The Newton eMate, (very) dead technology from Apple, is available for cheap on eBay. It's very rugged, having been designed for kids. It has a built-in keyboard and no disk drive (FLASH RAM instead). It runs for hours and hours on a charge. The LCD display is shock-mounted and highly readable in direct sunlight. It doesn't wash out, it just gets easier to read. There's a backlight too. And it's very light.
It's fun to use one of these on a long plane flight. After two hours, everyone else is either changing batteries or folding their laptops. After four hours, the battery-changers are folding up, and you're the only one still typing.
It doesn't have USB. Its only interface is serial. But the Palm Desktop software for the Mac, which is Claris Works in disguise, will export documents from the eMate either as ASCII text, or to Word.
I saw the original posting on the Cairo installfest, and as a longtime user of BSD-style UNIX (1978 or so), I was delighted. I had the pleasure of seeing Cairo a couple of years ago, and met a Linux devotee who was the son of an Egyptian family with whom I had dinner.
So, I followed the links to the website, and read a large number of the postings in the forum there. I don't suppose I should have been surprised at the infighting that seemed to be going on there - the noisy minorities usually dominate the forums, worldwide - but I was. What is the political climate inside the Egyptian open source world? Is it very highly factionalized?
Linus Torvalds needs to immediately drive to Montreal and get a parking ticket at one of these things while eating poutine or something. The circular irony of that is just too wonderful to pass up.
Sure enough, the Yellowstone caldera will blow high and mighty one of these millenia, and send an ash plume to the East Coast too. But we'll be able to pay handily for the damage by a simple use tax put on biogenetics companies, levied on the billions and billions of $$$ they're making from the genetics of the bacteria they pull out of the hot springs there. I've seen the little square of cyanobacterial mat they cut out of Octopus Pool up by Great Fountain geyser. That little square alone was worth billions, for the tailored bacteria they got out of it.
I deal with inboxes with 5,000-7,000 messages - not immense by today's standards (still boggled by the guy whose wife has 66,000 pending inbox messages) but large enough.
George Santayana keeps invading my consciousness. Most of today's mail readers are blindly taking the road that I abandoned 25 years ago. I don't want to read my mail using a database system. I want my mail to be a full-fledged member of UNIX society, not locked up inside a single application.
At RAND, we had a homebrew mail system that worked about like today's readers: mail was kept in a file, with a sidebar index file for quickly locating individual messages. It fell out of sync regularly, but on those dog-slow machines, rebuilding the index file was a coffee-break operation.
Norm Shapiro should be credited with the insight that UNIX already provided the cleanest solution to mail storage: messages are files, folders are directories. He and Bruce Borden hammered things out over about six months of conversations, then Bruce wrote the first version of the MH system over a weekend.
MH is ancient. There is no doubt about this. The original MH is as dead as T. Rex; people use NMH now. It's almost all text-only. It does have a MIME wart on the side, but just barely. If you want to use mice, scroll wheels, and other "modern" goodies you need to use a front end like EXMH.
BUT: 99.95% of all the legit email I get is text-only. "showproc" can deal with MIME mail that just asks for a different font, and EXMH does understand basic HTML. You can create MIME attachments if you need to.
And it's the skip-loader of email systems. It doesn't care if there are 8,000 messages in a folder. It just works. And it's fast.
On the Mac I use Mail.app. It does work (mostly, except when Apple is having one of its periodic days where WebDAV doesn't work, and they're in denial [nothing wrong here, move along please]). It has nice filtering features. It has threading.
It also feels like a toy. I get the feeling that if I pointed it at an 8,000-message inbox, it'd fold like a cheap suit. Certainly it'd be tough to deal with that many messages through that interface.
For the big time mail flows, I'm sticking with MH. Thanks again, Norm and Bruce.
Nothing you do is going to be fast when your mail database is that big. But the most efficient mail program I know of is the (now ancient) MH mail system. You could probably get it to run under cygwin. The problem with all other mail systems is that they're database systems, and you've got a database several orders of magnitude larger than what they're designed to work with. MH just deals in files and directories, so you get whatever the OS can do, performance-wise.
"Bubba" is a nickname for any random Southern cracker. It's so common it's trite, like calling all Russians "Ivan".
"Ho-tep" is a different matter. For one thing, it's spelled wrong. "Hotep" doesn't have a hyphen. It's really "htp", as Egyptian, like Hebrew, is written without vowels, and it means "gift". It's not a family name. A number of pharaohs were named "Amonhotep", and that means "gift of Amun". "Imhotep" was the vizier of the pharaoh Khufu, and is generally regarded as the architect of the Great Pyramid. He was also a physician, and was later deified as the god of physicians.
So BubbaHo-Tep would mean "gift of Bubba." Make of that what you will.:-)
(Me, I thought the movie was fun. Touching, too. Despite the ridiculous premise it had some depth.)
Yeah, that's what it looks like from the outside, but I bought a Radeon 9700, and had to run the bleeding-edge "-current" distribution of XFree86 to get drivers for it. In browsing the XFree86 CVS repository via CVSWeb, I noted that all of the Radeon-specific code supporting the 9700 in XFree86 had been checked in by an XFree86 committer with an address at ati.com.
So officially or not, it came from them. I was impressed.
Maddog indicates that legal shenanigans like the SCO suit are similar to the looting of the Iraqi museum because a few people bent on immediate profit are wrecking goods worth a lot more, in the long run, to the public at large.
Um, this happens all the time. Like it or not, it's part of our society. Capitalism does this. That's why we pass laws which act against immediate profiteers in favor of protecting, say, national parks.
Now, the first national park in the world (Yellowstone) was established by Congress in near-record time (for the day) back in the 1870s, when photos taken by William Henry Jackson of the territory that would become the park were printed up and sent to every member of Congress. And in the 1870s, the Wild West was in full swing. The Mining Act, which permits squatters to file mining claims on land they don't own, is also a product of that era, and a more Libertarian law (if that's not a contradiction in terms) would be hard to find.
But the only solution, in a capitalist society, to profit grabs which would result in great public harm is to pass great public laws against them. That's what the RIAA claims the new copyright laws do - keep the squatters off their land. The fact that a lot of people don't think it should be their land in the first place didn't pass muster in Congress.
So assuming Maddog's point is valid, what's to be done about it? Give every member of Congress a Linux system? I'd laugh a lot. Linux doesn't pass the "Mom" test or even come close. They'd scream for their Windoze boxes back.
I'd think OS X works better with HFS+ because it's optimized to work with HFS+.
If that's true, I wonder why? Why change FreeBSD to work better with HFS+?
I've read that HFS+ is more "modern" than UFS (aka BSD Fast File System), which was more important when hard drives were small and slow.
Now, that's the kicker. Is it true? The one piece of hard evidence I have says that it isn't: HFS+ needs to be linearized periodically for best performance, via Speed Disk or something similar. UFS/FFS doesn't, because it spreads inodes and block allocations evenly over the disk, and clusters things so that files in a directory are preferentially localized. HFS+ has a catalog B-tree, and unless that's split up and spread over the disk, it means that the disk will continuously seek as it looks up files and then accesses their content. No matter how fast disks are now, that has to be bad.
It's this sort of thing I'd like to hear Jordan's comments on.
I don't doubt Mac OS X performs better with HFS+. The questions are: 1) Why? and 2) Is this because HFS+ is better than UFS, or because the Mac OS X implementation of UFS is suboptimal?
It can be argued that Aqua's window system is a great deal more advanced than X. Heck, Rob Scheifler would probably agree with this. It's got a much more modern imaging model, is much more powerful, has a more unified architecture, and doesn't have 82 layers of cruft to fight through.
Only problem is, it's not network-aware. You wanna run a window on a foreign system, you either install X, and give up on remoting Aqua services entirely, or use Macintosh Manager, which does god-knows-what.
Sun dumped SunView 1 pretty damn quick when it became evident how mind-bogglingly useful X was in comparison. Can you comment on when we might see Aqua take the same step? Is it desirable?
The current OS X offering is more than a little schizophrenic. To wit, life in the Terminal app is ignorant of a whole bunch of stuff that's Apple-specific. If you tar up a bunch of files and untar them elsewhere, or even copy them, you can kiss the resource forks goodbye. The BSD side of the house doesn't even know that resource forks exist. And this is only one of the areas where Apple's worldview and BSD's worldview collide.
What about this? Are we going to see a system with a unified vision of life, both via Aqua and the shell?
Apple's going after the server market in a big way, for the first time in Apple's history. Mac OS X Server is their flagship (heck, their only) product in that department, and you'd think with FreeBSD's popularity, it'd be a slam dunk.
But on closer inspection, we see that the file system used in Mac OS X is, preferentially, HFS+. Now, UFS/FFS (aka the file system as performed by Kirk McKusick) has been tuned to within an inch of its life for close to 20 years to be able to do this, whereas, as far as I can tell, HFS+ is a) proprietary and b) hasn't ever been used seriously as a server file system before, having lived most of its life on desktops.
Soooooooooo...... what's with HFS+? How much of a performance hit, if any, do we take in using it instead of UFS? What would we see if we benchmarked the two of them in an "average" server?
Most of Mac OS X is FreeBSD, with Mach underpinnings to do the machine-dependent stuff. Memory management is also done by Mach. How does Mach's memory management stack up against the VM system in straight FreeBSD?
That's a heckuva piece of work, but the smiley appears to have been generated by parallel evolution. Several people seem to have come up with it independently. I first encountered it on Usenet around the same time period. I don't remember who it was who suggested it; all I remember is that it was a woman and hence couldn't have been Scott Fahlman.
The reference says that NoteCards provided "part of" the inspiration for HyperCard. My understanding was that it provided most of it. The question would be difficult to resolve, especially at this remove. I think it formed the basis of HyperCard - still, pretty good for a Travel Voucher hack.
Serial port access under Mac OS X seems quite dicey. An informal poll among Newton types, who either get to use USB-to-serial adapters or Ethernet if they're lucky, shows that access by OS 9 apps running under Classic to serial ports is an on-again, off-again affair. Currently, I have to boot to OS 9 to get access to a serial port.
In practice, it seems that serial port access depends on luck.
Apparently the beta version(s) of OS X had support for serial ports, and it was ripped out.
Getting access to the serial ports, via a USB-to-serial converter, is the sole reason I still boot OS 9. My Newtons need it. My GPS needs it. Just because Steve thinks serial is dead doesn't mean all other devices disappear. The day I can't get at my serial gear, by booting OS 9 or otherwise, is the day I quit upgrading my Apple gear.
Many, many years ago, when I was wandering through Xerox PARC for the first time, this one guy insisted on showing me this hack he'd put together in LISP (probably Interlisp) for filling out his Travel Expense Reports. He hated TERs because they're fussy and boring, so he'd created a system that let him create blank reports with rules, sort of like a generalized spreadsheet.
He called them Hypercards. Apparently this grew into quite something later, at Apple. Not bad for a hack for Travel Expense Reports.
This is the thing that really started to get me. Norton Systemworks, which includes Dantz Retrospect Express, comes on a bootable CD...bootable to OS 9. The OS 9 version of Retrospect cannot restore an OS X file system. Surprise!
Yes, there are all sorts of products which will clone a file system to an external drive, if one level of the most recent backup only will do. For me, that won't do...I was a sysadmin for too many years. I want a rack full of tapes.
Failing that, a backup set such as Retrospect provides will have to do. Which means that to back up an iBook I need two external Firewire drives - one to hold a bootable OS X plus Retrospect Express, and one to hold the backup set. Boy howdy, the bottom line just rose a tad!
The Achilles heel in all this is backup, especially for Mac OS X Server. Every other version of UNIX out there has a built-in backup solution (except, unaccountably, Linux, which has no dump/restore, last I checked). Mac OS X has dump/restore too, but they only understand the UFS file system. Apple rewrote 'fsck' to understand about HFS+ file systems, but not dump/restore.
That leaves Retrospect as the only sensible solution for backup: a third party product. And the regular Retrospect Mac OS X client won't dump a Mac OS X Server system! Instead you have to spend $800 (!!!) for the Server backup software. That software will also dump Windows 2000 and NT workstations, whoop-de-do.
Whatever happened to UNIX as a self-hosting, self-supporting system? Gaaaah. I'm thinking hard about wiping our Mac OS X Server machine and just installing the regular Mac OS X, where at least we can afford the backup software.
Or maybe just dumping Macs entirely and going to FreeBSD on a dual-processor Xeon box. All hail Amanda! At least I could back up a box like that.
I used (and programmed) every version of Newton device. There were several generations of Newton recognition software. The first generation was actually licensed from a Russian company called Paragraph.
I had a chance to talk to the folks at ParaGraph International at a mobile computing conference once. It was a very enlightening conversation. They all used to work at the Soviet Academy of Sciences, and had decided that rather than slowly starve in the post-Soviet era, they'd rather form a Russian equivalent of Bell Labs. Now, it turns out one thing they were really good at was curve-fitting. They could use higher-order polynomials to compress and characterize curves of arbitrary shape. As a demonstration, they'd taking a Picasso pencil sketch (in color) and compressed it down to 17K, then re-expanded it into something indistinguishable from the original.
These folks told me that when Apple first contracted with them,they were held at such arm's length that they didn't even know what kind of device they were writing a recognizer for. They never even saw a Newton until they hit market. Hence,they had no opportunity to tune the recognizer. Those who've used Newtons know that things were difficult at best until Newton OS 2.0 came out for the 120. After that, it got much better (and the Rosetta printed recognizer really helped). That was the release that used the 'tuned' cursive recognizer, and with further tweaking in Newton OS 2.1, it pretty much rocks. No more Egg Freckles.
Inkwell, of course, is based on Rosetta, not the ParaGraph recognizer, but the latter is available as a separate package for other PDAs.
Slashdot Mirror
The Newton eMate, (very) dead technology from Apple, is available for cheap on eBay. It's very rugged, having been designed for kids. It has a built-in keyboard and no disk drive (FLASH RAM instead). It runs for hours and hours on a charge. The LCD display is shock-mounted and highly readable in direct sunlight. It doesn't wash out, it just gets easier to read. There's a backlight too. And it's very light.
It's fun to use one of these on a long plane flight. After two hours, everyone else is either changing batteries or folding their laptops. After four hours, the battery-changers are folding up, and you're the only one still typing.
It doesn't have USB. Its only interface is serial. But the Palm Desktop software for the Mac, which is Claris Works in disguise, will export documents from the eMate either as ASCII text, or to Word.
I saw the original posting on the Cairo installfest, and as a longtime user of BSD-style UNIX (1978 or so), I was delighted. I had the pleasure of seeing Cairo a couple of years ago, and met a Linux devotee who was the son of an Egyptian family with whom I had dinner.
So, I followed the links to the website, and read a large number of the postings in the forum there. I don't suppose I should have been surprised at the infighting that seemed to be going on there - the noisy minorities usually dominate the forums, worldwide - but I was. What is the political climate inside the Egyptian open source world? Is it very highly factionalized?
Linus Torvalds needs to immediately drive to Montreal and get a parking ticket at one of these things while eating poutine or something. The circular irony of that is just too wonderful to pass up.
Sure enough, the Yellowstone caldera will blow high and mighty one of these millenia, and send an ash plume to the East Coast too. But we'll be able to pay handily for the damage by a simple use tax put on biogenetics companies, levied on the billions and billions of $$$ they're making from the genetics of the bacteria they pull out of the hot springs there. I've seen the little square of cyanobacterial mat they cut out of Octopus Pool up by Great Fountain geyser. That little square alone was worth billions, for the tailored bacteria they got out of it.
I deal with inboxes with 5,000-7,000 messages - not immense by today's standards (still boggled by the guy whose wife has 66,000 pending inbox messages) but large enough.
George Santayana keeps invading my consciousness. Most of today's mail readers are blindly taking the road that I abandoned 25 years ago. I don't want to read my mail using a database system. I want my mail to be a full-fledged member of UNIX society, not locked up inside a single application.
At RAND, we had a homebrew mail system that worked about like today's readers: mail was kept in a file, with a sidebar index file for quickly locating individual messages. It fell out of sync regularly, but on those dog-slow machines, rebuilding the index file was a coffee-break operation.
Norm Shapiro should be credited with the insight that UNIX already provided the cleanest solution to mail storage: messages are files, folders are directories. He and Bruce Borden hammered things out over about six months of conversations, then Bruce wrote the first version of the MH system over a weekend.
MH is ancient. There is no doubt about this. The original MH is as dead as T. Rex; people use NMH now. It's almost all text-only. It does have a MIME wart on the side, but just barely. If you want to use mice, scroll wheels, and other "modern" goodies you need to use a front end like EXMH.
BUT: 99.95% of all the legit email I get is text-only. "showproc" can deal with MIME mail that just asks for a different font, and EXMH does understand basic HTML. You can create MIME attachments if you need to.
And it's the skip-loader of email systems. It doesn't care if there are 8,000 messages in a folder. It just works. And it's fast.
On the Mac I use Mail.app. It does work (mostly, except when Apple is having one of its periodic days where WebDAV doesn't work, and they're in denial [nothing wrong here, move along please]). It has nice filtering features. It has threading.
It also feels like a toy. I get the feeling that if I pointed it at an 8,000-message inbox, it'd fold like a cheap suit. Certainly it'd be tough to deal with that many messages through that interface.
For the big time mail flows, I'm sticking with MH. Thanks again, Norm and Bruce.
Nothing you do is going to be fast when your mail database is that big. But the most efficient mail program I know of is the (now ancient) MH mail system. You could probably get it to run under cygwin. The problem with all other mail systems is that they're database systems, and you've got a database several orders of magnitude larger than what they're designed to work with. MH just deals in files and directories, so you get whatever the OS can do, performance-wise.
"Bubba" is a nickname for any random Southern cracker. It's so common it's trite, like calling all Russians "Ivan".
:-)
"Ho-tep" is a different matter. For one thing, it's spelled wrong. "Hotep" doesn't have a hyphen. It's really "htp", as Egyptian, like Hebrew, is written without vowels, and it means "gift". It's not a family name. A number of pharaohs were named "Amonhotep", and that means "gift of Amun". "Imhotep" was the vizier of the pharaoh Khufu, and is generally regarded as the architect of the Great Pyramid. He was also a physician, and was later deified as the god of physicians.
So BubbaHo-Tep would mean "gift of Bubba." Make of that what you will.
(Me, I thought the movie was fun. Touching, too. Despite the ridiculous premise it had some depth.)
Yeah, that's what it looks like from the outside, but I bought a Radeon 9700, and had to run the bleeding-edge "-current" distribution of XFree86 to get drivers for it. In browsing the XFree86 CVS repository via CVSWeb, I noted that all of the Radeon-specific code supporting the 9700 in XFree86 had been checked in by an XFree86 committer with an address at ati.com.
So officially or not, it came from them. I was impressed.
Maddog indicates that legal shenanigans like the SCO suit are similar to the looting of the Iraqi museum because a few people bent on immediate profit are wrecking goods worth a lot more, in the long run, to the public at large.
Um, this happens all the time. Like it or not, it's part of our society. Capitalism does this. That's why we pass laws which act against immediate profiteers in favor of protecting, say, national parks.
Now, the first national park in the world (Yellowstone) was established by Congress in near-record time (for the day) back in the 1870s, when photos taken by William Henry Jackson of the territory that would become the park were printed up and sent to every member of Congress. And in the 1870s, the Wild West was in full swing. The Mining Act, which permits squatters to file mining claims on land they don't own, is also a product of that era, and a more Libertarian law (if that's not a contradiction in terms) would be hard to find.
But the only solution, in a capitalist society, to profit grabs which would result in great public harm is to pass great public laws against them. That's what the RIAA claims the new copyright laws do - keep the squatters off their land. The fact that a lot of people don't think it should be their land in the first place didn't pass muster in Congress.
So assuming Maddog's point is valid, what's to be done about it? Give every member of Congress a Linux system? I'd laugh a lot. Linux doesn't pass the "Mom" test or even come close. They'd scream for their Windoze boxes back.
Maybe Apple should send them all G5s...
If that's true, I wonder why? Why change FreeBSD to work better with HFS+?
Now, that's the kicker. Is it true? The one piece of hard evidence I have says that it isn't: HFS+ needs to be linearized periodically for best performance, via Speed Disk or something similar. UFS/FFS doesn't, because it spreads inodes and block allocations evenly over the disk, and clusters things so that files in a directory are preferentially localized. HFS+ has a catalog B-tree, and unless that's split up and spread over the disk, it means that the disk will continuously seek as it looks up files and then accesses their content. No matter how fast disks are now, that has to be bad.
It's this sort of thing I'd like to hear Jordan's comments on.
I don't doubt Mac OS X performs better with HFS+. The questions are: 1) Why? and 2) Is this because HFS+ is better than UFS, or because the Mac OS X implementation of UFS is suboptimal?
...I was still doing the magazine column, so you could tell me in private that you weren't going to answer the more embarrassing questions? :-)
It can be argued that Aqua's window system is a great deal more advanced than X. Heck, Rob Scheifler would probably agree with this. It's got a much more modern imaging model, is much more powerful, has a more unified architecture, and doesn't have 82 layers of cruft to fight through.
Only problem is, it's not network-aware. You wanna run a window on a foreign system, you either install X, and give up on remoting Aqua services entirely, or use Macintosh Manager, which does god-knows-what.
Sun dumped SunView 1 pretty damn quick when it became evident how mind-bogglingly useful X was in comparison. Can you comment on when we might see Aqua take the same step? Is it desirable?
The current OS X offering is more than a little schizophrenic. To wit, life in the Terminal app is ignorant of a whole bunch of stuff that's Apple-specific. If you tar up a bunch of files and untar them elsewhere, or even copy them, you can kiss the resource forks goodbye. The BSD side of the house doesn't even know that resource forks exist. And this is only one of the areas where Apple's worldview and BSD's worldview collide.
What about this? Are we going to see a system with a unified vision of life, both via Aqua and the shell?
Apple's going after the server market in a big way, for the first time in Apple's history. Mac OS X Server is their flagship (heck, their only) product in that department, and you'd think with FreeBSD's popularity, it'd be a slam dunk.
But on closer inspection, we see that the file system used in Mac OS X is, preferentially, HFS+. Now, UFS/FFS (aka the file system as performed by Kirk McKusick) has been tuned to within an inch of its life for close to 20 years to be able to do this, whereas, as far as I can tell, HFS+ is a) proprietary and b) hasn't ever been used seriously as a server file system before, having lived most of its life on desktops.
Soooooooooo...... what's with HFS+? How much of a performance hit, if any, do we take in using it instead of UFS? What would we see if we benchmarked the two of them in an "average" server?
Most of Mac OS X is FreeBSD, with Mach underpinnings to do the machine-dependent stuff. Memory management is also done by Mach. How does Mach's memory management stack up against the VM system in straight FreeBSD?
That's a heckuva piece of work, but the smiley appears to have been generated by parallel evolution. Several people seem to have come up with it independently. I first encountered it on Usenet around the same time period. I don't remember who it was who suggested it; all I remember is that it was a woman and hence couldn't have been Scott Fahlman.
...and his twin plays, Dogg's Hamlet/Kohout's Macbeth.
The reference says that NoteCards provided "part of" the inspiration for HyperCard. My understanding was that it provided most of it. The question would be difficult to resolve, especially at this remove. I think it formed the basis of HyperCard - still, pretty good for a Travel Voucher hack.
...and sometimes it doesn't.
Serial port access under Mac OS X seems quite dicey. An informal poll among Newton types, who either get to use USB-to-serial adapters or Ethernet if they're lucky, shows that access by OS 9 apps running under Classic to serial ports is an on-again, off-again affair. Currently, I have to boot to OS 9 to get access to a serial port.
In practice, it seems that serial port access depends on luck.
Apparently the beta version(s) of OS X had support for serial ports, and it was ripped out.
Getting access to the serial ports, via a USB-to-serial converter, is the sole reason I still boot OS 9. My Newtons need it. My GPS needs it. Just because Steve thinks serial is dead doesn't mean all other devices disappear. The day I can't get at my serial gear, by booting OS 9 or otherwise, is the day I quit upgrading my Apple gear.
Many, many years ago, when I was wandering through Xerox PARC for the first time, this one guy insisted on showing me this hack he'd put together in LISP (probably Interlisp) for filling out his Travel Expense Reports. He hated TERs because they're fussy and boring, so he'd created a system that let him create blank reports with rules, sort of like a generalized spreadsheet.
He called them Hypercards. Apparently this grew into quite something later, at Apple. Not bad for a hack for Travel Expense Reports.
This is the thing that really started to get me. Norton Systemworks, which includes Dantz Retrospect Express, comes on a bootable CD...bootable to OS 9. The OS 9 version of Retrospect cannot restore an OS X file system. Surprise!
Yes, there are all sorts of products which will clone a file system to an external drive, if one level of the most recent backup only will do. For me, that won't do...I was a sysadmin for too many years. I want a rack full of tapes.
Failing that, a backup set such as Retrospect provides will have to do. Which means that to back up an iBook I need two external Firewire drives - one to hold a bootable OS X plus Retrospect Express, and one to hold the backup set. Boy howdy, the bottom line just rose a tad!
The Achilles heel in all this is backup, especially for Mac OS X Server. Every other version of UNIX out there has a built-in backup solution (except, unaccountably, Linux, which has no dump/restore, last I checked). Mac OS X has dump/restore too, but they only understand the UFS file system. Apple rewrote 'fsck' to understand about HFS+ file systems, but not dump/restore.
That leaves Retrospect as the only sensible solution for backup: a third party product. And the regular Retrospect Mac OS X client won't dump a Mac OS X Server system! Instead you have to spend $800 (!!!) for the Server backup software. That software will also dump Windows 2000 and NT workstations, whoop-de-do.
Whatever happened to UNIX as a self-hosting, self-supporting system? Gaaaah. I'm thinking hard about wiping our Mac OS X Server machine and just installing the regular Mac OS X, where at least we can afford the backup software.
Or maybe just dumping Macs entirely and going to FreeBSD on a dual-processor Xeon box. All hail Amanda! At least I could back up a box like that.
These folks told me that when Apple first contracted with them,they were held at such arm's length that they didn't even know what kind of device they were writing a recognizer for. They never even saw a Newton until they hit market. Hence,they had no opportunity to tune the recognizer. Those who've used Newtons know that things were difficult at best until Newton OS 2.0 came out for the 120. After that, it got much better (and the Rosetta printed recognizer really helped). That was the release that used the 'tuned' cursive recognizer, and with further tweaking in Newton OS 2.1, it pretty much rocks. No more Egg Freckles.
Inkwell, of course, is based on Rosetta, not the ParaGraph recognizer, but the latter is available as a separate package for other PDAs.