FreeMWare (ok, plex86 now) is not anymore an emulator than VMWare. It works by using the actual processor, just upping the security on the processor so that things stay controlled, and the other program/OS 'thinks' it has full control of the system. It is dosemu emulation, not snes9x emulation. So getting it to work on another platform is hard, unless that platform handles the x86 instruction set
I agree about the mouse buttons.. and I must add I don't see the point of newer, five-button mice.. should I start binding buttons to keys so I don't need to put my right hand on the keyboard anymore?
The extension manager is cool but it is required, because extensions break so many things. The majority of them (in my experience) are just patches to the toolkit, the interrupt table MacOS calls its system api. load about 30 extensions and I guarantee you won't have five days of uptime.
Most of my bad macintosh experience is through netscape.. I routinely couldn't make it through loading netscape without locking the machine. It took me a long time to realize this was Apple's fault, not Netscape's, because no application, even if it is written 'wrong' like Netscape apparently was, should cause a machine to deadlock like that.
I have heard this statement before, and I don't quite buy it. Java is not a platform in itself but is an abstraction for the underlying platform. Sure you have to port to the java platform, this is because all the OSes out there are implemented differently.
The main difference is in java, the 'core' is not as bad as writing for a specific OS, because it will run on any operating system. The main problem with java compatibility comes from non-pure code, which must be ported to every platform in order to work.
java has references, which is different.. they are equiv. to handles (pointer pointers) where the middle pointer is a table to the location of the object - so things can be moved around without the actual pointer value in the program changing.
hotspot and the IBM java both use generational garbage collectors, and use real pointers - the code will scan the stack and heap space for the old values and change them to the new values.
The point basically being, that while pointers are used to implement java, they are not available really for programmer use, just references are (and in java, everything that isn't a native type is a reference to an object)
the problem with vectors is that they really don't work well for small objects like icons. You'll notice that the interface for the MacOSX uses roster for all its icons and toolbars (although it is cool enough to have these be bilinear textures, so that they still scale)
assuming DVD (with hypothetical, less-laughable CSS2)->ASIC(on video card)->Monitor, yes.
There are problems with this though. First off, the ASIC will have to contain enough memory to both completely decrypt and then decode the MPEG-2 stream, then to stream it out (well, probably block-encrypted) to the monitor. If you do this, the encrypted data could only really be viewed fullscreen.
Also if the memory for decryption was not stored outside the chip, you could just set it up to copy out the pre-encoded frames from video memory. Or even better, just the same with the original MPEG2.
I think the goal for this would be set-top boxes, where you don't have an underlying OS to use to snoop. But unless digital TV comes out and intel supplies the chips, there will never be enough consumers 'duped' into buying televisions with this technology for it to be really used.
1. Serial ATA is an internal standard, Firewire is an external standard (although it can be used internally). Explain to me why Apple includes ATA drives inside the G4 when it has an internal Firewire port?
2a. You do not need topology internal to the computer. So hubs and daisychaining is meaningless.
2b. Since four ports of Serial ATA could technically operate in parallel (the bandwidth is not shared), four devices have a combined available bandwidth of 600 MB/s, as opposed to firewire's 400 Mbps (~40 MB/s)
2c. The max bandwidth of firewire, 3.2 Gbps, is ~320 MB/s, making it about the same speed as two Serial ATA ports running at the initial speed (which is not the maximum bandwidth)
3. Serial ATA is a point to point protocol. There is no master.
4. Serial ATA is already supported under both windows and linux, past the extent of IEEE1394. This is because Serial ATA uses the same protocol as regular ATA, and thus works exactly the same on an OS level. Unless you look at the chipset, the OS probably will not know it isn't using parallel ATA.
5. ATA is a cheaper solution than Firewire. Currently all firewire drives are actually ATA drives with a ATA<->Firewire chip inside, so Serial ATA reduces their cost, as well as requires a smaller engineering cost to change the chips on the drive to be native Serial ATA.
6. <i>What should be done? IMNSHO, they should scrap the proprietary Serial-ATA interface, and adopt 1394 as the official Serial ATA standard.</i>
What makes you think that Serial ATA is a proprietary interface?? Why would it be any less of an open standard than Parallel ATA? At least you don't need to pay licensing fees for ATA.
The blackdown release is still on Release Candidate 4 as of two hours ago. This is a joint venture between Sun and Borland/Imprise, mostly because Borland wants to have a high-quality JBuilder for linux.
IBM is a very nice company. I really hope they support Java 2 sometime soon, but to be honest I expect IBM's performance to drop some when they go up. Java 2 is just a heck of a lot more complicated in the VM, with things like weak references and the like now.
Much more like if you wrote binary code that would monitor your application's own usage, and dynamically relink code while running to inline functions, remove impossible-to-reach code blocks, etc.
No, it can't modify bitmaps with vectors. Think about it for a second, that is a type of Optical Recognition.
Apple's present PR engine's mantra is that Icons are represented up to 128x128, but by a bilinearly mipmapped graphic.. I.e. there is also a 64x64, 32x32, 16x16 (And possibly lower powers of 2) version, and it chooses and scales an icon based on these numbers and the closest version of the icon size-wise. No vectors. You really need 64x64 icons in order for vector icons to work, in my experience. Scaling lower than that just gives a mess.
I think people who don't like linux because of the lack of an 'established' support structure are definately missing the point. While the point that "if something breaks with Linux you have no-one to blame" is true, the question is - are you looking for someone to blame or a fix to your problem?
Read the 'rights' you have left over after the EULA gets done with you. The only thing you can do when Windows breaks is 'blame' microsoft.
So stand there pointing your finger at them, I'll be looking at something else that breaks with linux and know rather than having someone to 'blame', I have both the maintainers of the software to fix it (As I would with Microsoft) and also have the ability to contract someone out or fix it in-house, since the source code is open.
If you want someone to 'blame', get a support agreement. If you want to be capable of taking control of getting a problem fixed (as I imagine most people who's business and livelyhood comes from the continual operation of servers), take something with source code available. If you have a database break, it doesn't matter if 'a bunch of geeks from Berkeley' or 'a bunch of geeks working at Oracle' wrote it, unless you have a service agreement or some sort of liability from them, you are equally screwed.
And for gods sake, don't trust your livelyhood to an OS like Windows 2000. I won't install it on my home machine because of its stability, why in hell would I ever trust a server that has to actually stay up more than two hours to it?
hopefully you aren't using the drivers that came with win2k, because they offer no OGL or D3D support (due to them deciding at the absolute last minute to upgrade to DirectX 7).
Is anyone else worried about them upgrading to DirectX so late in the development cycle? There was practically no testing because the final DX7 didn't make it in until RC2, which is also why the few drivers that were in were ripped out for the release. Of course video stuff is the only subsystem that runs at kernel level, so I can see why they wouldn't want to test something that vital..:P
there are supposedly methods for altering hardware that has been released (both readers and writers) to let you read/write the 'protected' disk key and title key regions.
If Diamond indeed ships out a Webpad running Mobile Linux (their page lacks any press releases at this point), I think they surely deserve some sort of Hypocracy Award, if not an annual award than a 'Lifetime Acceivement' award. It would be truely ironic for a company to ship a webpad that only runs linux, after years and years of outright denying specs for things as trivial as the Rio uploading code (claiming it was "valuable intellectual property") to things like specs/help on video card drivers (based on S3 chips, the big help was needed because Diamond deviated in such fscked-up ways from the reference design).
Yes, Diamond deserves a nice trophy anyways, but if they really plan to ship an entire computer now (mobile or not) running the linux operating sytem when they have yet to support a single project under linux.. well, let me just say it is too early in the morning for me to fully imagine what the figurine at the top of the trophy should look like, and what it will be doing;-)
there was actual a full private-eye mystery novel that was published as a palindrome (although, just letters, not capitalization or spacing). I forget how many pages long it was, I know it was either over 100 or over 200. Not the most incredible reading from what I understand, though:)
the main (only?) disadvantage to a microkernel OS like Hurd is that in order to have all drivers be servers, you have a lot more Interprocess Communication. In general, this is a huge speed hit-things like video card drivers have go between process spaces twice just to draw a pixel (line, if you are talking accellerated cards). This is actually why between NT 3.51 and NT 4.0, Microsoft moved the video card drivers into the kernel - they went from one kernel and one IPC call to just a kernel call, sacrificing speed for a big video performance boost.
Of course, the ideal thing would be to figure out how to make IPC and context switches 'free' (i.e. very low cost as far as CPU time go), probably at the memory management and processor level. If you could cache say, three process structures in backup registers on a chip, you could potentially get something like a microkernel for no added cost CPU-time wise. That and the fast context switching of Sun's MAJC chip would probably completely change computer OS design.
Of course, after you do that, you are pretty close to just building memory management and task switching directly into the CPU - you are one step away from not needing an OS at all;-)
strange thing - I installed 3.9.17 this last weekend... where are these DRI drivers nvidia is talking about? shouldn't they be in XFree 4.0, thats what they are waiting for, isn't it? And I thought that 3.9.17 was going to be the last (or next to last) snapshot before 4.0
it should be no faster, whether you use a TNT or a GeForce. It does not utilize features of the card (like DMA transfer or GART texturing) so it will still be very driver limited.
The two new features appear to be 32 bpp support and improved texture management.
they haven't released their API or their source. There is a API similar to the one they use internally that they post on their website, it has no more features than the GLX driver as far as 3d accelleration/card access (no gart, no dma).
nvidia's GLX driver uses no low level API, it uses the card registers directly. The ideal solutions would be to release their low-level API of course, and let it be ported to linux (or at least publicly say they are making it XP for later release). If they had this library released, drivers would come out for new cards literally days from the release of a new API version.
Slashdot Mirror
FreeMWare (ok, plex86 now) is not anymore an emulator than VMWare. It works by using the actual processor, just upping the security on the processor so that things stay controlled, and the other program/OS 'thinks' it has full control of the system. It is dosemu emulation, not snes9x emulation. So getting it to work on another platform is hard, unless that platform handles the x86 instruction set
I agree about the mouse buttons.. and I must add I don't see the point of newer, five-button mice.. should I start binding buttons to keys so I don't need to put my right hand on the keyboard anymore?
The extension manager is cool but it is required, because extensions break so many things. The majority of them (in my experience) are just patches to the toolkit, the interrupt table MacOS calls its system api. load about 30 extensions and I guarantee you won't have five days of uptime.
Most of my bad macintosh experience is through netscape.. I routinely couldn't make it through loading netscape without locking the machine. It took me a long time to realize this was Apple's fault, not Netscape's, because no application, even if it is written 'wrong' like Netscape apparently was, should cause a machine to deadlock like that.
I'm looking forward to OS X
hard to tell - the Strongarm processor supports running both as a slave (device) and a host controller.
:)
maybe both?
I have heard this statement before, and I don't quite buy it. Java is not a platform in itself but is an abstraction for the underlying platform. Sure you have to port to the java platform, this is because all the OSes out there are implemented differently.
The main difference is in java, the 'core' is not as bad as writing for a specific OS, because it will run on any operating system. The main problem with java compatibility comes from non-pure code, which must be ported to every platform in order to work.
java has references, which is different.. they are equiv. to handles (pointer pointers) where the middle pointer is a table to the location of the object - so things can be moved around without the actual pointer value in the program changing.
hotspot and the IBM java both use generational garbage collectors, and use real pointers - the code will scan the stack and heap space for the old values and change them to the new values.
The point basically being, that while pointers are used to implement java, they are not available really for programmer use, just references are (and in java, everything that isn't a native type is a reference to an object)
the problem with vectors is that they really don't work well for small objects like icons. You'll notice that the interface for the MacOSX uses roster for all its icons and toolbars (although it is cool enough to have these be bilinear textures, so that they still scale)
please explain in what ways. Utah-GLX already has its own direct rendering code.
assuming DVD (with hypothetical, less-laughable CSS2)->ASIC(on video card)->Monitor, yes.
There are problems with this though. First off, the ASIC will have to contain enough memory to both completely decrypt and then decode the MPEG-2 stream, then to stream it out (well, probably block-encrypted) to the monitor.
If you do this, the encrypted data could only really be viewed fullscreen.
Also if the memory for decryption was not stored outside the chip, you could just set it up to copy out the pre-encoded frames from video memory. Or even better, just the same with the original MPEG2.
I think the goal for this would be set-top boxes, where you don't have an underlying OS to use to snoop. But unless digital TV comes out and intel supplies the chips, there will never be enough consumers 'duped' into buying televisions with this technology for it to be really used.
1. Serial ATA is an internal standard, Firewire is an external standard (although it can be used internally). Explain to me why Apple includes ATA drives inside the G4 when it has an internal Firewire port?
2a. You do not need topology internal to the computer. So hubs and daisychaining is meaningless.
2b. Since four ports of Serial ATA could technically operate in parallel (the bandwidth is not shared), four devices have a combined available bandwidth of 600 MB/s, as opposed to firewire's 400 Mbps (~40 MB/s)
2c. The max bandwidth of firewire, 3.2 Gbps, is ~320 MB/s, making it about the same speed as two Serial ATA ports running at the initial speed (which is not the maximum bandwidth)
3. Serial ATA is a point to point protocol. There is no master.
4. Serial ATA is already supported under both windows and linux, past the extent of IEEE1394. This is because Serial ATA uses the same protocol as regular ATA, and thus works exactly the same on an OS level. Unless you look at the chipset, the OS probably will not know it isn't using parallel ATA.
5. ATA is a cheaper solution than Firewire. Currently all firewire drives are actually ATA drives with a ATA<->Firewire chip inside, so Serial ATA reduces their cost, as well as requires a smaller engineering cost to change the chips on the drive to be native Serial ATA.
6. <i>What should be done? IMNSHO, they should scrap the proprietary Serial-ATA interface, and adopt 1394 as the official Serial ATA standard.</i>
What makes you think that Serial ATA is a proprietary interface?? Why would it be any less of an open standard than Parallel ATA? At least you don't need to pay licensing fees for ATA.
This is my opinion, moderate down accordingly
The blackdown release is still on Release Candidate 4 as of two hours ago. This is a joint venture between Sun and Borland/Imprise, mostly because Borland wants to have a high-quality JBuilder for linux.
IBM is a very nice company. I really hope they support Java 2 sometime soon, but to be honest I expect IBM's performance to drop some when they go up. Java 2 is just a heck of a lot more complicated in the VM, with things like weak references and the like now.
Much more like if you wrote binary code that would monitor your application's own usage, and dynamically relink code while running to inline functions, remove impossible-to-reach code blocks, etc.
oooh.. but the article is also in HTML, the delivery method of said security problem. Post this on every article!
haha..
why can't I have moderator points on the days I want them?
No, it can't modify bitmaps with vectors. Think about it for a second, that is a type of Optical Recognition.
Apple's present PR engine's mantra is that Icons are represented up to 128x128, but by a bilinearly mipmapped graphic.. I.e. there is also a 64x64, 32x32, 16x16 (And possibly lower powers of 2) version, and it chooses and scales an icon based on these numbers and the closest version of the icon size-wise. No vectors. You really need 64x64 icons in order for vector icons to work, in my experience. Scaling lower than that just gives a mess.
DOS attack? Well thats what they get for running a site that big on top of DOS.. I hope they didn' t use MS-DOS for it..
Maybe now they will upgrade to a more recent OS maybe one where the TCPIP stack isn't a TSR.
I think people who don't like linux because of the lack of an 'established' support structure are definately missing the point. While the point that "if something breaks with Linux you have no-one to blame" is true, the question is - are you looking for someone to blame or a fix to your problem?
Read the 'rights' you have left over after the EULA gets done with you. The only thing you can do when Windows breaks is 'blame' microsoft.
So stand there pointing your finger at them, I'll be looking at something else that breaks with linux and know rather than having someone to 'blame', I have both the maintainers of the software to fix it (As I would with Microsoft) and also have the ability to contract someone out or fix it in-house, since the source code is open.
If you want someone to 'blame', get a support agreement. If you want to be capable of taking control of getting a problem fixed (as I imagine most people who's business and livelyhood comes from the continual operation of servers), take something with source code available. If you have a database break, it doesn't matter if 'a bunch of geeks from Berkeley' or 'a bunch of geeks working at Oracle' wrote it, unless you have a service agreement or some sort of liability from them, you are equally screwed.
And for gods sake, don't trust your livelyhood to an OS like Windows 2000. I won't install it on my home machine because of its stability, why in hell would I ever trust a server that has to actually stay up more than two hours to it?
hopefully you aren't using the drivers that came with win2k, because they offer no OGL or D3D support (due to them deciding at the absolute last minute to upgrade to DirectX 7).
:P
Is anyone else worried about them upgrading to DirectX so late in the development cycle? There was practically no testing because the final DX7 didn't make it in until RC2, which is also why the few drivers that were in were ripped out for the release. Of course video stuff is the only subsystem that runs at kernel level, so I can see why they wouldn't want to test something that vital..
there are supposedly methods for altering hardware that has been released (both readers and writers) to let you read/write the 'protected' disk key and title key regions.
If Diamond indeed ships out a Webpad running Mobile Linux (their page lacks any press releases at this point), I think they surely deserve some sort of Hypocracy Award, if not an annual award than a 'Lifetime Acceivement' award. It would be truely ironic for a company to ship a webpad that only runs linux, after years and years of outright denying specs for things as trivial as the Rio uploading code (claiming it was "valuable intellectual property") to things like specs/help on video card drivers (based on S3 chips, the big help was needed because Diamond deviated in such fscked-up ways from the reference design).
;-)
Yes, Diamond deserves a nice trophy anyways, but if they really plan to ship an entire computer now (mobile or not) running the linux operating sytem when they have yet to support a single project under linux.. well, let me just say it is too early in the morning for me to fully imagine what the figurine at the top of the trophy should look like, and what it will be doing
over 100 million people computers in the US use DOS. They just call it windows 95 or windows 98.
there was actual a full private-eye mystery novel that was published as a palindrome (although, just letters, not capitalization or spacing). I forget how many pages long it was, I know it was either over 100 or over 200. Not the most incredible reading from what I understand, though :)
the main (only?) disadvantage to a microkernel OS like Hurd is that in order to have all drivers be servers, you have a lot more Interprocess Communication. In general, this is a huge speed hit-things like video card drivers have go between process spaces twice just to draw a pixel (line, if you are talking accellerated cards). This is actually why between NT 3.51 and NT 4.0, Microsoft moved the video card drivers into the kernel - they went from one kernel and one IPC call to just a kernel call, sacrificing speed for a big video performance boost.
;-)
Of course, the ideal thing would be to figure out how to make IPC and context switches 'free' (i.e. very low cost as far as CPU time go), probably at the memory management and processor level. If you could cache say, three process structures in backup registers on a chip, you could potentially get something like a microkernel for no added cost CPU-time wise. That and the fast context switching of Sun's MAJC chip would probably completely change computer OS design.
Of course, after you do that, you are pretty close to just building memory management and task switching directly into the CPU - you are one step away from not needing an OS at all
strange thing - I installed 3.9.17 this last weekend... where are these DRI drivers nvidia is talking about? shouldn't they be in XFree 4.0, thats what they are waiting for, isn't it? And I thought that 3.9.17 was going to be the last (or next to last) snapshot before 4.0
I don't think they are written yet.
it should be no faster, whether you use a TNT or a GeForce. It does not utilize features of the card (like DMA transfer or GART texturing) so it will still be very driver limited.
The two new features appear to be 32 bpp support and improved texture management.
they haven't released their API or their source. There is a API similar to the one they use internally that they post on their website, it has no more features than the GLX driver as far as 3d accelleration/card access (no gart, no dma).
nvidia's GLX driver uses no low level API, it uses the card registers directly. The ideal solutions would be to release their low-level API of course, and let it be ported to linux (or at least publicly say they are making it XP for later release). If they had this library released, drivers would come out for new cards literally days from the release of a new API version.