How To Make An Apple Tablet Today
There is a way to make an Apple tablet today.
The Basilesk Macintosh emulator works on a standard Microsoft tablet PC. All the TPC features work properly, with one exception noted below.
I have a Motion Computing TPC with a 12" display, 40 GB drive, wireless, 768KB of memory, running on an 866MHz Pentium IV processor. Using Basilesk, I an running an emulation of a Macintosh Quadra using Macintosh OS 8.1. The original Quadra had a 68040 processor. The emulator operates as though the Quadra had a clock rate of about 400 MHz - it is reasonably fast.
I'm able to run many Macintosh applications. I'm running MS Word 6.1, Excel, Mac Draw, Word Perfect, Claris Works, More 3.1, and a number of other Mac apps without problems.
The Basilesk emulation will accept input from the TPC pen when I use the on-screen keyboard. I'm also able to use hand writing recognition in the foreground while the Mac runs in the background. The only thing I haven't been able to do is transfer recognized handwritten text from the TPC into the Mac apps. The TPC recognizes and converts the handwriting and places it on the preview window, but when you tell the TPC to send the text into the Mac app, the text disappears. I suspect there is a simple tweak to Basilesk to fix this. But the on-screen keyboard works 100%.
This isn't as good as having a tablet with OS X and a PwerPC, but it works pretty well. Personnaly I'd rather use Mac OS 8.1 and the Mac applications than Windows XP and any of the newer PC apps.
The Mac emulator works so well that I pretty much use it for all work on the TPC. Once you start the emulator in full-screen mode, it's just like having a Mac OS 8.1 tablet.
There seems to be a new PowerPC emulator being developed for use on Linux-based computers. I know that many folks are running Linux on Microsoft architecture TPCs. When the PPC emulator is mature I'm going to see if I can use it and run OS 9.
Having used this for several months, I can only say that I wish Apple would get into the tablet business. At the very least, I wish someone would port Mac OS 9.1 to the TPC architecture.
The Basilesk Macintosh emulator works on a standard Microsoft tablet PC. All the TPC features work properly, with one exception noted below.
I have a Motion Computing TPC with a 12" display, 40 GB drive, wireless, 768KB of memory, running on an 866MHz Pentium IV processor. Using Basilesk, I an running an emulation of a Macintosh Quadra using Macintosh OS 8.1. The original Quadra had a 68040 processor. The emulator operates as though the Quadra had a clock rate of about 400 MHz - it is reasonably fast.
I'm able to run many Macintosh applications. I'm running MS Word 6.1, Excel, Mac Draw, Word Perfect, Claris Works, More 3.1, and a number of other Mac apps without problems.
The Basilesk emulation will accept input from the TPC pen when I use the on-screen keyboard. I'm also able to use hand writing recognition in the foreground while the Mac runs in the background. The only thing I haven't been able to do is transfer recognized handwritten text from the TPC into the Mac apps. The TPC recognizes and converts the handwriting and places it on the preview window, but when you tell the TPC to send the text into the Mac app, the text disappears. I suspect there is a simple tweak to Basilesk to fix this. But the on-screen keyboard works 100%.
This isn't as good as having a tablet with OS X and a PwerPC, but it works pretty well. Personnaly I'd rather use Mac OS 8.1 and the Mac applications than Windows XP and any of the newer PC apps.
The Mac emulator works so well that I pretty much use it for all work on the TPC. Once you start the emulator in full-screen mode, it's just like having a Mac OS 8.1 tablet.
There seems to be a new PowerPC emulator being developed for use on Linux-based computers. I know that many folks are running Linux on Microsoft architecture TPCs. When the PPC emulator is mature I'm going to see if I can use it and run OS 9.
Having used this for several months, I can only say that I wish Apple would get into the tablet business. At the very least, I wish someone would port Mac OS 9.1 to the TPC architecture.
Comments
I'm interested less in a Tablet and more in knowing if they're even necessary.
Originally posted by bunge
There seems to be a new PowerPC emulator being developed for use on Linux-based computers. I know that many folks are running Linux on Microsoft architecture TPCs. When the PPC emulator is mature I'm going to see if I can use it and run OS
Any links? I could have sworn that people here had said that either it was virtually impossible, or it would run so slow it woudn't even be worth it. Do to the fact there is such a large registor count on the Mac vs the PC.
Originally posted by bunge
Just out of curiosity, what do you use your tablet for? Is it a complete desktop replacement for you or just an extension of your normal machine? Or is it basically just a big Palm with more features?
I'm interested less in a Tablet and more in knowing if they're even necessary.
I would say the tablet has about as much use as a Palm. You think you need it, and you think you look cool because you have it. But once you get it, you realise you really didn't need it and it just becomes another "toy".
Originally posted by kupan787
I would say the tablet has about as much use as a Palm. You think you need it, and you think you look cool because you have it. But once you get it, you realise you really didn't need it and it just becomes another "toy".
This is kind of my thought as well, but that's why I'm asking this user. I'm curious to know how useful they actually can be.
I use the tablet primarily for work. I'm a research scientist, and the large majority of my work is writing and reviewing technical papers that are a mix of text, equations, and line drawings of computer hardware and software architectures. Most of my work from the last 25 years was originally done on Macintoshes, and I need access to that material.
I've found that the primary advantage of the tablet is that it frees me from being anchored to a desk when I want to work. I carry it with me, and when I have time between meetings or classes I can find a quiet place to read papers I've down-loaded from the web, or continue work on a paper under development. I find I'm getting more work done because I can make use of chunks of time that would otherwise be wasted. The opportunity to have all my notes and data in one place is a great help. The ability to work without a keyboard makes it much easier to find a place to work.
I have on my desk a G5 with a 20" Apple cinema display. It is a fairly high-end system with 2 GB of memory, 250GB of disk space, a scanner, a video capture system, external Firewire drives, and other. But I use it only for image processing and video editing. My main work machine is the tablet, even when I'm in the office, primarily because I can't affor dthe time to try to resynchronize everything between the two machines - i just leave the importnat stuff on the TPC.
I've had Palm and Pocket PC PDAs. I found they were useful for holding Outlook schedules and simple email, but they are not tools for serious work.
I haven't run the PPC emulator so I can't speak for it's speed. It's discussed in the Delphi forums. I'd be surprised if it wasn't about as fast emulating a PPC on an x86 architceture as Virtual PC is at emulating an 80586 on a PPC.
I'm a computer architect by profession. Don't believe everything you hear about the difficulties in emulating a PPC on an x86. There is a lot of uninformed speculation floating around on that topic.
Originally posted by illuminati18
The tablet has pretty much replaced my desk top except for video editting.
I use the tablet primarily for work. I'm a research scientist, and the large majority of my work is writing and reviewing technical papers that are a mix of text, equations, and line drawings of computer hardware and software architectures. Most of my work from the last 25 years was originally done on Macintoshes, and I need access to that material.
I could see a tablet coming very much in use for this. Especially writing equations out.
I carry it with me, and when I have time between meetings or classes I can find a quiet place to read papers I've down-loaded from the web, or continue work on a paper under development.
This I can't imagine. I type way faster then I could ever write (maybe you don't?). So I can't image holding a tablet (or sitting it on a desk) and trying to "write in" a paper I am working on. It seems a laptop would be much better served No table, no problem. Plus, with a keyboard, a lot more coudl be entered at a much faster rate.
I haven't run the PPC emulator so I can't speak for it's speed. It's discussed in the Delphi forums. I'd be surprised if it wasn't about as fast emulating a PPC on an x86 architceture as Virtual PC is at emulating an 80586 on a PPC.
There has been much talk abotu PPC emulators for the x86 camp for years. I woudl check out the Delphi forums, but you need to register, and I really don't feel like taking the time to do so right now.
I believe anything can be done, it just comes down to a matter of useability. I wish I could find some of the threads (might have been over at macnn...), but some technical people were exaplining the why, and I could have sworn the biggest issue was registers. Since the PPC has a multitude more registers than an x86 processor, the x86 would need to emulate those other registers in main memory. This would be a huge speed hit, as a register is magnitudes faster than main memory is.
And if all this about VPC and the G5 is true, then the reverse would be true for a P4, haveing to do the byte swapping for endian issues. If it will slow down the G5, then the P4 woudl see a very similar slow down.
I'm a computer architect by profession. Don't believe everything you hear about the difficulties in emulating a PPC on an x86. There is a lot of uninformed speculation floating around on that topic.
I would love to hear why you think it will be much easier than what I have come to understand.
[EDIT] ok I found this http://www.kearney.net/~mhoffman/basiliskII/bauer2.html
Here is one interesting bit:
OSEH: There have been many Macintosh PowerPC emulation promises, but none of them ever came to fruition; e.g., SoftMac from Emulators Online (http://www.emulators.com), eMac, and Fusion (http://www.microcode-solutions.com). What is your opinion of these offerings? Since the promise of PowerMac emulation has been around for such a long time with no real concrete products materializing, can you tell us what is so difficult about writing a PowerMac emulator?
CB: Maybe the release of MacOS X has slowed things down. Getting PPC MacOS 9.x and earlier versions to work is not much different from what is done in 68k Mac emulators. MacOS X is a different story. Since it is Unix it requires a more complete "virtual machine" implementation which is more difficult than the hacks which are sufficient for 68k MacOS.
No, I don't have plans for making it run MacOS X, either. :-)
Picture it this way: at the end of each clock cycle in a computer the complete contents of the memory of the computer (including its registers) can be viewed as a data set representing the state of that machine. in the next clock cycle, some of the contents of that data set are transformed into new values based on whatever instruction the processor performed. The memory contents of the computer after that clock cycle are a new data set. A famous software architect named Hoare represented it this way:
[M(n)] P [M(n+1)]
where M(n) and M(n+1) are the memory contents before and after the clock cycle, and P is the processor instruction.
There is no requirement that the emulator mimic the register contents, or any other feature of the emulated processor, while transforming M(n) to M(n+1). All that matters is that, in the emulation, the new data set M(n+1) accurately reflect the effect of the operation P. You (the programmer) can perform P in software anyway that you like, and the emulation will be correct as long as M(n+1) is what the emulated processor would have calculated in the real computer that is being simulated.
When you write an emulator you write code that simulates transforming the input data M(n) to the data M(n+1). In doing this, there is no need to establish a 1:1 corrspondence between the native processor registoers (or other resources) and the emulated processor resources. You simpy write the code to perform M(n)-> M(n+1).
When you compile your emulator the compiler for the native processor will try to optimize the code for the resources of that processor. If you examine the object code you'll find that in many cases a good compiler will assign native processor registers to data that represents emulated registers. But this is an artifact of the compiler doing what is called an "isomorphic mapping" of the native processor architecture to the emulated machine architecture. It is not something that the programmer tries to explicitely control.
OS X is based on a form of Unix derived from the original Mach operating system. I have worked with the virtual machine for the Mach OS and MIPS processors, and it isn't any more (or less) complicated than the virtual machine for Windows/x86.
There is one exception to the M(n) P M(n+1) description above. Early processors allowed asynchronous modifications of the memory space based on input from external devices. As far as I know, no modern processor allows this to happen. The PPC, the x86, and other processors force synchornization of input for reliability purposes, and Hoare's method works for those machines.
Sorry for the length of this, but some of the rumors of the difficulties of emulating the PPC have not been well-founded in science. The PPC isn't any more difficult to emulate than any other modern processor with pipellining and a large internal memory set. It isn't any harder to emulate a PPC than it is to emulate an x86. The methods used in Virtual PC easily apply to the PPC emulators.
Originally posted by illuminati18
Sorry for the length of this, but some of the rumors of the difficulties of emulating the PPC have not been well-founded in science. The PPC isn't any more difficult to emulate than any other modern processor with pipellining and a large internal memory set. It isn't any harder to emulate a PPC than it is to emulate an x86. The methods used in Virtual PC easily apply to the PPC emulators.
My question then becomes (and follow the line of questionings asked of the BasiliskII emulator) why have a number of groups started projects, but none have barred fruit? If, as you put it, it is just as easy to emulate the PPC as it is to emulate the x86, shouldn't something have been put out by now (seeing as how the PPC has been around for quite awhile now). The x86 emulator space has seen quite a few programs (SoftWindows, VPC, Blue Label, Bochs, and probably more...) Some were open source, some sucked, and a small number have succeed and are still around today.
I would think that if a PPC emulator was of similar difficuilty to making an x86 one, then there would be something out there.
Microsoft is able throw a lot of resources at VPC because it is a profitable business. To them it isn't a technical issue, but a business case.
The Mac emulators have largely been shareware. The folks who developed Basilesk did an admirable job and were technically excellent. It doesn't appear that anyone made money on the 68K emulators, and conseequently there hasn't been any financial incentive to develop a Mac PPC emulator. The folks who knew how to do emulation well seem to have moved on and their interests now lie elsewhere. There just doesn't seem to be a market for a PPC/OS X emulator.
As best I can tell from the few web sites that discuss Mac emulation, the handful of companies that claimed to be working on PPC emulation were under-capitalized and short of technical staff.
In my previous note I didn't mean to imply that emulation is easy. All emulators are hard to develop and require excellent technical staff to succeed. It's just that all of them are equally hard, and most are not fully developed unless there is a market for them (like PPC).
Originally posted by illuminati18
My personal opinion is that it is just a question of resources.
Microsoft is able throw a lot of resources at VPC because it is a profitable business. To them it isn't a technical issue, but a business case.
Microsoft just recently got VPC, they have yet to make any releases of VPC (besides the rebranding update), and didnt put an ounce of time into it in the past 10 years. If I am remembering things correctly, only two of the x86 emulators have been done by "big" companies (SoftWindows and VPC). Blue Label, PCx, and DOSbox are either shareware or freeware. Bochs was opensource, and I believe started as a UNIX project, backed by no large company at all.
The Mac emulators have largely been shareware. The folks who developed Basilesk did an admirable job and were technically excellent. It doesn't appear that anyone made money on the 68K emulators, and conseequently there hasn't been any financial incentive to develop a Mac PPC emulator. The folks who knew how to do emulation well seem to have moved on and their interests now lie elsewhere. There just doesn't seem to be a market for a PPC/OS X emulator.
Hardly anyone makes money on the multitude of console emulators out there, yet they are still made. And they are made for such obscure systems. I would bet the "interest" for a PPC emulator is much greater than there ever was for half the emulator projects ever attempted.
I would love to see what, if anything, Programmer could contribute to this conversation.
As best I can tell from the few web sites that discuss Mac emulation, the handful of companies that claimed to be working on PPC emulation were under-capitalized and short of technical staff.
In my previous note I didn't mean to imply that emulation is easy. All emulators are hard to develop and require excellent technical staff to succeed. It's just that all of them are equally hard, and most are not fully developed unless there is a market for them (like PPC).
I know of emulation projects done by one man shows. Not to trivialize it, emulating a system takes a lot of technical know how. But I think the reason those PPC emulation sites have failed is do to more that being "under-capitalized and short of technical staff". There is a lot more info out there about the PPC than things like the N64 or PSX, yet those emulators were able to be made (and made quite well).