Slashback: Benchmarks, Sobig, Blob 285
A good excuse to file purchase orders, too. Eug writes "Writing in this Ars thread, Craig Hunter of NASA gives details about his much-quoted dual-G5 Power Mac benchmarks listed here. This should answer some of the questions posed around the net about the methodology and potentially the validity of his benchmarks."
The lines between viruses and spam is thin enough already. Joe Stewart writes "There have been a lot of news stories lately about how Sobig and spam are tied together. I actually revealed this in a paper two months ago. Now with the widespread Sobig.e, it seems to have become a topic again. However, the major antivirus companies have once again left out the whole story - most of them currently rate Sobig.e as 'low damage.' This is because they haven't fully understood how the real payload of Sobig.e is delivered. I've written a followup paper describing the entire mechanism that Sobig.e uses to facilitate spam, identity theft and bank fraud. Sobig has evolved, and it is much harder to stop than before."
Is this the beginning of a long goodbye? inertia@yahoo.com writes "Microsoft has updated their Mactopia Web Site to include a section on Virtual PC. It's taken them since February 2003 to do this. On the site, they mention, 'In August 2003, Virtual PC for Mac will be available through standard Microsoft channels of distribution.' So it looks like they aren't killing it after all."
Simplicity itself is a nice ideal. webword writes "Building Accessible Websites by Joe Clark is now available online. As you might recall, Joe was interviewed on Slashdot back in December. Good stuff if you care about accessibility."
Not yet billions and billions served, but getting there. nzilla writes "The Internet Book List, which announced its creation earlier this year on /. has now reached 10,000+ entries and is still going strong. The Internet Book List (IBList) strives to be the IMDb of books. IBList is maintained exclusively by volunteers around the world."
Girlfriends drive strange endeavors. ceejayoz writes "This interesting article on MSNBC.com details the Degree Confluence Project - a project to gather a photographic record of the points on Earth where latitude and longitude lines meet. The article has links to some of the more interesting points. The project's website also has an interesting map showing all the completed confluence points."
We mentioned this project quite some time ago, and it's progressed quite a bit since then.
Uh, sir, you have some blubber on your collar there. Scoria writes "Chilean scientists have determined that a 12-meter mass of flesh discovered recently on a Pacific beach is actually a sperm whale, not an obscure 'giant octopus' as many researchers speculated. Scientists performing research at the Museum of Natural History in Santiago were the first to develop this conclusion after observing the presence of dermal glands unique to the species."
Code that pays tribute to the money in television. mondainx writes "Following(?) in the footsteps of Linksys, Tivo has made their source available for versions 2.0 through 4.0. Get the GPL source here. Sweet!"
That's about as fair as it gets re: G5 speed. (Score:5, Informative)
Here are Craig's final numbers, as posted on Ars's website:
dual G4-1GHz Xserve (single CPU only): 105
dual G4-1GHz Xserve (both CPUs): 207
dual G4-1.25GHz PowerMac (single CPU only): 129
dual G4-1.25GHz PowerMac (both CPUs): 256
dual G5-2GHz PowerMac (single CPU only): 254
dual G5-2GHz PowerMac (both CPUs): 498
single P4 2GHz: 192
single P4 2.66GHz: 255
single P4 3.2GHz (extrapolated): 307
These numbers seem entirely reasonable to me. A single G5/2GHz G5 is approximately equivalent to a single P4/2.66GHz. This rings true to me -- Intel has never been known to squeeze every last bit of performance out of every chip, instead opting to continually push for higher and higher raw MHz. Thus, on a purely MHz/performance basis, Apple wins (as has been the case for years.)
However, in the dual-processor arena, things get muddier. Intel should have dual 3.4GHz Xeons by the time Apple's G5s are shipping. In raw performance, based on these numbers, the Xeon will have an edge over the G5. Plus, it will be priced lower... I priced a dual Xeon 2.4GHz with 1GB of RAM and a 120GB hard drive for a company that is buying a game server from us, and even with a 1U form factor (which is more expensive than a standard desktop case), the price came to $1705... a bit more than half the cost of the dual G5/2.0GHz. There is no question that the dual Xeon will outperform the G5 both in terms of raw performance and cost. The P4, however, doesn't have much edge over the G5 except for the cost.
For most of us, who are probably sitting on machines around 1-2GHz, almost all of the machines above, including the P4/2.66 and a single G5, will be a healthy upgrade. Despite Apple's high price point, I for one am happy to see them get back into the game... and I'm happy to see Intel have some real competition. A big thanks to Craig for doing the benchmarks... I'm sure this is just the first of many arguments about which machine is better!
Re:Poor Sperm Whales (Score:5, Informative)
The sperm whale has a huge reservoir of liquid in its head, with an oily sheen and a translucent, pale white color. The liquid solidifies under pressure (when the whale dives); current scientific thinking has it that the change in the liquid's density helps the whale adjust its buoyancy.
When early whaling crews first killed one of the beasties and slit it open, they encountered the oily stuff in its head but didn't know what it was....being sex-starved sailors, they jumped to conclusions, called the substance spermaceti, and named the whale after his unique feature: gallons and gallons of sperm in its head!
Re:Whales (Score:5, Informative)
ISBN.nu (Score:5, Informative)
Re:That's about as fair as it gets re: G5 speed. (Score:5, Informative)
Re:Bad Compilers for Apple G5 (Score:1, Informative)
Re:Poor Sperm Whales (Score:5, Informative)
Actually, I'm sure it was natural philosophers who coined the actual term "spermaceti"; the sailors called it jizz or cum or whatever one called the male ejaculate, circa 1600.
From Thomas Beale's "The Natural History of the Sperm Whale," 1840:
"What spermaceti is," (says Sir Thomas Brown, in his work published in 1686, third book, chap. xxv. p.139,) "men might justly doubt, since the learned Hofmannus, in his work of thirty years, saith plainly nescio quid sit, and therefore need not wonder at the variety of opinions, while some conceived it to be flos maries, and many a 'bituminous substance floating upon the sea.' That it was not the spawn of the whale, according to vulger conceit or nominal appellation, philosophers have always doubted, not easily conceiving the seminal humour of animals should be inflammable, or of a floating nature.
So you see, even before sperm were known to people, spermaceti (though probably not known by that name) was considered to be somehow involved with the reproduction of the whale, in much the same way that human semen was known to be involved with the reproduction of humans even though its exact nature was unknown.
Re:What about Virtual PC for Windows? (Score:2, Informative)
TiVo a leader, not a follower (Score:3, Informative)
TiVo has been a leader in releasing the required source and a little more; they also provide the compiler toolchain used to build the kernel (which is not required, but a nice touch, since it allows users to easily build additional binaries with the same toolchain).
Re:The IBlist is kind of poor (Score:3, Informative)
Hopefully this will.
Actually, the hobbit was listed as 1937. I would think if they were gonna make an IMDB kinda thing, they would have added some more of the cool stuff from the start.
Re:That's about as fair as it gets re: G5 speed. (Score:3, Informative)
Re:And in other news... (Score:5, Informative)
Re:Nasa G5 Benchmarks (Score:3, Informative)
Not only did the score of the G5 with both CPUs make me say "Holy shit" out loud in front of my comp (seriously!), but it also kicks the piss out of the P4! So, Apple does have one of the fastest machines around!
so two 2.0 GHz cpus are faster than a single 3.2 GHz cpu? i fail to see why this is surprising.
Wrong on Tivo (Score:3, Informative)
Although they give you the source code for the kernel, that doesn't mean that you can change it -- not and still expect your Tivo to work, anyway. The boot PROM (think "BIOS") in the Series 2 checks that the kernel it's booting is signed with Tivo's key. Then, a program in the initrd checks everything on the root partition to see that it's not modified, either. With the initial software that came out with the Series 2, it was possible to get around this by setting BASH_ENV as a kernel option in the drive's boot page, but they "fixed" that in the next revision.
Now, to hack a Series 2, you have to either stick to old software, play two-card monte with the kernel, or reprogram the PROM -- which requires desoldering it from the motherboard, since it can't be done in software from the Tivo.
I've done the kmonte thing, and it works well -- in that context, the kernel source is actually useful, since you can boot anything you like as the second kernel. But you still have to devote a couple partitions to the old software (after first getting a copy of it) that allows the BASH_ENV hack. Doubtless this will not work once there's a Tivo Series 3.
Oh, and it probably goes without saying, but Tivo's GPL'ed software doesn't include the main applications -- the bits that actually handle TV.