Have you checked out SGI's patches for NFSV3 against the 2.2.10 kernel?
We use a LOT of nfs at work. I already have Trond's NFS V3 patches applied to my kernels, and they work OK, sort of. I can also run an NFS V3 server with the knfsd patches (or Neil Brown's new patches). But the linux nfs V3 server, while working with Digital and SGI NFS V3 clients, cannot be seen by linux nfs v3 clients, OR linux nfs v2 clients.
I feel safe in saying linux nfs v2 server and client are working well. V3 would sure help a lot, but I just don't see it happening in any sort of stable well-maintained way - yet. Of course, the speed differences are not great until you get to 100 BT lines anyway...
This has HUGE raminifications for the usability of linux as a file server in a distributed Unix environment. NFS servers from linux are behind virtually EVERY other flavor of Unix. I suspect they will come around. But nfs is not exactly improving fast.
How about fixing NFS on SMP too. That's been broken ever since 2.2.13. It seems like Alan was working on it in September and then he just lost interest in it.
2.2.13 must be what - at least a month old ?
NFS on SMP is working JUST FINE on my SMP box, running 2.2.12 with the knfsd-1.4.7 patches. And it has been for about 2.5 months. Most of the knfsd patch functionality has been merged into the 2.2.14-pre tree, so 2.2.14 ought to be a fairly stable NFS branch, even for SMP users. At least for NFS version two, which is a fairly old standard.
If you are really interested, there is a separate mailing list for nfs users that has been posted to the kernel mailing list, and the user space utilities have been evolving from knfsd-1.4.7 to nfs-utils-0.15 or so. Linux nfs now has locking and everything. Still, you'd have to consider nfs on linux a real weakness compared to other Unices. NFS version 3 is still a pipe-dream for clients or servers, and version two is just now stabilizing in the 2.2 tree.
ESR contends that the DVDCA wants to protect players since PC's will compete against DVD players. This argument holds very little water. For starters, Windows has a much larger marketshare than Linux, and Windows machines will compete more against DVD players, but Windows ALREADY had many, many software DVD players. Furthermore, these Windows players are essentially free of cost -- every video card in the world comes with a software DVD player.
That is just part of the conspiracy. Not all DVD players can decode all the keys. There are regional aspects to distribution. A Japanese DVD would not play in a US player. The Windows DVD players are in line with this. The linux players are not. If the encryption is broken then a DVD in Japan is just as good as a DVD here.
Also, is it really true that you can make a bit-for-bit copy? My understanding was that this required specialized hardware, and that commdity DVD reader hardware was not capable of reading special tracks.
Bit for bit READS can be made currently. The writes require a very expensive machine to do double sided double layering 17 GBytes per disk. For not so much money you could copy a DVD onto four disks though. Bit for bit. That is part of the point though - currently it costs a LOT more to make a copy of a disk at home than it costs to buy the disk. This will likely change in a few more years though.
TW, OSF UNIX is ancient. It it now called Tru64 Unix, and is rather different than all old OSF. In fact, I'm suprised that you still run OSF: OSF -> Digital Unix -> Tru64 Unix
Here is the output on a brand spanking new XP-1000 for "uname -a" OSF1 [$HOSTNAME] V4.0 1091 alpha
fsck does not fix all problems. However in your original post, you said that AdvFS did not have any utilities (simply because it lacks a "fsck"), when in fact it does. Your problems with the badly broken filesystem are beyond me at the moment, since I don't know (or care) about your setup, your specific problems, and whatnot. My post was simply to inform that utilities do exist, since again, you stated that they did not.
I did NOT say AdvFS had no utilities. I said it could not be fixed. We have some 60-70 networked 4.0[de] machines mostly running AdvFS. For the most part, it had been great. The only fileset corruption we got could not be fixed with existing tools. We are 0 for 1 in fixing corrupted filesets. In our experience, it is not so hot at fixing broken filesets. I'd much rather have a fixable UFS and deal with it.
It is also nice to know you care about troubles for DecPaq users:)
Tell that to my four bad inodes, on which AdvFS tools have failed to remove.
Actually, let me clarify. I have four bad files. When I try to run ls, I get a
ls: [bad file name] : Invalid argument
It also appears stat fails on the files. Verify failed to remove them. Now, there files are on an NFS mounted home directory that contains home directories for hundreds of people, and running verify took out the directories for the better part of the morning. The man page for salvage indicates it is intended for use for salvaging unmountable domains. This domain is mountable - it only has four bad files which are not important anyway.
Postings to osf1 newsgroups have resulted in the advice to manually copy the remaining usable files, destroy the domain, and recreate it.
From the salvage man page
The salvage utility helps you recover file data after an AdvFS file domain has become unmountable due to some type of data corruption. Errors that could cause data corruption of a file domain include I/O errors in file system metadata, the accidental removal of a volume, or any I/O error that produces a panic Use the salvage utility as a last resort.You should first repair domain structures by using the verify utility. If that repair method is unsatisfactory, attempt to recover fileset data from backup media. Only if both methods are unsatisfactory should you employ the salvage utility.
The lengths to which we've had to go and the difficulty in performing the equivalent of a very simple fsck has made us seriously reconsider our use of the Advanced File System. The file system, when damaged, is non-trivial to fix, if it is indeed fixable. The verify command did NOT work, and the salvage man page suggests trying a manual backup first instead !!!!
Lay down the crack pipe and take a deep breath and read what I wrote again. Insanely idiotic OS and component licensing scheme. The licensing is insane, not the OS.
Maybe you'd have a leg to stand on if Linux supported the enterprise features that Digital UNIX does. Unfortunately, it doesn't. Example: High performance, dynamically resizable, journalling filesystem.
AdvFS cannot be fscked. This in turn has fscked me. Get the picture. If the FS breaks, you keep both pieces. Thank you Digital for such a wonderful advance in computing. Bad inodes just get to live indefinitely on your system until you copy all the files to a different partition (dump and tar choke), and reformat. We haven't had such advances since, well, DOS. Needless to say, we are going back to UFS for our OSF needs, which is only quite a bit slower than ext2. But at least when it breaks we can fix it.
OSF also doesn't ship with a reasonably modern interface. CDE, MWM, and TWM are simply not enough.
It certainly has its niche for ultra high end computing where user interfaces are just not viewed as important. But the OS is chock full of holes. And I run into them from time to time.
Example. glibc call for wordexp does a complete shell-like expansion in C. Libc shipped with OSF does the expansion by shelling a command to ksh. Why should libc depend on ksh for its integrity ??
Example. Recursive scripts fill up the process table and lock the system in OSF. Not so in linux.
There are also good things. CC is a really fast compiler. So is the ForTran compiler. If you want to run processes real fast, OSF is a good choice. If you want a large number of CPUs, OSF is a good choice. If you want a decent user interface, reasonable speed, and a journaled file system that can be fixed, linux is a good choice.
Alpha costs more? When's the last time you priced a top of the line Intel Xeon? I found a PIII Xeon 550/2MB priced at $ 3795.95 - and that's just the processor! (and these guys are cheap!)
Compare apples to apples. A full system with a PIII 500MHz is around $2500-$3k including a complement of RAM and hard disk space. The top alpha is ONLY available from Compaq. The fast motherboard is the XP-1000 for about $10k without the RAM or monitor. And add-ons for the Compaq alphas machines are EXPENSIVE. One generation late you can get a 21164 processor which is still fast in a machine for about $3k. Video card support under OSF or alpha linux is very poor. Unless you really need the supercomputing, it just doesn't make a lot of sense to buy an alpha. Now, if Compaq somehow made third party peripherals supportable for alphas they would sell a WHOLE lot more alphas without losing the number crunching crowd.
The alpha processors are not changing their niche in the computer market. They are ripping fast - and Dec first and now Compaq plays to the supercomputing crowd. The XP series motherboards and 21264 chips simply rip any other motherboard/chipset out there.
However, they cost too much for anyone except a supercomputing hound. If Compaq would drop Dec's insanely idiotic OS and component licensing scheme and aid linux on alphas, they might stand a chance of making a LOT of money selling hardware. As is, people buy ten times more alphas one chip generation late and run linux instead of OSF.
Anyone interested should see the linux alpha compilers available. cc is a small improvement, and ForTran is a LARGE improvement.
http://www.unix.digital.com/linux/software.htm
But still, Itanium will come out, and an Itanium box will offer slightly less than half the floating point speed, and it will cost about 1/4th of the fast alpha box from Compaq. And the alpha motherboards will still make it tough to support third party peripherals. And Itanium will dominate the 64 bit market. And Alpha will own the supercomputing market.
To get a really ergonomic mouse they ought to construct one that doesn't require use of finger or thumb muscles. Like place the buttons on the near side of the mouse. The ergonomic issue is 1) use larger muscles whenever possible
BTW, this is also why it is recommended never to rest your wrists while you type. Keeping the wrists up allows forearm movement to do some of the work of typing.
Any mouse work that requires click and drag or lots of double clicking is asking for lots of trouble.
Various groups have now shown that there are substantial interactions between sensory representational abnormalities and loss of motor function in repetitive stress injuries. The emerging hypothesis is that in this most common type of dysfunction, attention and motor action cause sensory representations to become abnormal. This in turn tweeks motor feedback, which then worsens. This in turn causes further erosion of sensory representations. And so on. In the worst cases focal dystonia can result which NEVER GOES AWAY and NEVER GETS BETTER.
Unless the underlying sensory representation is restored.
But the bottom line is that once your hands begin to feel weird in a repetitive stress injury, continued use will lead you down the spiral towards loss of control. Stop, and take a break. And fairly soon effective sensory training will be able to reverse the negative effects of the repetitive movement.
The article states that stroke victims are at risk of thrombosis, yet these devices are made out of glass capillaries, which is thrombogenic.
I don't think they are made out of glass capillaries. Glass is not so biocompatible. I would be extremely surprised if the glass is not insulated with a polymer (like parylene) or is actually a plastic tube to begin with.
Lastly, the critic mentioned in the article is correct: long-term studies are needed. These lil things can become lodged in joints, escape into the bloodstream, etc. They might be carcinogenic. Who knows? Its time for a randomized trial.
The concept of these lil things escaping the muscle capsule is pretty out there. The more likely confound is encapsulation in the form of a foreign body reaction, and even that would not be so bad, and very undo-able (^_). The risks are pretty low, and the potential gain is quite high. I think they will gain approval without much trouble.
That said, it IS a step in the right direction. Current science is 99% focused on the chemistry of neurology...theyre forgetting the electrical aspects of it. Not until doctors looked at the heart as an electrical organ did they really do anything that can be described as "cardiology," and the brain is every bit as electrical as the heart.
So nitric oxide treatment, bypass surgery, and angioplasty are not really cardiology ??
Really, cardiology, like other disciplines, works in a gradient descent method. Techniques that lead to improvements in short time frame get worked on the most. Electrically, as well as chemically. There are already numerous electrical techniques at work in brain pathologies - like stimulating electrodes for Parkinson's patients.
I work for a company closely affiliated with the one who developed this device (about a 7 year process). Here's two URLS to the people involved, although information there is sparse. http://www.advancedbionics.com These people also market Cochlear implants that restore hearing to totally deaf patients. http://www.aemf.org This is the non-profit R&D foundation that has actually done most of the development. If anyone really wants to know more about these devices, please contact them... Dave Marsh Systems Engineer Medical Research Group, Inc.
Interesting. Gerald Loeb was part of the team that developed the cochlear implant technology while he was at UCSF some 15-20 years ago. That technology led to patents that form the basis of the first generation cochlear implants used at advanced bionics. The muscle implants should be fairly easy. Muscle tissue is not so hard to implant, and much more difficult things are already being done in nervous system tissue.
How predetermined is out number of senses and limbs? If I was to implant some device like this into my brain would it always a) feel unnatural b) necessarily interfere with the movement of my arm or c) eventually behave entirely unconciously just like the movement of my arm does now.
Our number of senses and limbs is fairly predetermined. Early sensory pathways are guided by cues that appear to be mostly genetic. Within each area the potential for change is enormous. People have little to no problem adapting to using cochlear implants, for example. In cases in which implants allow movements of devices, I would be surprised if the person didn't feel an attachment of sorts to the device they are moving. Heck, if you use a pencil a lot you feel as though you are touching the paper with the pencil tip instead of feeling the vibrations that move through the pencil. In golf you feel as though you are striking a golf ball, whereas you are really only feeling a club.
Is the same true of senses? If I attached a little I/O port into my brain would it eventually be integrated as a 6th sense or are out brains to preprogramed for this to happen?
It is very unclear on how one would attach an I/O port to the brain. Are you just going to put a serial port on the brain and plug in ?? Implants rely on causing actions based on activity in existing neurons. Sensory implants activate sensory afferents.
This is no surprise at all. The surprise is how many of these exist that are not publicized.
And I doubt it stops there. I would think every major ISP is tracking hits from every user, particularly to correlate web purchases with web site visits. That is very valuable info. Your ISP can record EVERY MOVE YOU MAKE if they are so inclined. And it is worth a lot, so in business terms they would be idiots not to track you.
There is a little oversight on the technology by this professor.
There are already recent high profile experiments by John Chapin's group in Philadelphia in which rats are trained to control lever arms by activating parts of their brains in which electrodes have been placed.
Other cortical implants in primary motor cortex of humans show a lot of promise for quadrapelegics to be able to control robotics around them. The last I saw humans were quite able to make LED displays light up in certain patterns, and it is a small step from there to controlling robotics. In both cases a small number of bits of information is measured from the CNS and converted to simple movements.
But the converse, computer control of humans, is in a much much more primitive state. Likening the ability of implanted chips to enable the lights to be turned on, to giving the computer control of your limbs is at this point a crack pipe dream, and there is far more advanced work going on than this press release. For example, advanced bionics makes cochlear implants that allow high fidelity sound reception. There is comparable work on retinal implants at a much less advanced state. But the problem is much more advanced, since the human arm has more than a few bits worth of degrees of freedom, and its natural control from motor cortex to movement is not even close to fully understood.
As for controlling emotion, that would be a far simpler problem, since there are a few well localized centers of neuromodulators, and implanted stimulating electrodes already exist. It will be fairly easy to make someone hostile and aggressive, or sleepy, or passive, or edgy. Making someone alert and intelligent would be much harder. There would be a lot of unreversible psychosis caused in the pilot experiments, so don't expect them anytime soon.
> So, we're out with CUPS. Now if we could just > get Adobe to give us some Linux versions of > their tools, life might be (almost) perfect.
Adobe tools suck rocks. Their acroread is poorly implemented in a lot of ways. They patent their standards (like pdf) to prevent others from being able to make pdfs without paying the man.
They embrace and extend their proprietary standards (PS and PDF) to keep open source tools from being able to work with full functionality.
I do a lot of PS hacking with open source tools, and Adobe has about the worst attitude towards open source PS hackers as you can get.
First, I wish you the best in your search for painless typing. Be aware, however, that many people who experience pain while typing do not have carpal tunnel, and most of them finally resort to fewer and shorter bouts of typing.
If it feels like your fingers are about to tie themselves in knots - take a break.
Don't rest your wrists on anything. Keep them in the air and force your forearms to do some of the work. Bigger muscle groups are much less susceptible to repetitive stress injuries.
And a last thought to those who may have severe problems. A retraining strategy that focusses on sensory retraining of the hands has a very high efficacy after diligent training for 2-3 months. Many people who thought they would never use their hands the same way again now can. This therapy was pioneered by Nancy Byl and Mike Merzenich at UCSF. It supposes that a large portion of the problem is sensory reorganization during heavily attended heavily repeated motor actions, and the lack of sensory feedback wreaks havoc with motor control.
> I've been stuck with Word for years b/c my the > equations and graphs in my existing > documents never came out right in other on > other word processors and spreadsheets.... > until StarOffice, which seems to work pretty > well.
Nothing beats TeX and its derivatives for ease of use in making equations.
This is not to say you shouldn't try Word, WordPerfect, StarOffice... but Knuth developed TeX BECAUSE existing equation editors sucked rocks. And still, TeX has the easiest to use and most flexible implementation. And it was made to make documents - not made to draft a one page letter.
> This isn't troll talk - it is slow compared to > MS Office. On Linux it takes about a minute to > load, on windows, around about the same. ewww. > MS Office takes like 2 seconds (per app).
Windows generally preloads the libraries for Office into RAM so that bloated pigs like Word will start faster. Try closing the Office toolbar and then opening Word and look at the difference in load time.
It says next to nothing about security of linux or NT except that if you set up your machine like a complete total moron you are likely to get cracked.
The funniest thing about this stunt is that PCWeek has just badly eroded its credibility.
To verify this, try going around and asking your PhD professors of engineering how many of them are licensed PEs.
And for electrical and/or computer engineering it is more a rarity than common among professionals to have a PE.
For MOST things related to EE or CS, the PE will be viewed in a similar light as an MSCE. It is kinda nice to have, but doesn't tell you very much about the quality of the job candidate.
In addition, the PE test (the first phase of certification) is almost exclusively based on the core curriculum for Civil and Mechanical Engineers. Yes there is an electrical engineering section - and it is almost exclusively based on things taught in the first intro EE course that most civil and mechanical engineers are forced to take.
If I were a graduating EE or CS major, I would forget about the PE. As a mechanical or civil engineer, I would probably do it.
Slashdot Mirror
Have you checked out SGI's patches for NFSV3 against the 2.2.10 kernel?
We use a LOT of nfs at work. I already have Trond's NFS V3 patches applied to my kernels, and they work OK, sort of. I can also run an NFS V3 server with the knfsd patches (or Neil Brown's new patches). But the linux nfs V3 server, while working with Digital and SGI NFS V3 clients, cannot be seen by linux nfs v3 clients, OR linux nfs v2 clients.
I feel safe in saying linux nfs v2 server and client are working well. V3 would sure help a lot, but I just don't see it happening in any sort of stable well-maintained way - yet. Of course, the speed differences are not great until you get to 100 BT lines anyway...
This has HUGE raminifications for the usability of linux as a file server in a distributed Unix environment. NFS servers from linux are behind virtually EVERY other flavor of Unix. I suspect they will come around. But nfs is not exactly improving fast.
How about fixing NFS on SMP too. That's been broken ever since 2.2.13. It seems like Alan was working on it in September and then he just lost interest in it.
2.2.13 must be what - at least a month old ?
NFS on SMP is working JUST FINE on my SMP box, running 2.2.12 with the knfsd-1.4.7 patches. And it has been for about 2.5 months. Most of the knfsd patch functionality has been merged into the 2.2.14-pre tree, so 2.2.14 ought to be a fairly stable NFS branch, even for SMP users. At least for NFS version two, which is a fairly old standard.
If you are really interested, there is a separate mailing list for nfs users that has been posted to the kernel mailing list, and the user space utilities have been evolving from knfsd-1.4.7 to nfs-utils-0.15 or so. Linux nfs now has locking and everything. Still, you'd have to consider nfs on linux a real weakness compared to other Unices. NFS version 3 is still a pipe-dream for clients or servers, and version two is just now stabilizing in the 2.2 tree.
Bit for bit READS can be made currently.
No, they can't. At least not on consumer hardware. They can't read all the information.
It is this simple. Stick the DVD in your linux
box with DVD drive. Have an empty 18 GByte
ready.
dd if=/dev/dvdrom of=/dev/sdb
VOILA ! You've made a bit for bit copy (at least,
after a few hours you have) Perhaps the Windows drivers are less lenient - I wouldn't know.
All this and more at http://www.opendvd.org/
Expect a demo of this in court on the 14th.
ESR contends that the DVDCA wants to protect players since PC's will compete against DVD players. This argument holds very little water. For starters, Windows has a much larger marketshare than Linux, and Windows machines will compete more against DVD players, but Windows ALREADY had many, many software DVD players. Furthermore, these Windows players are essentially free of cost -- every video card in the world comes with a software DVD player.
That is just part of the conspiracy. Not all DVD players can decode all the keys. There are regional aspects to distribution. A Japanese DVD would not play in a US player. The Windows DVD players are in line with this. The linux players are not. If the encryption is broken then a DVD in Japan is just as good as a DVD here.
Also, is it really true that you can make a bit-for-bit copy? My understanding was that this required specialized hardware, and that commdity DVD reader hardware was not capable of reading special tracks.
Bit for bit READS can be made currently. The writes require a very expensive machine to do double sided double layering 17 GBytes per disk. For not so much money you could copy a DVD onto four disks though. Bit for bit. That is part of the point though - currently it costs a LOT more to make a copy of a disk at home than it costs to buy the disk. This will likely change in a few more years though.
This and more from http://www.opendvd.org
Just like the book Zodiac by Neal Stephenson.
/.
Just had to do it.
Mandatory Stephenson reference on
TW, OSF UNIX is ancient. It it now called Tru64 Unix, and is rather different than all old OSF. In fact, I'm suprised that you still run OSF: OSF -> Digital Unix -> Tru64 Unix
Here is the output on a brand spanking new XP-1000 for "uname -a"
OSF1 [$HOSTNAME] V4.0 1091 alpha
ysyi@alf.dec.com wrote:
:)
fsck does not fix all problems. However in your original post, you said that AdvFS did not have any utilities (simply because it lacks a "fsck"), when in fact it does. Your problems with the badly broken filesystem are beyond me at the moment, since I don't know (or care) about your setup, your specific problems, and whatnot. My post was simply to inform that utilities do exist, since again, you stated that they did not.
I did NOT say AdvFS had no utilities. I said it could not be fixed. We have some 60-70 networked 4.0[de] machines mostly running AdvFS. For the most part, it had been great. The only fileset corruption we got could not be fixed with existing tools. We are 0 for 1 in fixing corrupted filesets. In our experience, it is not so hot at fixing broken filesets. I'd much rather have a fixable UFS and deal with it.
It is also nice to know you care about troubles for DecPaq users
Tell that to my four bad inodes, on which AdvFS tools have failed to remove.
Actually, let me clarify. I have four bad files.
When I try to run ls, I get a
ls: [bad file name] : Invalid argument
It also appears stat fails on the files.
Verify failed to remove them. Now, there files are on an NFS mounted home directory that contains home directories for hundreds of people, and running verify took out the directories for the better part of the morning. The man page for salvage indicates it is intended for use for salvaging unmountable domains. This domain is mountable - it only has four bad files which are not important anyway.
Postings to osf1 newsgroups have resulted in the advice to manually copy the remaining usable files, destroy the domain, and recreate it.
From the salvage man page
The salvage utility helps you recover file data after an AdvFS file domain has become unmountable due to some type of data corruption. Errors that could cause data corruption of a file domain include I/O errors in file system metadata, the accidental removal of a volume, or any I/O error that produces a panic Use the salvage utility as a last resort.You should first repair domain structures by using the verify utility. If that repair method is unsatisfactory, attempt to recover fileset data from backup media. Only if both methods are unsatisfactory should you employ the salvage utility.
The lengths to which we've had to go and the difficulty in performing the equivalent of a very simple fsck has made us seriously reconsider our use of the Advanced File System. The file system, when damaged, is non-trivial to fix, if it is indeed fixable. The verify command did NOT work, and the salvage man page suggests trying a manual backup first instead !!!!
Try reading the man pages for verify and salvage. AdvFS does not need fsck.
Tell that to my four bad inodes, on which AdvFS tools have failed to remove.
``Insanely idiotic OS''??
Lay down the crack pipe and take a deep breath and read what I wrote again. Insanely idiotic OS and component licensing scheme. The licensing is insane, not the OS.
Maybe you'd have a leg to stand on if Linux supported the enterprise features that Digital UNIX does. Unfortunately, it doesn't. Example: High performance, dynamically resizable, journalling filesystem.
AdvFS cannot be fscked. This in turn has fscked me. Get the picture. If the FS breaks, you keep both pieces. Thank you Digital for such a wonderful advance in computing. Bad inodes just get to live indefinitely on your system until you copy all the files to a different partition (dump and tar choke), and reformat. We haven't had such advances since, well, DOS. Needless to say, we are going back to UFS for our OSF needs, which is only quite a bit slower than ext2. But at least when it breaks we can fix it.
OSF also doesn't ship with a reasonably modern interface. CDE, MWM, and TWM are simply not enough.
It certainly has its niche for ultra high end computing where user interfaces are just not viewed as important. But the OS is chock full of holes. And I run into them from time to time.
Example. glibc call for wordexp does a complete shell-like expansion in C. Libc shipped with OSF does the expansion by shelling a command to ksh. Why should libc depend on ksh for its integrity ??
Example. Recursive scripts fill up the process table and lock the system in OSF. Not so in linux.
There are also good things. CC is a really fast compiler. So is the ForTran compiler. If you want to run processes real fast, OSF is a good choice. If you want a large number of CPUs, OSF is a good choice. If you want a decent user interface, reasonable speed, and a journaled file system that can be fixed, linux is a good choice.
Alpha costs more? When's the last time you priced a top of the line Intel Xeon? I found a PIII Xeon 550/2MB priced at $ 3795.95 - and that's just the processor! (and these guys are cheap!)
Compare apples to apples. A full system with a PIII 500MHz is around $2500-$3k including a complement of RAM and hard disk space. The top alpha is ONLY available from Compaq. The fast motherboard is the XP-1000 for about $10k without the RAM or monitor. And add-ons for the Compaq alphas machines are EXPENSIVE. One generation late you can get a 21164 processor which is still fast in a machine for about $3k. Video card support under OSF or alpha linux is very poor. Unless you really need the supercomputing, it just doesn't make a lot of sense to buy an alpha. Now, if Compaq somehow made third party peripherals supportable for alphas they would sell a WHOLE lot more alphas without losing the number crunching crowd.
The alpha processors are not changing their niche in the computer market. They are ripping fast - and Dec first and now Compaq plays to the supercomputing crowd. The XP series motherboards and 21264 chips simply rip any other motherboard/chipset out there.
However, they cost too much for anyone except a supercomputing hound. If Compaq would drop Dec's insanely idiotic OS and component licensing scheme and aid linux on alphas, they might stand a chance of making a LOT of money selling hardware. As is, people buy ten times more alphas one chip generation late and run linux instead of OSF.
Anyone interested should see the linux alpha compilers available. cc is a small improvement, and ForTran is a LARGE improvement.
http://www.unix.digital.com/linux/software.htm
But still, Itanium will come out, and an Itanium box will offer slightly less than half the floating point speed, and it will cost about 1/4th of the fast alpha box from Compaq. And the alpha motherboards will still make it tough to support third party peripherals. And Itanium will dominate the 64 bit market. And Alpha will own the supercomputing market.
To get a really ergonomic mouse they ought to construct one that doesn't require use of finger or thumb muscles. Like place the buttons on the near side of the mouse. The ergonomic issue is
1) use larger muscles whenever possible
BTW, this is also why it is recommended never to rest your wrists while you type. Keeping the wrists up allows forearm movement to do some of the work of typing.
Any mouse work that requires click and drag or lots of double clicking is asking for lots of trouble.
Various groups have now shown that there are substantial interactions between sensory representational abnormalities and loss of motor function in repetitive stress injuries. The emerging hypothesis is that in this most common type of dysfunction, attention and motor action cause sensory representations to become abnormal. This in turn tweeks motor feedback, which then worsens. This in turn causes further erosion of sensory representations. And so on. In the worst cases focal dystonia can result which NEVER GOES AWAY and NEVER GETS BETTER.
Unless the underlying sensory representation is restored.
But the bottom line is that once your hands begin to feel weird in a repetitive stress injury, continued use will lead you down the spiral towards loss of control. Stop, and take a break. And fairly soon effective sensory training will be able to reverse the negative effects of the repetitive movement.
The article states that stroke victims are at risk of thrombosis, yet these devices are made out of glass capillaries, which is thrombogenic.
I don't think they are made out of glass capillaries. Glass is not so biocompatible. I would be extremely surprised if the glass is not insulated with a polymer (like parylene) or is actually a plastic tube to begin with.
Lastly, the critic mentioned in the article is correct: long-term studies are needed. These lil things can become lodged in joints, escape into the bloodstream, etc. They might be carcinogenic. Who knows? Its time for a randomized trial.
The concept of these lil things escaping the muscle capsule is pretty out there. The more likely confound is encapsulation in the form of a foreign body reaction, and even that would not be so bad, and very undo-able (^_). The risks are pretty low, and the potential gain is quite high. I think they will gain approval without much trouble.
That said, it IS a step in the right direction. Current science is 99% focused on the chemistry of neurology...theyre forgetting the electrical aspects of it. Not until doctors looked at the heart as an electrical organ did they really do anything that can be described as "cardiology," and the brain is every bit as electrical as the heart.
So nitric oxide treatment, bypass surgery, and angioplasty are not really cardiology ??
Really, cardiology, like other disciplines, works in a gradient descent method. Techniques that lead to improvements in short time frame get worked on the most. Electrically, as well as chemically. There are already numerous electrical techniques at work in brain pathologies - like stimulating electrodes for Parkinson's patients.
I work for a company closely affiliated with the one who developed this device (about a 7 year process). Here's two URLS to the people involved, although information there is sparse. http://www.advancedbionics.com These people also market Cochlear implants that restore hearing to totally deaf patients. http://www.aemf.org This is the non-profit R&D foundation that has actually done most of the development. If anyone really wants to know more about these devices, please contact them... Dave Marsh Systems Engineer Medical Research Group, Inc.
Interesting. Gerald Loeb was part of the team that developed the cochlear implant technology while he was at UCSF some 15-20 years ago. That technology led to patents that form the basis of the first generation cochlear implants used at advanced bionics. The muscle implants should be fairly easy. Muscle tissue is not so hard to implant, and much more difficult things are already being done in nervous system tissue.
Our number of senses and limbs is fairly predetermined. Early sensory pathways are guided by cues that appear to be mostly genetic. Within each area the potential for change is enormous. People have little to no problem adapting to using cochlear implants, for example. In cases in which implants allow movements of devices, I would be surprised if the person didn't feel an attachment of sorts to the device they are moving. Heck, if you use a pencil a lot you feel as though you are touching the paper with the pencil tip instead of feeling the vibrations that move through the pencil. In golf you feel as though you are striking a golf ball, whereas you are really only feeling a club.
Is the same true of senses? If I attached a little I/O port into my brain would it eventually be integrated as a 6th sense or are out brains to preprogramed for this to happen?
It is very unclear on how one would attach an I/O port to the brain. Are you just going to put a serial port on the brain and plug in ?? Implants rely on causing actions based on activity in existing neurons. Sensory implants activate sensory afferents.
This is no surprise at all. The surprise is how
many of these exist that are not publicized.
And I doubt it stops there. I would think every
major ISP is tracking hits from every user,
particularly to correlate web purchases with
web site visits. That is very valuable info. Your
ISP can record EVERY MOVE YOU MAKE if they are
so inclined. And it is worth a lot, so in business
terms they would be idiots not to track you.
There is a little oversight on the technology by this professor.
There are already recent high profile experiments by John Chapin's group in Philadelphia in which rats are trained to control lever arms by activating parts of their brains in which electrodes have been placed.
Other cortical implants in primary motor cortex of humans show a lot of promise for quadrapelegics to be able to control robotics around them. The last I saw humans were quite able to make LED displays light up in certain patterns, and it is a small step from there to controlling robotics. In both cases a small number of bits of information is measured from the CNS and converted to simple movements.
But the converse, computer control of humans, is in a much much more primitive state. Likening the ability of implanted chips to enable the lights to be turned on, to giving the computer control of your limbs is at this point a crack pipe dream, and there is far more advanced work going on than this press release. For example, advanced bionics makes cochlear implants that allow high fidelity sound reception. There is comparable work on retinal implants at a much less advanced state.
But the problem is much more advanced, since the human arm has more than a few bits worth of degrees of freedom, and its natural control from motor cortex to movement is not even close to fully understood.
As for controlling emotion, that would be a far simpler problem, since there are a few well localized centers of neuromodulators, and implanted stimulating electrodes already exist. It will be fairly easy to make someone hostile and aggressive, or sleepy, or passive, or edgy. Making
someone alert and intelligent would be much harder. There would be a lot of unreversible psychosis caused in the pilot experiments, so don't expect them anytime soon.
> So, we're out with CUPS. Now if we could just
> get Adobe to give us some Linux versions of
> their tools, life might be (almost) perfect.
Adobe tools suck rocks. Their acroread is poorly
implemented in a lot of ways. They patent their
standards (like pdf) to prevent others from being
able to make pdfs without paying the man.
They embrace and extend their proprietary standards (PS and PDF) to keep open source tools
from being able to work with full functionality.
I do a lot of PS hacking with open source tools,
and Adobe has about the worst attitude towards
open source PS hackers as you can get.
First, I wish you the best in your search for
painless typing. Be aware, however, that many people who experience pain while typing do not
have carpal tunnel, and most of them finally resort to fewer and shorter bouts of typing.
If it feels like your fingers are about to tie themselves in knots - take a break.
Don't rest your wrists on anything. Keep them in the air and force your forearms to do some of the work. Bigger muscle groups are much less susceptible to repetitive stress injuries.
And a last thought to those who may have severe problems. A retraining strategy that focusses on sensory retraining of the hands has a very high efficacy after diligent training for 2-3 months. Many people who thought they would never use their
hands the same way again now can. This therapy was pioneered by Nancy Byl and Mike Merzenich at UCSF.
It supposes that a large portion of the problem is
sensory reorganization during heavily attended heavily repeated motor actions, and the lack of sensory feedback wreaks havoc with motor control.
email me if you are interested in more info.
> I've been stuck with Word for years b/c my the
> equations and graphs in my existing
> documents never came out right in other on
> other word processors and spreadsheets....
> until StarOffice, which seems to work pretty
> well.
Nothing beats TeX and its derivatives for ease
of use in making equations.
This is not to say you shouldn't try Word,
WordPerfect, StarOffice... but Knuth developed
TeX BECAUSE existing equation editors sucked
rocks. And still, TeX has the easiest to use
and most flexible implementation. And it was
made to make documents - not made to draft
a one page letter.
> This isn't troll talk - it is slow compared to
> MS Office. On Linux it takes about a minute to
> load, on windows, around about the same. ewww.
> MS Office takes like 2 seconds (per app).
Windows generally preloads the libraries
for Office into RAM so that bloated pigs
like Word will start faster. Try closing the
Office toolbar and then opening Word and
look at the difference in load time.
This was a marketing stunt.
It says next to nothing about security
of linux or NT except that if you set up
your machine like a complete total moron
you are likely to get cracked.
The funniest thing about this stunt is that
PCWeek has just badly eroded its credibility.
It most certainly is not a license to practice.
To verify this, try going around and asking
your PhD professors of engineering how many of
them are licensed PEs.
And for electrical and/or computer engineering
it is more a rarity than common among professionals to have a PE.
For MOST things related to EE or CS, the PE
will be viewed in a similar light as an MSCE.
It is kinda nice to have, but doesn't tell you
very much about the quality of the job
candidate.
In addition, the PE test (the first phase of
certification) is almost exclusively based on
the core curriculum for Civil and Mechanical
Engineers. Yes there is an electrical engineering
section - and it is almost exclusively based
on things taught in the first intro EE course
that most civil and mechanical engineers are
forced to take.
If I were a graduating EE or CS major, I would
forget about the PE. As a mechanical or civil
engineer, I would probably do it.