From the linux-2.4.20-18.9 (Red Hat) kernel source:
find . -type f -exec grep "Return the base of the allocated space" {}/dev/null \;./arch/ia64/sn/io/ate_utils.c: * Return the base of the allocated space.
Keyboards and mice work fine (I'm using them now). But after that, there's a lot of configuration and little documentation when something goes wrong. Setting up my Zaurus is easier under RH9 than 8, but it's still non-obvious how to do so.
With usb-storage, there's the ever popular:
usb-uhci.c: ENXIO 80000480, flags 0, urb f3234e80, burb edfae580 usbdevfs: USBDEVFS_CONTROL failed dev 4 rqt 128 rq 6 len 18 ret -6
Which is about as descriptive as "PC LOAD LETTER".
USB support is typically flaky, sometimes working on some chipsets, but not working on others.
Firewire support can crash your machine if you're not careful, and hotplugging support is all but undocumented unless you're a kernel developer.
Take a look at the Neuros. It's a bit bigger than the ipod, but it's backpack based. I have a 128M flash backpack I can use while exercising, while using the 20G backpack while in the car. It plays Vorbis and MP3 files quite nicely and will also broadcast over FM to a variety of channels (YMMV outside the US).
Linux friendly (there are 3 apps to sync over Linux) as well. Much of the Linux and Vorbis work are being done by Xiph.
I actually found a good use for budweiser. Not for drinking, but for cooking. You can make a real good gravy using 1/2 part bud and 1/2 part stock (chicken or beef).
Otherwise, make mine Guiness or Sam Adams. Aside from unfiltered beers or home brew, beer does not contain any yeast, and the carbonation is a result of injecting CO2 into the beer before capping it. You'll know if there's yeast in it if you find crunchies at the bottom of the bottle.
Oh yea, beer is just mostly empty carbs. Almost no nutritional value. At best, I'd put it down in the same category as 'snack food'.
If the code is as ingrained as SCO is leading us to believe, then the changes will be neither minor nor quiet. SCO is most likely watching the kernel mailing list to see what's going on, and will see if there are any substancial changes to what they think is theirs.
The code would then have to be tested extensively to make sure it works at least as well as the old code.
In addition, anyone who saw the SCO code would be tainted, preventing them from being able to make alternate code to do the same thing.
According to SCO in MozillaZine, this is a contract dispute between IBM and SCO. If so, then Linux users have nothing to worry about, as SCO would have to tell Linux users what the offending code they own is. At that point, it can be replaced.
Even so, SCO (Caldera) has been releasing various forms of Linux for years. The 2.4 tree has been out for 2.5 years (early '01). If Caldera has been distributing this code in their release for 30 months without knowing it, then they must have some very poor programmers. One would hope a vendor would know what they're selling, right? I could understand if it were 6 months or a year, but this is a bit strange.
Because every time the community thinks they know what SCO is talking about, they change their tune. First it was changes only in IBM's distribution, then it was in Red Hat, then it was in the stock 2.4 kernel, now it's in every kernel since 2.4.
It's NUMA, it's RCU, it's JFS, it's low-level subroutines, but they won't tell you exactly what it is. How can you change something when you don't know what needs changing?
Uhm...you misread. They don't have any of those repair kits NOW. NASA is looking at what to bring up in case this happens again. What is light enough, what will handle the most general case, what can be done by astronauts in EVA suits, etc.
NPR had a report last thursday covering the possibilities of repair in space. There's a lot of options, from filling the wing cavity with heat-resistent foam to wrapping the wing in titanium which will burn off during reentry (like the heat shields of Mercury, Apollo, etc.).
It was the prevailing theory of the media. NASA didn't have evidence either way, but now that they've done the testing and looked at the results, it's now the best theory as to what happened.
The failure of VA Linux doesn't mean much. Consider the success of Red Hat. VA Linux's problem (yes, I used to work for them) was that by the time they were putting out hardware, Linux ran on just about everything. Why buy a VA Linux server for a 50% premium when you can buy the same style hardware from IBM and know Linux will work on it?
We use Blue Socket boxes behind our WAPs, so while anyone can get an IP address from our WAP, you won't be able to get anywhere until you authenticate (via SSL). Since the wireless network is outside our firewall, you have to either use a VPN or SSL-web access to get your e-mail.
It's sad really. Public places have been removing chilled water fountains and replacing them with either warm water fountains or selling bottled water (that's cold).
At home, I buy a 24-pk of bottles for cheap, then refill them from my filtered tap. Refil 5-10 times until the bottle gets crusty.
A few years ago I replaced an SGI cluster with a Linux cluster. Due to the reduced cost of hardware and software (they even bought from VA Lin^H^H^HSoftware), they paid less to purchase the Linux cluster than the yearly lease fee of the SGI cluster.
And the Linux cluster had at least twice (probably 3-4x) the performance.
Slashdot Mirror
From the linux-2.4.20-18.9 (Red Hat) kernel source:
/dev/null \; ./arch/ia64/sn/io/ate_utils.c: * Return the base of the allocated space.
find . -type f -exec grep "Return the base of the allocated space" {}
You forgot Darlhat
'an historic' is one of many rules-meant-to-be-broken in English:
http://www.bartleby.com/64/C003/001.html
Keyboards and mice work fine (I'm using them now). But after that, there's a lot of configuration and little documentation when something goes wrong. Setting up my Zaurus is easier under RH9 than 8, but it's still non-obvious how to do so.
With usb-storage, there's the ever popular:
usb-uhci.c: ENXIO 80000480, flags 0, urb f3234e80, burb edfae580
usbdevfs: USBDEVFS_CONTROL failed dev 4 rqt 128 rq 6 len 18 ret -6
Which is about as descriptive as "PC LOAD LETTER".
USB support is typically flaky, sometimes working on some chipsets, but not working on others.
Firewire support can crash your machine if you're not careful, and hotplugging support is all but undocumented unless you're a kernel developer.
Anyhting having to do with USB or Firewire support
Take a look at the Neuros. It's a bit bigger than the ipod, but it's backpack based. I have a 128M flash backpack I can use while exercising, while using the 20G backpack while in the car. It plays Vorbis and MP3 files quite nicely and will also broadcast over FM to a variety of channels (YMMV outside the US).
Linux friendly (there are 3 apps to sync over Linux) as well. Much of the Linux and Vorbis work are being done by Xiph.
I actually found a good use for budweiser. Not for drinking, but for cooking. You can make a real good gravy using 1/2 part bud and 1/2 part stock (chicken or beef).
Otherwise, make mine Guiness or Sam Adams. Aside from unfiltered beers or home brew, beer does not contain any yeast, and the carbonation is a result of injecting CO2 into the beer before capping it. You'll know if there's yeast in it if you find crunchies at the bottom of the bottle.
Oh yea, beer is just mostly empty carbs. Almost no nutritional value. At best, I'd put it down in the same category as 'snack food'.
If the code is as ingrained as SCO is leading us to believe, then the changes will be neither minor nor quiet. SCO is most likely watching the kernel mailing list to see what's going on, and will see if there are any substancial changes to what they think is theirs.
The code would then have to be tested extensively to make sure it works at least as well as the old code.
In addition, anyone who saw the SCO code would be tainted, preventing them from being able to make alternate code to do the same thing.
According to SCO in MozillaZine, this is a contract dispute between IBM and SCO. If so, then Linux users have nothing to worry about, as SCO would have to tell Linux users what the offending code they own is. At that point, it can be replaced.
Even so, SCO (Caldera) has been releasing various forms of Linux for years. The 2.4 tree has been out for 2.5 years (early '01). If Caldera has been distributing this code in their release for 30 months without knowing it, then they must have some very poor programmers. One would hope a vendor would know what they're selling, right? I could understand if it were 6 months or a year, but this is a bit strange.
Would it measure ego in ESRs, RMSs, SCOs, or Perens?
What's the conversion rate between those units anyway?
A good portion of the USB subsystem seems to be the result of a college project.
Try finding out what this means:
usbdevfs: USBDEVFS_CONTROL failed dev 10 rqt 128 rq 6 len 18 ret -110
Google has lots of questions asking, but noone answering.
Because every time the community thinks they know what SCO is talking about, they change their tune. First it was changes only in IBM's distribution, then it was in Red Hat, then it was in the stock 2.4 kernel, now it's in every kernel since 2.4.
It's NUMA, it's RCU, it's JFS, it's low-level subroutines, but they won't tell you exactly what it is. How can you change something when you don't know what needs changing?
For the same reason Clinton is blamed for this recession.
Right. You give $300 to your daughter and then give $95,000 to your Rich Uncle Moneybags.
But were weren't talking about now. We're talking about 4 years ago.
You mean a slanted way of saying "it's about time we paid off the debt incurred by Reagan".
Just because you have some extra cash doesn't mean you immediately have to start giving it all back.
If I have extra money at the end of the month, I don't go buy a big-screen TV, I put extra money to my car loan/mortgage/credit cards.
It'll be a sad day when George AWOL Bush gets a carrier named after him.
Uhm...you misread. They don't have any of those repair kits NOW. NASA is looking at what to bring up in case this happens again. What is light enough, what will handle the most general case, what can be done by astronauts in EVA suits, etc.
NPR had a report last thursday covering the possibilities of repair in space. There's a lot of options, from filling the wing cavity with heat-resistent foam to wrapping the wing in titanium which will burn off during reentry (like the heat shields of Mercury, Apollo, etc.).
It was the prevailing theory of the media. NASA didn't have evidence either way, but now that they've done the testing and looked at the results, it's now the best theory as to what happened.
The failure of VA Linux doesn't mean much. Consider the success of Red Hat. VA Linux's problem (yes, I used to work for them) was that by the time they were putting out hardware, Linux ran on just about everything. Why buy a VA Linux server for a 50% premium when you can buy the same style hardware from IBM and know Linux will work on it?
We use Blue Socket boxes behind our WAPs, so while anyone can get an IP address from our WAP, you won't be able to get anywhere until you authenticate (via SSL). Since the wireless network is outside our firewall, you have to either use a VPN or SSL-web access to get your e-mail.
It shouldn't be. That's why you have air circulation.
It's sad really. Public places have been removing chilled water fountains and replacing them with either warm water fountains or selling bottled water (that's cold).
At home, I buy a 24-pk of bottles for cheap, then refill them from my filtered tap. Refil 5-10 times until the bottle gets crusty.
But isn't it too late for this? Haven't they been negligent?
No. Hormel has been very clear for years about using the word spam...err..SPAM(tm).
Don't you worry about my hair, let me worry about blank.
A few years ago I replaced an SGI cluster with a Linux cluster. Due to the reduced cost of hardware and software (they even bought from VA Lin^H^H^HSoftware), they paid less to purchase the Linux cluster than the yearly lease fee of the SGI cluster.
And the Linux cluster had at least twice (probably 3-4x) the performance.