Meanwhile, IBM continues to locate all of its sites exactly as far away from public transportation as possible. The Toronto site, 10 years ago, went so far as to move *further* away from the downtown area.
here's what my representative had to say about her vote in favor of the bill:
Thank you for contacting me to express your opposition to H.R. 6304, the Foreign Intelligence Surveillance Act (FISA) Amendments Act of 2008. I appreciate the benefit of your views, and I welcome the opportunity to respond.
In March 2008 the House of Representatives passed legislation to modernize FISA. This well-crafted proposal would have outlawed warrantless surveillance and did not provide immunity for telecommunications companies that participated in the Administration's illegal Terrorist Surveillance Program. I strongly supported this legislation.
The Senate ignored the well-crafted House bill and instead passed its own ill-conceived legislation that would have granted warrantless spying, allowed the Administration to conduct surveillance activities outside the purview of the FISA Court, and provided blanket immunity for telecommunications companies. I strongly opposed the Senate bill.
The compromise legislation the House considered, H.R. 6304, is far closer to the House legislation and a significant improvement over current law. It would require judicial approval before surveillance could begin; prohibit the Administration from conducting surveillance outside the purview of the FISA Court; prevent reverse-targeting, a tactic used to spy on Americans without a warrant; include unprecedented reporting and oversight requirements; and update FISA to incorporate modern communications technology.
For those reasons I supported H.R. 6304. It does not provide immunity for federal officials who administered the Terrorist Surveillance program, but unfortunately it does include provisions that will likely lead to immunity for telecommunications companies. I do not support immunity. However, the bill would be such an improvement over current law, I determined it would be better to support this bill than to maintain the status quo and allow for warrantless spying on Americans.
Throughout the FISA debate, my goal has been to give law enforcement and the intelligence community the tools it requires to combat the very real threat of international terrorism while preserving our civil liberties. These are not mutually exclusive goals, and I believe that H.R. 6304 accomplishes both. While I oppose the inclusion of immunity provisions, it prevents warrantless spying on Americans by requiring individual warrants based on probable cause, protects American citizens' rights while abroad, and places the entire surveillance program under judicial review. The compromise, while far from perfect, is a vast improvement over current law.
[...] just as web page formats are pretty much irrelevant to current web users.
this is because the protocol (HTTP), which is very relevant to current web users, is open and standard. the file format (HTML) isn't relevant to current web users, because it is hidden behind a wire protocol. whenever you turn a remote data source to a local one, however, the file format is exposed. and it then becomes relevant.
are you suggesting that in the future, we won't have any local repositories? or that the local repositories will be accessed via wire protocols? at some point, the distinction between a protocol and a file format seems unimportant.
i've always wondered why people say "six and one half dozen of the other". Shouldn't it be "six of one, half dozen of the other" -- with the comma being the equals sign, you thus state a tautology to reinforce that you "could care less" about the difference?
My dad phrased it the former way my whole childhood. It was one of those linguistic ambiguities, you know the ones that haunt you, hehe -- am I just mishearing that, or is there something strange about it?
as far as I can tell, the article only gives one quantification of the scale these folks are dealing with: on the order of tens of thousands for that one case. is this what is considered large? the point of visualizations is to show patterns of nodes (and patterns of paths) in graphy structures. at some point, one runs into one or the other of various limits:
pixels on the display: 2 million or so.
insufficiency of the clustering algorithms: showing one pixel per node and random placement, or placement by DFS traversal? for trees, or for graphs where classification is the primary concern, then tree-map or "Csoft" views scale relatively well in this regard, but what about for more general problems?
implementations (or algorithms) that don't scale: e.g. graphviz uses n^2 (n=#nodes) space for its graph layout!
one must always think about the summarization criteria: what aren't you going to show? how will you indicate that summarization has occured? how do you denote drill-down capability? what will the form of drill-down be? what heuristics should you use to selectively deaggregate, in order to highlight potentially interesting subgraphs? for large-scale info, this is as important as what you will be showing, and how it will be shown!
for our stuff, we have graphs with tens of millions of nodes.
it reminds me that one of the neat things about the movie Aliens was how the hero didn't overcome by some inherent, over even trained, super skills. in fact, it is the inherently skilled (the android bishop) or trained ones who all die. it's so rare to have superheros these days that overcome through perseverance and wisdom and intuition (not knowing fancy fighting skills like they show in the Serenity trailer, or how to build things or hotwire --- just knowing what makes sense to do at any one time)
i also liked the whole inverted parenting themes (super corporations and military as caretakers; the aliens breeding by growing inside of foreign hosts, ripley being a good mother to someone she doesn't even know)
you didn't seem to compile the c++ code with optimizations. i suggest compiling with -O2.
for the kind of code you chose, you used a bad provider for java. i suggest the IBM 1.4.2, which is flat out the fastest x86 VM on the kind of code you provided.
you didn't use the VM version for the java you chose. if you are going to run a sun-based VM, then, for the kind of number-crunching code you provided, you should use the "-server" option. the hotspot-based JIT is optimized for a different kind of code (UI, interactive)
on a 1.6GHz pentium-m running linux, i get:
c++: 2 seconds
IBM 1.4.2 java: 2.4 seconds
sun 1.5.0 -server java: 4.8 seconds
sun 1.5.0 -client java: 7.7 seconds
(times are best of 4 consecutive runs using, as you did the "time" command)
nick
i'm 6'5" and had a honda insight for a year (before i sold it to go carless); it has more legroom and headroom than a honda accord. this is because the seat is lower, and because it is a two seater.
next excuse?
you should be able to do better on the freeway, too: since you only really need torque while accelerating, and since the electric motor provides this, you can get away with a smaller gasoline engine.
are you sure about that? i'd believe it if you replaced i486 with i586. the pentium was in-order issue, and had "typed" dual pipelines, so that, while it could execute two instructions in parallel, only if the compiler perfectly scheduled things to use the dual issue slots.
the 686 was out-of-order, and had 5 (6?) pipelines, and more than one of each type.
therefore, codegen for 586 wasn't particularly useful for 686. codegen for 486 should work fine on 686?
hehe, i can't tell if you're jokey or not. the grandparent talks about misuse of begging the question, and you zing him for begging the question? bravo, sir!
does the NYT have a logic to their use of dots in abbreviations?
e.g. they say "A.M.D."
they also say "U.S.B."
but they say "DVD", and i've even seen a mix of "DVD" and "U.S.B." in the same article. Is there some underlying logic?
I can get from my house in white plains, new york to my brother's in wilmington, delaware by trains as fast as driving. And I mean it when I say trains (plus walking!): white plains to NYC, then subway from grand central to penn station, then amtrak. Costs $90 round trip. If my brother were willing to pick me up in trenton, new jersey, then it'd only cost $30.
Even if it took 30-60 minutes longer, it's so much less stressful than driving on the jersey turnpike! And the speed is not just cause of traffic on the 'pike, it's cause of restroom pit stops, and cause the trains go fast, zoomy!
I grew up on the west coast, and took caltrain for four months. Yes, it sucks (though the limited stop version is much better). I took BART for four years from san francisco to berkeley; it's pretty good. Metro-North here in new york is a better commuter railway than both, though (i think).
I suspect that, if this dude has a high-end built-in, then the dude has the microwave running on its own circuit. wouldn't this avoid the problem you mention, anon?
hey, I just read that, while popcorn has a high glycemic index (72), it has a low glycemic load (8). Meaning that it has a high impact on blood sugar, but is pretty low in carbs -- so you'd have to eat a hecka big load of popcorn to impact your blood sugar levels appreciably.
In comparison, a baked potato has a GI or 85 and a GL of 26. Carrots are 49 and 2, lentils are 30 and 5, brown rice is 64 and 23. Raw broccoli is basically zero on both figures.
so I say go for the popcorn, dude! I'm not sure whether the GL figures account for the high fiber content of popcorn (compared to a potato). If it doesn't, then all the better!
I am an abashed subscribed to the print version of Stereophile. This whole subthread is a little silly, given that one of this magazine's strengths lies in how it combines the qualitative with the quantitative. The head editor, John Atkinson, runs a battery of measurements against many of the components that they review. In particular, he ran these tests against the iPod. It came out measuring pretty well!
I don't know whether the online version includes these measurements. In general, the online version has the measurement pages at the end of the click-throughs. I can't check it out, since stereophile seems to have gotten spanked:)
hi! ok, I'm a delany #1 fan, but dhalgren is a truly awesome reading experience. it's a good bit different than nova and triton (which I think are very different, between themselves!) but that's just me.
PBD has the potential to make computers easier to use by allowing any user to automate repetitive tasks simply by demonstrating the task on concrete examples.
The 390 machines use this not for performance, but actually for security. Each transaction uses a fresh JVM heap, so there is no possibility that errant code might mistakenly (who, me? hehe) read something is shouldn't, like the previous transaction's cleartext user info.
nick
(i believe this post is past the SAT, the Slashdot Attention Threshold, hehe, lesse if anyone reads it)
Hi, if you have them handy, I'd totally appreciate if you can share any performance numbers. I'm not being facetious at all, I am the anti-slashdot, and am truly interested! I'm doing research on memory leak debugging, so data is the drug I desire.:D
thanks!
nick
Slashdot Mirror
Meanwhile, IBM continues to locate all of its sites exactly as far away from public transportation as possible. The Toronto site, 10 years ago, went so far as to move *further* away from the downtown area.
Maybe the editor posted before previewing?
this is because the protocol (HTTP), which is very relevant to current web users, is open and standard. the file format (HTML) isn't relevant to current web users, because it is hidden behind a wire protocol. whenever you turn a remote data source to a local one, however, the file format is exposed. and it then becomes relevant.
are you suggesting that in the future, we won't have any local repositories? or that the local repositories will be accessed via wire protocols? at some point, the distinction between a protocol and a file format seems unimportant.
My dad phrased it the former way my whole childhood. It was one of those linguistic ambiguities, you know the ones that haunt you, hehe -- am I just mishearing that, or is there something strange about it?
-
pixels on the display: 2 million or so.
- insufficiency of the clustering algorithms: showing one pixel per node and random placement, or placement by DFS traversal? for trees, or for graphs where classification is the primary concern, then tree-map or "Csoft" views scale relatively well in this regard, but what about for more general problems?
- implementations (or algorithms) that don't scale: e.g. graphviz uses n^2 (n=#nodes) space for its graph layout!
one must always think about the summarization criteria: what aren't you going to show? how will you indicate that summarization has occured? how do you denote drill-down capability? what will the form of drill-down be? what heuristics should you use to selectively deaggregate, in order to highlight potentially interesting subgraphs? for large-scale info, this is as important as what you will be showing, and how it will be shown! for our stuff, we have graphs with tens of millions of nodes.i guess it looks kinda fun
it reminds me that one of the neat things about the movie Aliens was
how the hero didn't overcome by some inherent, over even trained,
super skills. in fact, it is the inherently skilled (the android
bishop) or trained ones who all die. it's so rare to have superheros
these days that overcome through perseverance and wisdom and intuition
(not knowing fancy fighting skills like they show in the Serenity
trailer, or how to build things or hotwire --- just knowing what makes
sense to do at any one time)
i also liked the whole inverted parenting themes (super corporations
and military as caretakers; the aliens breeding by growing inside of
foreign hosts, ripley being a good mother to someone she doesn't even
know)
but serenity looks like it might be fun
earthquake!!
why are disks that store email so much more expensive than disks that store anything else?
- you didn't seem to compile the c++ code with optimizations. i suggest compiling with -O2.
- for the kind of code you chose, you used a bad provider for java. i suggest the IBM 1.4.2, which is flat out the fastest x86 VM on the kind of code you provided.
- you didn't use the VM version for the java you chose. if you are going to run a sun-based VM, then, for the kind of number-crunching code you provided, you should use the "-server" option. the hotspot-based JIT is optimized for a different kind of code (UI, interactive)
on a 1.6GHz pentium-m running linux, i get:- c++: 2 seconds
- IBM 1.4.2 java: 2.4 seconds
- sun 1.5.0 -server java: 4.8 seconds
- sun 1.5.0 -client java: 7.7 seconds
(times are best of 4 consecutive runs using, as you did the "time" command) nicki'm 6'5" and had a honda insight for a year (before i sold it to go carless); it has more legroom and headroom than a honda accord. this is because the seat is lower, and because it is a two seater. next excuse?
you should be able to do better on the freeway, too: since you only really need torque while accelerating, and since the electric motor provides this, you can get away with a smaller gasoline engine.
are you sure about that? i'd believe it if you replaced i486 with i586. the pentium was in-order issue, and had "typed" dual pipelines, so that, while it could execute two instructions in parallel, only if the compiler perfectly scheduled things to use the dual issue slots. the 686 was out-of-order, and had 5 (6?) pipelines, and more than one of each type. therefore, codegen for 586 wasn't particularly useful for 686. codegen for 486 should work fine on 686?
hehe, i can't tell if you're jokey or not. the grandparent talks about misuse of begging the question, and you zing him for begging the question? bravo, sir!
does the NYT have a logic to their use of dots in abbreviations? e.g. they say "A.M.D." they also say "U.S.B." but they say "DVD", and i've even seen a mix of "DVD" and "U.S.B." in the same article. Is there some underlying logic?
things done by you or on your behalf may not be as you expected, or intended.
Even if it took 30-60 minutes longer, it's so much less stressful than driving on the jersey turnpike! And the speed is not just cause of traffic on the 'pike, it's cause of restroom pit stops, and cause the trains go fast, zoomy!
I grew up on the west coast, and took caltrain for four months. Yes, it sucks (though the limited stop version is much better). I took BART for four years from san francisco to berkeley; it's pretty good. Metro-North here in new york is a better commuter railway than both, though (i think).
I suspect that, if this dude has a high-end built-in, then the dude has the microwave running on its own circuit. wouldn't this avoid the problem you mention, anon?
In comparison, a baked potato has a GI or 85 and a GL of 26. Carrots are 49 and 2, lentils are 30 and 5, brown rice is 64 and 23. Raw broccoli is basically zero on both figures.
so I say go for the popcorn, dude! I'm not sure whether the GL figures account for the high fiber content of popcorn (compared to a potato). If it doesn't, then all the better!
nick
I don't know whether the online version includes these measurements. In general, the online version has the measurement pages at the end of the click-throughs. I can't check it out, since stereophile seems to have gotten spanked :)
hate jar jar
hi! ok, I'm a delany #1 fan, but dhalgren is a truly awesome reading experience. it's a good bit different than nova and triton (which I think are very different, between themselves!) but that's just me.
(hi tessa!) nick
The 390 machines use this not for performance, but actually for security. Each transaction uses a fresh JVM heap, so there is no possibility that errant code might mistakenly (who, me? hehe) read something is shouldn't, like the previous transaction's cleartext user info.
nick (i believe this post is past the SAT, the Slashdot Attention Threshold, hehe, lesse if anyone reads it)Hi, if you have them handy, I'd totally appreciate if you can share any performance numbers. I'm not being facetious at all, I am the anti-slashdot, and am truly interested! I'm doing research on memory leak debugging, so data is the drug I desire. :D
thanks!
nick