worked. My favorite was the little setup that we did in LabView. Took about 30 minutes to make, and would solve this quite easilly. It was very nice to see the result in color intensity with frequency along the x axis, and time across the y. Seeing the data in there was quite easy even though the SNR was so low that you couldn't even hear most of the tones.
In short, if you want to do this, read the data into matlab (or octive), do an FFT on a small window of the data, slide the window over, repeat. Take all the FFT's and plot in an intenisty graph (not sure what the matlab term is, but I am sure it has such a thing).
It would also be a fun hobby and a great way for open source to venture into the hardware realm.
Open sources has extended into the hardware realm. I am sure there more sites, but I have found Open Cores to be quite helpful. They have all kinds of things, and some are GPL.
First of all, broken link. I assume you were refering to this. Now, you first mentioned 42bit VIRTUAL ADDRESSES. There is a BIG DIFFERENCE between virtual and real addresses.
There are three address types: Effective: This is the address that the program sees. This may (and often does) have very little to do with the real address Virtual: This is used only within the MMU and Page tables. Physical (Real): This is the actual address in the memory.
Now, the link you provided states 42bits of real address. This does not mean that the virtual address is 42bits. The PEM is very acurate in these cases since it describes the page table setup that the OS has to take care of and maintain. The same file you reference states that the 970 has 64bit effective addresses. This is consistent with the PEM! Notice the PEM also states that the physical addresses are ?64bits. This is also consistent with the 970.
Lastly, 42bit real addressing is only a limit in the amount of physical ram that the chip can address. This is not a limit in the amount of total memory (physical and virtual). This means that your Ram, OS, and HD space is the only practical limit in allocating massive amounts of memory. Assuming you have enough HD to handle 2^64 bytes, then you can allocate that much. The chip is not a limiting factor. Also, if you look at the specifics of the chip's memory management, you will see that you can change it to use other PTEG's if you change some of the registers, so if you are willing to go through the context switch expense; 2^64 bytes is a limit per executable.
Actually, it is 52 bit for the older PowerPC's: Page 7-35 It is 64 or 80 for the 970: Page 258.
It is also important to note that these values are a per program basis if you are willing to change around some special registers in your context switches.
As a Fire developer myself, I thought that I could contribute a little more to this. We have started to participate in a discussion on the best way to do encryption over IM protocols. This discussion can be found here: http://www.chat.solidhouse.com/smsn/. The GAIM-E author has even contributed to this discussion.
Also, we have drastically improved the way that the GPG encryption is handled. It now works on more protocols and will be more consistent. My favorite is that we now correctly recognize a gpg installed by fink.
Here is how I invision this in the end. Assuming that AOL didn't use PGP (or GPG), then we (OS Client Authors) should try to support their protocol, along with PGP (or GPG) which would be considered more secure.
Backup: $40 online backup - does anyone do this? Whats wrong w/ burning your 'keeper' stuff to a CDR? You certainly wont have room to store your apps/OS, whats the point?Besides, to you want Apple in possession of your personal data - they have nosy admins also you know...
The online backup is for the trial users. If you had bothered to read the info, you would have discovered that full users (ones who paid) can use the backup to CD's or DVD's.
i didn't have to carry a backpack that obviously has a laptop in it (so it's less likely to be stolen).
Excelent point. That is why I got one that is made by Jansport and looks identical to my other Jansport backback that doesn't hold a laptop. Upon first seeing it, my friends all thought that I just shoved my computer into the backpack until they saw the extra little padded compartment inside.
The last thing I want is to carry around a 16.1" diagonal behemoth
Good point. I have a Titanium which has a 15.2 inch screen. It is a wider screen than most laptops so that would make it narrower than other 15" screens. Whenever I want to take my computer to campus (to use their bandwidth), I do noticed that the computer barely fits in the backpack which has one of those compartments for laptops. The velcro flap cannot fully close because the computer is too long. Get much bigger than this, and it will not fit in the backpack at all.
I also have an older laptop which is not quite as big but weighs >9 pounds. It is easier to carry around in the backpack because it fit in the compartment completely.
I would say that 15" is big enough for a portable computer until they figure out how to make rollup screens (like the globals on EFC). It is nice having a large screen on a laptop, but I prefer the portability. If I wanted a larger screen, I would buy one for home to connect to the computer and leave it there.
The testing doesn't look for holes in the operating system, but rather evaluates what security features are built into it.
I guess they needed this so that Windows could be used.
If that is the case, OS X should not have any trouble at all. Let's look at some of the security features:
Root disabled by default
SSH remote login
Telnet not easy to turn on (should use SSH instead)
Can disable auto login
Any major system changes require authentication even if the current user is an admin
Built in Firewall
I know this is a short list, but it demonstrates the point well. OS X has many security features that are inherant of a unix based OS. It will be nice to see OS X more accepted amoung the government.
But the new rendering engine, called Quartz Extreme, could tax some systems running older graphics cards. For optimum performance, the new engine requires at minimum an Nvidia GeForce2MX or greater graphics card or any 32MB ATI Radeon graphics card. The hefty graphics requirements could present problems for some Apple notebooks or older desktops.
Jobs clearly stated in his keynote that the new Quartz Extreme will run the same on the older hardware as the current Quartz. Only those with the required graphics hardware will see any change and it will be an improvement. Also, the 32MB of VRAM is not required, just recomended.
Kay also questioned how many Mac OS X 10.1 users would move to Jaguar
Simple: Almost all of them. I know I will!
Already, Apple is the first computer maker to offer wireless networking on all its systems.
Nice to see that they have gotten over Dell's lies.
Adding Bluetooth to Mac OS X would put Apple ahead of Microsoft, which has yet to incorporate support for the technology into Windows XP.
A sign of things to come. With OS X, Apple has the easiest development platform out there. Once it start to get fully ramped up (still getting there), development on OS X should take off without any end in sight.
The register has posted another article on this saga. Here we read what some readers have said. One points out the fact that both the present and future cash flows must be calculated when looking at whether such a venture is possible. A few suggest that those complaining should just buy windows or get linux instead. I was rather shocked to see one mention that MS Office doesn't support Unicode! If this is true, then this is in no way Apple's fault. Nice try MS. My favorite was the mention of Koffice supporting it. If the open source community can do it, why can't you MS, or are you too busy making claims that open source software compromises security?
As usual with all comments, there are some good ones, and there are some bad ones.
Well, I got the full screen version, and it said that it doesn't allow saving. So... I did a get info on the movie, and lo and behold, there is an entry called source which is a URL. I typed this into wget, and it is downloading. For those who want to save it, here is the akamai URL:
I assume that you are not talking about something to play QT files as it is already free. The pro, well, you can find key generators if you look.
Sorenson Pro for windows, well, I don't really know since I am a MacOS X user.
I find it interesting that you are looking for a free version of it after you said that it sucks. Oh well. Is it that you think that it is good enough if you can get it for free, only you can't justify shelling out money for it?
I perfectly understand the difference between a file format and a codec. But neither of the two that are related to QuickTime are published standards. Therefore, at least in my opinion, they suck. Both of them. THEREFORE, I can say that QuickTime files, as a whole, suck.
It is true that the codec is not a publish standard, but the file format is! I have read descriptions of the file format and C code that reads it, all from Apple's web site. Perhaps you should do some research before you go on an uninformed tirade.
I for one would prefer that Apple keeps to the spec. It maintains a better operating client because then they don't have to maintaing two branches of the file reading routines (one to handle the official, and one to handle the hacked). Also, with the DivX video codec, I imagine that it will not be included with QT because of its shady history. In its current incarnation, it has a chance if the lovely people at divx.com would get their act together and put out a decent version of the codec for QT.
If you want to play DivX WELL in MacOS X, get vlc. It plays VCD's, SVCD's, DVD, DivX, MPEG 1, MPEG 2, and supposedly MPEG 4. The MPEG 4 part is the only aspect I have yet to test in this client.
Apple has the problem of supporting stuff and at the same time keeping in good legal standing with other companies that might try to sue them. If they were to release a DivX codec, they might be walking a tightrope. Also, they may have code under a licence to read the avi file that doesn't allow them to modify to read VBR MP3's.
In short, put yourself in their place and try to figure out why they may not do something before you critisize them. I am not sure which country you are from, but in the US (where Apple is based) and most other countries, a companies primary goal is to maximize profit. I doubt that VBR MP3 in avi is a profitable venture for them, and so they are not likely to do it.
One problem, next time you launch IE, it becomes the default. Trust me, if it were this easy, I would not be editing the plist. Just IE set itself as the default every time it was launched.
Slashdot Mirror
I grew up in Los Alamos and I worked there during my high school years through some of graduate school. The article completely failed to mention one of the main culprits for a lot of these problems: The Department of Energy. While I do not have knowledge beyond what is in the press for most of the incidents mentioned, the ones where I do mostly include a major role in the problem played by DOE ranging from their screwed up policies to direct involvement. Given this, a new contractor can only do so much.
The Roku rates pretty low in terms of scriptability/hackabilty IMHO. Yes, it has an API, but through numerous poor choices on Roku's part, this is not as useful as other platforms. Their Brightscript is not well designed, and the VM is buggy. I can cause the device to hard reset with one line of code (an accidental discovery trying to make the thing work). The library availability for the language is poor. Heaven help you if you want to do something that's provided in libraries on the net in nearly every language but not available on Brightscript. If you want to do something very very simple, you can pull it off, but more complicated programs are an exercise in frustration. Additionally, their provided displays are missing key functionality. For example, the Grid Display cannot be stacked without crashing the device, you cannot set the selection on it before it is displayed, and there is no callback for the display of the grid so you can set the selection then. Also, use the debugger too much, and you'll lock up the device, requiring pulling the power to correct it. Lastly, you cannot have more than one development app on the device at a time.
I went to a GoogleTV, and it's hackability is vastly superior to the Roku. It uses a language which has a large library availability (java) and the programming environment is inside an IDE (Eclipse). The remote debugger is more useful, synchronizing the execution with the location in the code and displaying local variables, in stark contrast to using telnet on a port to step through which is essentially a stripped down command line gdb (which is missing many of the more useful commands, such as breakpoints).
I cannot speak for the other devices, but if they were to provide any API for third party development, I would be surprised if they were not superior to the Roku. It has a development environment you would expect from the previous decade, not this one.
> later when they upgraded the OS (without breaking any compatibility)
This caused me to laugh. I have a Roku 2 and experienced a regression on http live streaming resulting in audio dropouts, video freezing, and even the device rebooting. My dad has a Roku 1, and experienced regressions on two separate "updates." This is not isolated; just see the forums and the number of requests to revert the firmware and Roku's refusal to allow it.
It is being tested... The problem is that the changes rely on the ssl in the libcurl library that 10.1 does not have. Fire is maintaing compatibility with 10.1, proteus does not. Expect a release today.
For my senior design project, we had to make some dedicated hardware to exactly this. It was solved using an FFT inside a Vertex-II Pro FPGA using a sliding window.
Now, as far as this contest goes, this can be done quite easilly in a computer. In fact, we had made a bitwise simulation of the FPGA using matlab before doing any Verilog to make sure that our design worked. My favorite was the little setup that we did in LabView. Took about 30 minutes to make, and would solve this quite easilly. It was very nice to see the result in color intensity with frequency along the x axis, and time across the y. Seeing the data in there was quite easy even though the SNR was so low that you couldn't even hear most of the tones.
In short, if you want to do this, read the data into matlab (or octive), do an FFT on a small window of the data, slide the window over, repeat. Take all the FFT's and plot in an intenisty graph (not sure what the matlab term is, but I am sure it has such a thing).
Now I have a great prank to play on friends... I wonder how long it will take them to figure out how to fix it :)
It would also be a fun hobby and a great way for open source to venture into the hardware realm.
Open sources has extended into the hardware realm. I am sure there more sites, but I have found Open Cores to be quite helpful. They have all kinds of things, and some are GPL.
Correction... The ?64bits should be =64bits. Guess other OS's still have problems with other characters.
First of all, broken link. I assume you were refering to this. Now, you first mentioned 42bit VIRTUAL ADDRESSES. There is a BIG DIFFERENCE between virtual and real addresses.
There are three address types:
Effective: This is the address that the program sees. This may (and often does) have very little to do with the real address
Virtual: This is used only within the MMU and Page tables.
Physical (Real): This is the actual address in the memory.
Now, the link you provided states 42bits of real address. This does not mean that the virtual address is 42bits. The PEM is very acurate in these cases since it describes the page table setup that the OS has to take care of and maintain. The same file you reference states that the 970 has 64bit effective addresses. This is consistent with the PEM! Notice the PEM also states that the physical addresses are ?64bits. This is also consistent with the 970.
Lastly, 42bit real addressing is only a limit in the amount of physical ram that the chip can address. This is not a limit in the amount of total memory (physical and virtual). This means that your Ram, OS, and HD space is the only practical limit in allocating massive amounts of memory. Assuming you have enough HD to handle 2^64 bytes, then you can allocate that much. The chip is not a limiting factor. Also, if you look at the specifics of the chip's memory management, you will see that you can change it to use other PTEG's if you change some of the registers, so if you are willing to go through the context switch expense; 2^64 bytes is a limit per executable.
Actually, it is 52 bit for the older PowerPC's: Page 7-35 It is 64 or 80 for the 970: Page 258.
It is also important to note that these values are a per program basis if you are willing to change around some special registers in your context switches.
As a Fire developer myself, I thought that I could contribute a little more to this. We have started to participate in a discussion on the best way to do encryption over IM protocols. This discussion can be found here: http://www.chat.solidhouse.com/smsn/. The GAIM-E author has even contributed to this discussion.
Also, we have drastically improved the way that the GPG encryption is handled. It now works on more protocols and will be more consistent. My favorite is that we now correctly recognize a gpg installed by fink.
Here is how I invision this in the end. Assuming that AOL didn't use PGP (or GPG), then we (OS Client Authors) should try to support their protocol, along with PGP (or GPG) which would be considered more secure.
Glad to run across another satisfied Fire user.
Backup: $40 online backup - does anyone do this? Whats wrong w/ burning your 'keeper' stuff to a CDR? You certainly wont have room to store your apps/OS, whats the point?Besides, to you want Apple in possession of your personal data - they have nosy admins also you know...
The online backup is for the trial users. If you had bothered to read the info, you would have discovered that full users (ones who paid) can use the backup to CD's or DVD's.
Excelent point. That is why I got one that is made by Jansport and looks identical to my other Jansport backback that doesn't hold a laptop. Upon first seeing it, my friends all thought that I just shoved my computer into the backpack until they saw the extra little padded compartment inside.
Good point. I have a Titanium which has a 15.2 inch screen. It is a wider screen than most laptops so that would make it narrower than other 15" screens. Whenever I want to take my computer to campus (to use their bandwidth), I do noticed that the computer barely fits in the backpack which has one of those compartments for laptops. The velcro flap cannot fully close because the computer is too long. Get much bigger than this, and it will not fit in the backpack at all.
I also have an older laptop which is not quite as big but weighs >9 pounds. It is easier to carry around in the backpack because it fit in the compartment completely.
I would say that 15" is big enough for a portable computer until they figure out how to make rollup screens (like the globals on EFC). It is nice having a large screen on a laptop, but I prefer the portability. If I wanted a larger screen, I would buy one for home to connect to the computer and leave it there.
I guess they needed this so that Windows could be used.
If that is the case, OS X should not have any trouble at all. Let's look at some of the security features:
- Root disabled by default
- SSH remote login
- Telnet not easy to turn on (should use SSH instead)
- Can disable auto login
- Any major system changes require authentication even if the current user is an admin
- Built in Firewall
I know this is a short list, but it demonstrates the point well. OS X has many security features that are inherant of a unix based OS. It will be nice to see OS X more accepted amoung the government.But the new rendering engine, called Quartz Extreme, could tax some systems running older graphics cards. For optimum performance, the new engine requires at minimum an Nvidia GeForce2MX or greater graphics card or any 32MB ATI Radeon graphics card. The hefty graphics requirements could present problems for some Apple notebooks or older desktops.
Jobs clearly stated in his keynote that the new Quartz Extreme will run the same on the older hardware as the current Quartz. Only those with the required graphics hardware will see any change and it will be an improvement. Also, the 32MB of VRAM is not required, just recomended.
Kay also questioned how many Mac OS X 10.1 users would move to Jaguar
Simple: Almost all of them. I know I will!
Already, Apple is the first computer maker to offer wireless networking on all its systems.
Nice to see that they have gotten over Dell's lies.
Adding Bluetooth to Mac OS X would put Apple ahead of Microsoft, which has yet to incorporate support for the technology into Windows XP.
A sign of things to come. With OS X, Apple has the easiest development platform out there. Once it start to get fully ramped up (still getting there), development on OS X should take off without any end in sight.
The register has posted another article on this saga. Here we read what some readers have said. One points out the fact that both the present and future cash flows must be calculated when looking at whether such a venture is possible. A few suggest that those complaining should just buy windows or get linux instead. I was rather shocked to see one mention that MS Office doesn't support Unicode! If this is true, then this is in no way Apple's fault. Nice try MS. My favorite was the mention of Koffice supporting it. If the open source community can do it, why can't you MS, or are you too busy making claims that open source software compromises security?
As usual with all comments, there are some good ones, and there are some bad ones.
I am using QT 6 Public Preview. I just did a get info, and it showed the whole thing. I guess it is a QT6 thing then.
Well, I got the full screen version, and it said that it doesn't allow saving. So... I did a get info on the movie, and lo and behold, there is an entry called source which is a URL. I typed this into wget, and it is downloading. For those who want to save it, here is the akamai URL:
a / a1a1aaa2198c627970773d80669d84574a8d80d3cb12453c02 589f25382f668c9329e0375e81787e85abb28970c7aee1d900 465d736aa1c81ec4bb91e8bf063cc3f6ed037/star_trek_ne mesis_fs.l.mov.
http://a1536.g.akamai.net/5/1536/51/979a9ce9df5cd
Enjoy. BTW, if this goes 404, just follow the procedure that I use listed above.
I find it amusing that they use IIS for their main web site, but when they need bandwidth, they use apache. Interesting revelation here.
BTW, it is still forbidden. Guess they are not going to let us see the trailer after all.
I assume that you are not talking about something to play QT files as it is already free. The pro, well, you can find key generators if you look.
Sorenson Pro for windows, well, I don't really know since I am a MacOS X user.
I find it interesting that you are looking for a free version of it after you said that it sucks. Oh well. Is it that you think that it is good enough if you can get it for free, only you can't justify shelling out money for it?
I perfectly understand the difference between a file format and a codec. But neither of the two that are related to QuickTime are published standards. Therefore, at least in my opinion, they suck. Both of them. THEREFORE, I can say that QuickTime files, as a whole, suck.
It is true that the codec is not a publish standard, but the file format is! I have read descriptions of the file format and C code that reads it, all from Apple's web site. Perhaps you should do some research before you go on an uninformed tirade.
> Second, use native codecs when possible. MPEG-2, for example.
QT 6 final is supposed to have MPEG-2 encode and decode support. Supposedly, it barely missed being included in the preview. Just wait, it is comming.
Or... You could get the DivX codec (v. 5) from divx.com and then convert to MPEG 4 in QT.
Warning: it is Alpha, and I have not tested it on QT 6. I currently use vlc to play DivX and have not come across a file that it cannot play.
I for one would prefer that Apple keeps to the spec. It maintains a better operating client because then they don't have to maintaing two branches of the file reading routines (one to handle the official, and one to handle the hacked). Also, with the DivX video codec, I imagine that it will not be included with QT because of its shady history. In its current incarnation, it has a chance if the lovely people at divx.com would get their act together and put out a decent version of the codec for QT.
If you want to play DivX WELL in MacOS X, get vlc. It plays VCD's, SVCD's, DVD, DivX, MPEG 1, MPEG 2, and supposedly MPEG 4. The MPEG 4 part is the only aspect I have yet to test in this client.
Apple has the problem of supporting stuff and at the same time keeping in good legal standing with other companies that might try to sue them. If they were to release a DivX codec, they might be walking a tightrope. Also, they may have code under a licence to read the avi file that doesn't allow them to modify to read VBR MP3's.
In short, put yourself in their place and try to figure out why they may not do something before you critisize them. I am not sure which country you are from, but in the US (where Apple is based) and most other countries, a companies primary goal is to maximize profit. I doubt that VBR MP3 in avi is a profitable venture for them, and so they are not likely to do it.
One problem, next time you launch IE, it becomes the default. Trust me, if it were this easy, I would not be editing the plist. Just IE set itself as the default every time it was launched.