How would you know that the calculations were wrong?
You know later. It's similar to things like the EPIC arch where you might (for example) execute both sides of a branch in parallel, but you don't 'commit' the results until you get the results of the predicate bits, at which point you'll throw away one side's results and commit the other side's.
Predicated execution is used to decrease the occurrence of branches and to increase the speculative execution of instructions. In this feature, branch conditions are converted to predicate registers which are used to kill results of executed instructions from the side of the branch which is not taken.
Honestly, I hold out little hope for Congress changing their reading habits (or lack thereof). *However*, having a viewcvs kind of interface (annotate/diff/etc through a web interface) *would* make it far easier for people to see who's making what changes and holding them accountable later on.
Imagine being able to have people point to changes (for instance, on their blogs, or in emails) made in a bill the same way we can point to code diffs now thanks to tools like view{cvs,svn} - it'd be great!
Not to go all black-helicopter, but keep in mind that lots of amazing things are done outside of the public spotlight. Some of the smartest people I've known ended up at places like the NSA - no, they don't get in the papers, but no, that doesn't mean they're not doing great things.
The task manager has to be somewhat end-user focused, and tries to be smart about not letting you kill things the typical user probably doesn't want to actually kill ("my compute is slow, i'll start killing random processes until it speeds up!")
If you want more "just kill it already", and want to stay in a GUI (since kill/taskkill/etc exist), I'd recommend Process Explorer, where most of those checks that try to help out the typical user aren't done.
I use those links to explain security backports all the time, even for clients. It's well-phrased and then tend to "get it" once it's phrased the way it is in that FAQ
AFAICT there's only a problem if you require the source code to be distributed. Can we then just make a GPL v2.5 (or whatever) that doesn't require source code availability, but instead says that one of 2 things must be provided: 1) source code *or* 2) rot13 of source code.
Since the source code is no longer required under this approach, we can still get the spirit of the GPL while working around MS's clause.
we have logic and analog sim tools now. Rather than evaluating on actual fpga's, hook in the GA logic with logic simulators (ikos would be happy to help:) and once the logic "stage" produces a set of good designs, move onto a later stage where fitness functions are then evaluated with (slower) analog simulations (min/max/typ for V/T/etc initially, but adding more test cases perhaps at a separate third stage would be easy enough) which could then "harden" the designs.
Yes, it's silly to trust something eval'd on actual fpga's when we can do so much better in simulations.
Course, it'd suck if you decided to switch your fpga provider (altera/actel/xilinx/etc) and couldn't just toss your verilog into synplify and click synthesize:)
IMHO, there's a ton of bargains on the market at the moment. As much as I love Bob Young and RH (just had a chat with Bob in the hall today), RH is not one of the extreme bargains that are out there. Combine "bargain rate" with "volatility" into your formula of choice and you'll get a nice selection of 5-10 companies (perhaps Dell) that would be solid portfolio choices. Bob probably believes RH is still a potential bargain and hence his lack of total sell-off (assuming he's not hitting SEC limits) and hence he's still holding onto large chunks.
The economy is in the tank, and it is *totally* a buyer's market. I continue to drool over the choices I have to toss my cash into:)
with the acquisition of cygnus, glibc is now basically a redhat-internal project these days, Ulrich being @redhat.com and all.
Think Taco wants to pimp the work of RH instead of the distro he loves that benefits from it? The kernel's leader is outside RH (although obviously many, if not most of the major contributers are inside RH), so slashdot can pimp the kernel without pimping RH.
Oh, and did I mention VA Linux is an RH competitor?:)
What wavelength range (in angstroms, please) are these colors? I would imagine it is what we normally tag as ultraviolet... otherwise, most remote controls would be quite bizarre for these people:)
Not at the present time, although Porivo Technologies has a sweepstakes running. This sweepstakes is currently a $2000 shopping spree from buy.com! Also, better than other sweepstakes that I've heard about, the second and third place winners even get $1000 themselves! I love great prizes:)
One of the nicest things I've noticed is that Porivo's client doesn't actually nail your CPU (and subsequently your power bill:) like other clients. Their main focus appears to be on Web site performance testing. Since I'm a big-time gamer that's much more interested in keeping my CPU free and latencies low, the very slight amount of bandwidth needed doesn't affect me. Combined with the bonus of a nicer sweepstakes prize, they've got me hooked:)
boot a software RAID meaning boot a kernel on a s/w raid partition? If so, it'll only work with the 0.90 RAID (not the 0.4x that ships in kernels, at least AFAIK) and it will only work with the lilo with RH 6.1 (unless you add a disk= section for md0 geometry).
Slashdot Mirror
IMHO, rejecting Google App Engine at this stage is a bit myopic, but it's your choice. :)
Among the more open-standards-focused cloud offerings, there is Red Hat's OpenShift
- https://www.openshift.com/
- https://www.openshift.com/developers/java
- https://www.openshift.com/developers/pricing
- http://www.jboss.org/openshift.html
likely easier to swallow if you rephrase it as "lower rates for areas with lower costs of living"
Certainly in the United States, people are used to rates being higher in places with higher costs of living (NYC, Bay Area, etc).
Alternatively, realize that it already applies if you rephrase it as "lower rates for poorer states" (and you get a rhyme as well!)
How would you know that the calculations were wrong?
You know later. It's similar to things like the EPIC arch where you might (for example) execute both sides of a branch in parallel, but you don't 'commit' the results until you get the results of the predicate bits, at which point you'll throw away one side's results and commit the other side's.
http://en.wikipedia.org/wiki/Explicitly_parallel_instruction_computing
Predicated execution is used to decrease the occurrence of branches and to increase the speculative execution of instructions. In this feature, branch conditions are converted to predicate registers which are used to kill results of executed instructions from the side of the branch which is not taken.
http://en.wikipedia.org/wiki/Speculative_execution
You can tell by the subject that GP was referring to Yuri with the "Wouldn't he qualify as first geek in space?" comment...
For the curious, it was an audio gain issue. Details on Rob Chambers' blog:
2 479.aspx
http://blogs.msdn.com/robch/archive/2006/07/29/68
Honestly, I hold out little hope for Congress changing their reading habits (or lack thereof). *However*, having a viewcvs kind of interface (annotate/diff/etc through a web interface) *would* make it far easier for people to see who's making what changes and holding them accountable later on.
Imagine being able to have people point to changes (for instance, on their blogs, or in emails) made in a bill the same way we can point to code diffs now thanks to tools like view{cvs,svn} - it'd be great!
Not to go all black-helicopter, but keep in mind that lots of amazing things are done outside of the public spotlight. Some of the smartest people I've known ended up at places like the NSA - no, they don't get in the papers, but no, that doesn't mean they're not doing great things.
:)
Don't make that (il)logical leap
The task manager has to be somewhat end-user focused, and tries to be smart about not letting you kill things the typical user probably doesn't want to actually kill ("my compute is slow, i'll start killing random processes until it speeds up!")
o rer.html
If you want more "just kill it already", and want to stay in a GUI (since kill/taskkill/etc exist), I'd recommend Process Explorer, where most of those checks that try to help out the typical user aren't done.
http://www.sysinternals.com/Utilities/ProcessExpl
% TZ=GMT perl -le 'print scalar localtime 2**31-1'
Tue Jan 19 03:14:07 2038
% TZ=GMT perl -le 'print scalar localtime 2**31'
Fri Dec 13 20:45:52 1901
Tue Jan 19 03:14:08 2038 GMT isn't representable by a 32-bit signed time_t like many of us have now.
http://www.debian.org/security/faq#oldversiont p://www.debian.org/security/faq#version
ht
I use those links to explain security backports all the time, even for clients. It's well-phrased and then tend to "get it" once it's phrased the way it is in that FAQ
AFAICT there's only a problem if you require the source code to be distributed. Can we then just make a GPL v2.5 (or whatever) that doesn't require source code availability, but instead says that one of 2 things must be provided: 1) source code *or* 2) rot13 of source code.
Since the source code is no longer required under this approach, we can still get the spirit of the GPL while working around MS's clause.
At least, IMHO, IANAL, YMMV, etc.
we have logic and analog sim tools now. Rather than evaluating on actual fpga's, hook in the GA logic with logic simulators (ikos would be happy to help :) and once the logic "stage" produces a set of good designs, move onto a later stage where fitness functions are then evaluated with (slower) analog simulations (min/max/typ for V/T/etc initially, but adding more test cases perhaps at a separate third stage would be easy enough) which could then "harden" the designs.
:)
Yes, it's silly to trust something eval'd on actual fpga's when we can do so much better in simulations.
Course, it'd suck if you decided to switch your fpga provider (altera/actel/xilinx/etc) and couldn't just toss your verilog into synplify and click synthesize
Isn't the 25th the XP launch date? Lots of fronts for open source fighting on that day.
IMHO, there's a ton of bargains on the market at the moment. As much as I love Bob Young and RH (just had a chat with Bob in the hall today), RH is not one of the extreme bargains that are out there. Combine "bargain rate" with "volatility" into your formula of choice and you'll get a nice selection of 5-10 companies (perhaps Dell) that would be solid portfolio choices. Bob probably believes RH is still a potential bargain and hence his lack of total sell-off (assuming he's not hitting SEC limits) and hence he's still holding onto large chunks.
:)
The economy is in the tank, and it is *totally* a buyer's market. I continue to drool over the choices I have to toss my cash into
one theory:
:)
with the acquisition of cygnus, glibc is now basically a redhat-internal project these days, Ulrich being @redhat.com and all.
Think Taco wants to pimp the work of RH instead of the distro he loves that benefits from it? The kernel's leader is outside RH (although obviously many, if not most of the major contributers are inside RH), so slashdot can pimp the kernel without pimping RH.
Oh, and did I mention VA Linux is an RH competitor?
Just some thoughts.
it's a great thing to stream while you're chillin' at home, too ::)
OpenPGP != OpenP2P
how many zeroes in a billion? :) heh
Definitely a family search engine when the first result on a search for "sex" is a ministry :)
:) to intenseromance.com... maybe not so much of a family engine after all?
Course, link #11 is a broken link (shame on them
nevermind... although "between red and green" doesn't help much
What wavelength range (in angstroms, please) are these colors? I would imagine it is what we normally tag as ultraviolet... otherwise, most remote controls would be quite bizarre for these people :)
Not at the present time, although Porivo Technologies has a sweepstakes running. This sweepstakes is currently a $2000 shopping spree from buy.com! Also, better than other sweepstakes that I've heard about, the second and third place winners even get $1000 themselves! I love great prizes :)
:) like other clients. Their main focus appears to be on Web site performance testing. Since I'm a big-time gamer that's much more interested in keeping my CPU free and latencies low, the very slight amount of bandwidth needed doesn't affect me. Combined with the bonus of a nicer sweepstakes prize, they've got me hooked :)
One of the nicest things I've noticed is that Porivo's client doesn't actually nail your CPU (and subsequently your power bill
You may remember Porivo from their earlier coverage, here on Slashdot
Testing? What's that? If it compiles, it is good. If it boots up it is perfect -- Linus Torvalds
boot a software RAID meaning boot a kernel on a s/w raid partition? If so, it'll only work with the 0.90 RAID (not the 0.4x that ships in kernels, at least AFAIK) and it will only work with the lilo with RH 6.1 (unless you add a disk= section for md0 geometry).
.11, .12, .13) and www.redhat.com/~mingo/ has patches for 2.2.14 (raid-2.2.14-B1) and 2.3.40 (ibc-ext2-raid-2.3.40-N1)
kernel.org/pub/linux/daemons/r aid/alpha/ has the 0.90 raidtools and patches for 2.2.11 (works on
2003, actually, since 2001=3*23*29
19:59:59, actually :) and yes, still prime :)