Re:craigslist.org for bay area
on
Honest Job Sites?
·
· Score: 3, Informative
I've gotten three jobs, and innumerable interviews, from craigslist. Companies that not only know about it, but also post openings there, are companies that are likely ready to hire. Highly, highly recommended.
Every is a well-known crank, with a vanishingly small understanding of technology. While I'm all for pointless, iconoclastic stands on principle, MacKiDo is basically the rantings of a brainless loon.
Sendmail? Are you thinking of the same Sendmail that I am? The group of science rocketists who've given us *this*:
R$* $: $(dequote "" $&{client_addr} $) R$-.$-.$-.$- $: $[ $4.$3.$2.$1.rbl.maps.vix.com $] R$-.$-.$-.$-.rbl.maps.vix.com. $#error $@ 5.7.1 $: 550 no access from [$4.$3.$2.$1], see http://mail-abuse.org/rbl/
?
I can't think of a anyone, anywhere, less competent to write a complex, featureful, information interchange mechanism. They've had their chance, and that's as good as they could come up with? Egad!
And then, in a brilliantly ironic twist, I can't post those sendmail.cf snippets as text, because of Slashdot's indescribable "lameness" filter. It's the irresistible force versus immovable Taco, and we all end up losing. Figures.
What's next? Pudge on John Stuart Mill's "On Liberty"? CmdrTaco dissecting "A History Of Sexuality"? The intersection of academe and Slashdot is too terrible to imagine...
The sperm whale is the largest toothed whale in the world, and likely the largest predator ever on earth. They are known to primarialy eat smaller squid (although a 13' squid is only smaller in comparison to Architheuthis), as well as fish, and the occasional marine mammal, like seals or porpoises.
Trust CNN to get the basic science wrong. If a whale has a sucker scar, it's from a desparate squid trying to escape, not from a brush with death. Look at the sizes of the things: how would a 250kg squid handle a 60 ton whale?
SPEC is not a neutral performance benchmark. It is clearly biased as it does not measure floating point performance-- it essentially measures clock speed.
So a processor that's faster will do always better on SPEC? And it doesn't measure floating point performance? Pardon me, sparky? Maybe you'd better get on the horn with IBM:
1.1ghz POWER4 708/1075 (int/FP) 1.1ghz Pentium III 427/268 (int/FP)
So what's that second number mean? Frist Post? Fried plantains? How about FLOATING POINT, you dullard? Is SPEC biased because it can be gamed by clever compiler writers? Or is it biased because it returns a result that makes you feel bad? This is an important point to realize, if anyone is still reading this: there's BitGeek, claiming that SPEC is a clockrate limited measure of integer performance, and then there're the SPEC members (no-name hype-driven marketing companies like IBM, Sun, Intel, Digital^WCom^H^H^HHP, Apple, &c) who view it as a reasonable, if imperfect, tool for measuring general microprocessor performance. Who do you believe?
A processor with a simpler instruction set can execute more instructions in parallel during a clock cycle, dipshit.
What in God's name does this mean? It's totally nonsensical, a simplification so gross as to be completely meaningless. What about the number of execution units? Will a single execution unit R3000 (about the perfect RISC microprocessor) outperform a multiple execution unit Motorola 68060 (about the perfect CISC processor)? What about the pipeline depth? And, more trenchantly, so what? What microprocessor performance is, is instructions issued and retired in a given unit of time: ie, work done.
Do you even know what RISC is? what it means?
RISC is tangential to this "discussion" because the P4 is a RISC processor (with a small number of consistently sized instructions), with an x86 decoder stage early in the instruction pipeline. In other words, it sees the putative benefits of RISC (increased performance) whilst maintaining backwards compatibility. Certainly there might be performance advantages to pulling the decoding stage and issuing 'native' P4 instructions -- the image that you seem to have, that the decoder is a separate 386 processor shrunk down and magically inserted into the die is strange and hilarious -- but what's the point? Intel gets plenty of performance from their current design, and a full reengineering of the entire project would likely not show performance benefits worth the cost. Even if your people from Planet Tinfoil-Hat view the x86 world's convincing performance superiority to be as conspiratorial myth, Steve "Reality At Arms-Length" Jobs understands it: it's why you'll never see a dollar for dollar performance comparison at a MW "bakeoff", let alone a straight-up top of the line Photoshop extravaganza.
Someone who didn't know shit from fat meat about microprocessors might call the x86 decoding stage on the P4 a "coprocessor" (what might they say about the Athlon, one wonders, rhetorically.) They might also believe that SPEC is just another measure of clock speed and integer performance. They probably believe that "RISC" means "fast", and that the fastest G4 is faster than the fastest P4. They probably don't know the difference between SP and DP floating point, of course, to say nothing of Altivec's fairly significant advantages over SSE2.
In other words, they'd be a typical Slashbot, who's attempt at iconoclasm means adherence to the Mac platform, rather than Linux. Which is basically idiocy, but then, it's certainly true that this is not a technical forum, and one can't expect any sort of reasonable technical discussion here.
The current generation of Powermacs very limited by their anemic memory subsystem. Rumor has it that Apple will be introducing "true" DDR G4 Towers at MWNY next week (as opposed to the Xscale).
That's what I'm doing. I have the money budgeted, but I can't really stomach spending it on a dual 1.0ghz box that's still stuck on PC-133 SDRAM.
And the rule of thumb when buying from Apple is get as little as possible from the Apple store and buy the rest of the components elsewhere. Nowadays, that mostly means memory (cheaper from Crucial) and disk (cheaper from anywhere.)
Oops. Of course clock speed is not a meaningful measure of performance across architectures. SPEC, the only really useful general purpose benchmark, is.
Altivec is a wondrous thing, but the majority of work that your computer is going to do is not vectorizable, or requires DP floating point, in which case, Altivec buys you nothing. And the pathetic memory subsystem on the Macintosh really hurts when you/do/ get to use Altivec, as it can't feed the G4 fast enough.
And trying to trot out the term "RISC" in this discussion is meaningless, and only highlights your ignorance (to say nothing of "386 co-processor".) Do yourself a favor and learn about CPU architectures before you post on the subject again, thanks.
Expect almost a linear scaling for FSAA. Note that at 4x, the GF4 would be pushing out around 40 fps. The 6k? About 13. At 8x? Let's be generous, and call it 8. Yes, the machines being tested are very different (a 1.3ghz Athlon vs. an 800mhz P3), but at those resolutions, you're very close to being 100% CPU bound.
I admire the meaningless iconoclasm that would lead one to tout an evolutionary dead-end like the 6000 as the be-all end-all of video cards, but in the future, you would be better served by appealing to the Voodoo's superior blast capacity, or the "warmth" of its image, rather than trying to make a technical argument without even the slimmests of legs to stand on.
Huh. That's the first bad word I've ever heard about Directron. I had to RMA two things with them, and a quick phone call settled things perfectly both times.
YMMV, of course, but I find them one of the very best.
Slashdot Mirror
I know the picture. But the Simpsons episode in question was pretty clearly a "Rashomon" parody.
'j
"Lola Rennt"? I think you misspelled "Rashomon".
'jfb
I've gotten three jobs, and innumerable interviews, from craigslist. Companies that not only know about it, but also post openings there, are companies that are likely ready to hire. Highly, highly recommended.
'jfb
'jfb
I had no idea, but it surely makes sense. His "technical" analyses always had the stink of the blue denim logo shirt and Gap khaki crowd to me.
'jfb
Every is a well-known crank, with a vanishingly small understanding of technology. While I'm all for pointless, iconoclastic stands on principle, MacKiDo is basically the rantings of a brainless loon.
'jfb
Tiny little pressure suits?
'j
Sendmail? Are you thinking of the same Sendmail that I am? The group of science rocketists who've given us *this*:
R$* $: $(dequote "" $&{client_addr} $)
R$-.$-.$-.$- $: $[ $4.$3.$2.$1.rbl.maps.vix.com $]
R$-.$-.$-.$-.rbl.maps.vix.com. $#error $@ 5.7.1 $: 550 no access from [$4.$3.$2.$1], see http://mail-abuse.org/rbl/
?
I can't think of a anyone, anywhere, less competent to write a complex, featureful, information interchange mechanism. They've had their chance, and that's as good as they could come up with? Egad!
And then, in a brilliantly ironic twist, I can't post those sendmail.cf snippets as text, because of Slashdot's indescribable "lameness" filter. It's the irresistible force versus immovable Taco, and we all end up losing. Figures.
'j
Love that ruby the diameter of a trash can lid. The genius of the 1st Ed.
'j
What's next? Pudge on John Stuart Mill's "On Liberty"? CmdrTaco dissecting "A History Of Sexuality"? The intersection of academe and Slashdot is too terrible to imagine ...
'jfb
The sperm whale is the largest toothed whale in the world, and likely the largest predator ever on earth. They are known to primarialy eat smaller squid (although a 13' squid is only smaller in comparison to Architheuthis), as well as fish, and the occasional marine mammal, like seals or porpoises.
'jfb
Trust CNN to get the basic science wrong. If a whale has a sucker scar, it's from a desparate squid trying to escape, not from a brush with death. Look at the sizes of the things: how would a 250kg squid handle a 60 ton whale?
Jeez.
'j
It's a quote from a Tom Lehrer song, actually. Maybe a bit oversubtle.
'jfb
We taught them a lesson, in 1918,
and they've hardly bothered us since then. - tl
I don't know why I bother.
SPEC is not a neutral performance benchmark. It is clearly biased as it does not measure floating point performance-- it essentially measures clock speed.
So a processor that's faster will do always better on SPEC? And it doesn't measure floating point performance? Pardon me, sparky? Maybe you'd better get on the horn with IBM:
1.1ghz POWER4 708/1075 (int/FP)
1.1ghz Pentium III 427/268 (int/FP)
So what's that second number mean? Frist Post? Fried plantains? How about FLOATING POINT, you dullard? Is SPEC biased because it can be gamed by clever compiler writers? Or is it biased because it returns a result that makes you feel bad? This is an important point to realize, if anyone is still reading this: there's BitGeek, claiming that SPEC is a clockrate limited measure of integer performance, and then there're the SPEC members (no-name hype-driven marketing companies like IBM, Sun, Intel, Digital^WCom^H^H^HHP, Apple, &c) who view it as a reasonable, if imperfect, tool for measuring general microprocessor performance. Who do you believe?
A processor with a simpler instruction set can execute more instructions in parallel during a clock cycle, dipshit.
What in God's name does this mean? It's totally nonsensical, a simplification so gross as to be completely meaningless. What about the number of execution units? Will a single execution unit R3000 (about the perfect RISC microprocessor) outperform a multiple execution unit Motorola 68060 (about the perfect CISC processor)? What about the pipeline depth? And, more trenchantly, so what? What microprocessor performance is, is instructions issued and retired in a given unit of time: ie, work done.
Do you even know what RISC is? what it means?
RISC is tangential to this "discussion" because the P4 is a RISC processor (with a small number of consistently sized instructions), with an x86 decoder stage early in the instruction pipeline. In other words, it sees the putative benefits of RISC (increased performance) whilst maintaining backwards compatibility. Certainly there might be performance advantages to pulling the decoding stage and issuing 'native' P4 instructions -- the image that you seem to have, that the decoder is a separate 386 processor shrunk down and magically inserted into the die is strange and hilarious -- but what's the point? Intel gets plenty of performance from their current design, and a full reengineering of the entire project would likely not show performance benefits worth the cost. Even if your people from Planet Tinfoil-Hat view the x86 world's convincing performance superiority to be as conspiratorial myth, Steve "Reality At Arms-Length" Jobs understands it: it's why you'll never see a dollar for dollar performance comparison at a MW "bakeoff", let alone a straight-up top of the line Photoshop extravaganza.
Someone who didn't know shit from fat meat about microprocessors might call the x86 decoding stage on the P4 a "coprocessor" (what might they say about the Athlon, one wonders, rhetorically.) They might also believe that SPEC is just another measure of clock speed and integer performance. They probably believe that "RISC" means "fast", and that the fastest G4 is faster than the fastest P4. They probably don't know the difference between SP and DP floating point, of course, to say nothing of Altivec's fairly significant advantages over SSE2.
In other words, they'd be a typical Slashbot, who's attempt at iconoclasm means adherence to the Mac platform, rather than Linux. Which is basically idiocy, but then, it's certainly true that this is not a technical forum, and one can't expect any sort of reasonable technical discussion here.
Best,
'jfb
http://www.fatalsteps.com/ac/tetris_japan_finals.m peg
'j
The current generation of Powermacs very limited by their anemic memory subsystem. Rumor has it that Apple will be introducing "true" DDR G4 Towers at MWNY next week (as opposed to the Xscale).
That's what I'm doing. I have the money budgeted, but I can't really stomach spending it on a dual 1.0ghz box that's still stuck on PC-133 SDRAM.
And the rule of thumb when buying from Apple is get as little as possible from the Apple store and buy the rest of the components elsewhere. Nowadays, that mostly means memory (cheaper from Crucial) and disk (cheaper from anywhere.)
'jfb
We'll let the numbers speak for themselves:
/do/ get to use Altivec, as it can't feed the G4 fast enough.
SPEC CPU 2000
P4 2.0 ghz : 745/743
G4 1.0ghz : 306/187 (!)
Oops. Of course clock speed is not a meaningful measure of performance across architectures. SPEC, the only really useful general purpose benchmark, is.
Altivec is a wondrous thing, but the majority of work that your computer is going to do is not vectorizable, or requires DP floating point, in which case, Altivec buys you nothing. And the pathetic memory subsystem on the Macintosh really hurts when you
And trying to trot out the term "RISC" in this discussion is meaningless, and only highlights your ignorance (to say nothing of "386 co-processor".) Do yourself a favor and learn about CPU architectures before you post on the subject again, thanks.
'jfb
Links:
http://www.heise.de/ct/english/02/05/182/ (G4 SPEC, unofficial)
http://www.spec.org (P4 SPEC)
Some numbers for you.
5 .html
5 83&p =9
Q3A, 1600x1200x32 bit (no FSAA)
GF4 Ti4600 : 160.6 fps
V5 6000 : 58.7 fps
Expect almost a linear scaling for FSAA. Note that at 4x, the GF4 would be pushing out around 40 fps. The 6k? About 13. At 8x? Let's be generous, and call it 8. Yes, the machines being tested are very different (a 1.3ghz Athlon vs. an 800mhz P3), but at those resolutions, you're very close to being 100% CPU bound.
I admire the meaningless iconoclasm that would lead one to tout an evolutionary dead-end like the 6000 as the be-all end-all of video cards, but in the future, you would be better served by appealing to the Voodoo's superior blast capacity, or the "warmth" of its image, rather than trying to make a technical argument without even the slimmests of legs to stand on.
Best,
'jfb
Links:
V5 6k benchmarking: http://www.voodooextreme.com/hw/previews/v5_6000/
GF4 numbers: All over, but I used these:
http://www.anandtech.com/showdoc.html?i=1
Huh. That's the first bad word I've ever heard about Directron. I had to RMA two things with them, and a quick phone call settled things perfectly both times.
YMMV, of course, but I find them one of the very best.
'jfb
Also excellent are
http://www.mwave.com
... and, for more esoteric case-related parts,
http://www.directron.com
There's no point in dealing with the hassles that other online retailers will subject you to.
'jfb
... or high quality cans. I endorse the find products from Grado Labs. Mmmmm.
'jfb
Yikes. Either way, we're fucked.
'jfb
I'd be more worried about my trousers catching on fire than the titanium melting ...
'jfb
Huge aquatic beast? Never before encountered? Stirring at a time when nuclear trouble is in the air?
Duh. It's clearly Godzilla.
'j