Forgot your password?
typodupeerror

GPUs To Power Supercomputing's Next Revolution 78

evanwired writes "Revolution is a word that's often thrown around with little thought in high tech circles, but this one looks real. Wired News has a comprehensive report on computer scientists' efforts to adapt graphics processors for high performance computing. The goal for these NVidia and ATI chips is to tackle non-graphics related number crunching for complex scientific calculations. NVIDIA announced this week along with its new wicked fast GeForce 8800 release the first C-compiler environment for the GPU; Wired reports that ATI is planning to release at least some of its proprietary code to the public domain to spur non-graphics related development of its technology. Meanwhile lab results are showing some amazing comparisons between CPU and GPU performance. Stanford's distributed computing project Folding@Home launched a GPU beta last month that is now publishing data putting donated GPU performance at 20-40 times the efficiency of donated CPU performance."

"Couchsurfing" Travel Takes Off On the Web 145

mikesd81 writes "The Associated Press writes about a growing network of people online who've gone a step beyond hotels, hostels and even apartment swapping in their travel planning: They sleep on each others' couches. A number of Web sites have sprung up to help pair travelers searching for a place to crash and hosts with a spare couch. Sites like hospitalityclub.org, couchsurfing.com, globalfreeloaders.com and place2stay.net are often free, serving only as middlemen and offering tips on how to find successful matches. The sites are largely the creations of 20-somethings bitten with wanderlust and the hope of helping to bring together people from different cultures. They often depend on volunteer administrators to help manage the Web operations."

Comment Re:Signal to Noise ratio (Score 1) 173

I'm browsing the RFC, and it sounds like they're planning on having people's firewalls spit out IIALP messages in response to port scans, etc. In my opinion, this is a really bad idea!


You don't know the half of it. Slashdot actually port scans every computer that posts anonymously looking for proxies. If this protocol went into affect, Slashdot would be at the top of the list, continually proxy scanning the hundreds of anonymous posters it gets per minute. If ISPs implemented it, no one would even be able to route to Slashdot.

Consider the effectiveness of Slashdot's own "excessive bad posting" filter, which (to overcome DHCP) has subnet-banned class C's all over the Internet, blocking hundreds of people from posting.

I'd say that just based on the examples give above, this kind of banlisting would be a false-positive-ridden nightmare scenario.

Comment A bit offtopic, but I need to vent (Score 4, Informative) 509

Perhaps someone on the Qt/Mac or equivalent GTK project could answer this for me. Why is it that when these toolkits get ported to another other platform, be it Windows, MacOS, BeOS or what have you, they insist on looking and acting as GTK or Qt applications rather than native apps? A Qt/Windows or GTK/Windows app would be much more useful and usable to me if it used native Windows widgets and thus fit in with every other program I use.

As an example, I use gaim on FreeBSD because its tabbed interface is simply the best I've come across. I would love to use it instead of Trillian when I'm forced into using Windows. But the Windows port of gaim, which uses GTK+/Windows, works horribly. The GTK theme doesn't match my XP settings, widgets draw slowly and work clumsily (tooltips in particular seem to spontaneously appear and refuse to go away, even when the program is minimized!), and all in all it feels like a cheap Wal-Mart knockoff.

GTK+ widgets offer no benefits over standard Windows controls -- they draw slower, they don't match the environment, and Windows is just as themable as GTK is. Going back on-topic, this Qt/Mac port of Konqueror likewise eschews native widgets for the entirely out-of-place Qt look. All I can ask is Why? Wouldn't it be far easier for Qt/ and GTK/Windows or /Mac to simply wrap native widgets, rather than poorly ape them?

Comment My GOD! Can we say "security risk" (Score 1) 302

So, basically, they want you to run a keystroke logger. The logger will then report back the number of times you push each key. Are we supposed to just trust them to ignore the order? I have to really doubt the sanity (or common sense, anyway) of the person who would voluntarily give away his security like that.

Slashdot Top Deals

How many Unix hacks does it take to change a light bulb? Let's see, can you use a shell script for that or does it need a C program?

Working...