Like I said... you don't decaffeinate the beer, you decaffeinate whatever it is you're adulterating^Wbrewing the beer with. Either you add caffeine to beer or you don't. It's not a naturally occuring ingredient in beer.
And yes, being a USian, I haven't hit too many beer festivals in the UK. I have tried well over 200 different kinds of beer at the Flying Saucer, though.
I've heard of caffeinated beer, but removing the caffeine from beer is kinda like removing the lead from gasoline—you just don't put it in to begin with. That's why they call it unleaded, not deleaded.
Ok, so then in those fake elections, don't send your vote rigging crew in, and it looks like they're fine. Then, in November, rig it. How have you solved anything?
Voting machines should have a paper trail. I personally would do this in triplicate: A receipt handed to the voter, a log in the machine, and then a combined log of all the machines at a polling place. In the event of a recount, you could then go back to the paper logs. Without paper logs, you can only ask the machine to give you the number it already gave you--not very useful.
Perhaps. My main point, though I could've done better articulating it, is that the only "news" here is that "Some web search results are dangerous; don't expect (Google|MSN|Ask|whoever) to be your nanny." The fact that most of the search engines had differences that were near the noise margin means there's no real strong conclusions you can draw.
MSN's number is an obvious anomaly that could be explained multiple ways: Microsoft heavily vets its advertisers; advertisers don't think MSN's worth advertising on; and Microsoft doesn't seek advertisers as aggressively because it doesn't need them. All three are plausible, and perhaps the truth is a mixture of the three.
Those are the most annoying ads: Meta ads that claim to have "everything you're looking for" and end up just being some crappy sort of meta page that has nothing immediately useful.
Ebay's pretty obvious when it comes up. There are others that are more annoying.
And, the results aren't too surprising. MSN came in lowest, but at the same time, it's probably the least dependent on generating ad revenue. Google's the biggest, but has a repuation to maintain, so it probably does at least some filtering of advertisers. Ask keeps tring to reinvent itself back to relevance...
Or it could just be a baffled reporter writing what they thought they understood, or Boyd dumbed it down too much for the reporter, or both. Notice it wasn't a direct quote?
Electromigration can cause wires to "pull apart." This happens when the switching current generates too strong of an electromagnetic field, and it literally pulls the wire along.
On reading the WikiPedia page more closely, I see where we're talking past each other. You're referring to a mandatory precedent, and I'm thinking of the broader concept. An initial decision of a lower court can be construed as a persuasive or advisory precedent.
At any rate, to have this case kicked out of the appellate court would establish a negative mandatory precedent. With a vacated conviction, even the persuasive precedent doesn't stand. The prosecution can wait for a stronger case to come by to establish the desired mandatory precedent.
If someone convicted of murder appeals, and before the appeal is heard the prosecuter becomes aware of new evidence that exonerates the defendant and moves to vacate the conviction, do you think that all future murders in the relevant jurisdiction should suddenly become legal?
Oh, certainly not! But what happens to any precedent established by the initial conviction? Does that evaporate, or does it remain based on the merits of the case and available evidence at the time it was decided?
The reason I raise the question is that one possible motivation for vacating a conviction is that it doesn't set the precedent you want, or that it establishes the precendent weakly and you'd prefer a stronger case to underpin it. I suspect the latter was the motivation here.
What effect does vacating a conviction like this have on precedent? That is, if the appeal proceded and the original conviction was overturned, the precedent would clearly side with McDanel, under some legal theory to be articulated in the judgment handed down by the appellate court. But, given that the conviction was vacated, does that mean the case sets no precedent whatsoever? How does this work?
That's what I was thinking. Also, notice a lot of bots and viruses and such come from Eastern Europe, especially Russia.... What was that saying about idle hands again?
Basically, what you're saying is that Parnas' calculus was only strong enough to determine if program X correctly produces the output expected by algorithm Y, without respect to the actual complexity. Is that correct? I'm curious: Does it prove that it actually implements algorithm Y, or just the same outputs as algorithm Y? For example, bubble sort and stabilized quicksort both produce the same sorted lists, but with very different different time complexity. OTOH, two different quicksort implementations might vary in execution cost due to quality of implementation, even if they go through all the same intermediate states.
If it proves program X implements algorithm Y (and not just a result-quivalent fascimile), then you can make broad statements about run time, but to only slightly better granularity than O() notation. It doesn't sound strong enough to predict anything about the actual execution time of the program.
Now, if you can analyze the algorithms involved in terms of "# of instances of each of these primitives" (where the primitives might be things like "read sensor", "compare string", "dereference structure pointer") such that you could write complexity equations for each algorithm that were more detailed than O(), you could then limit much of your analysis to measuring the costs of the primitives and then plugging-and-chugging on the complexity equations. Still not much fun.
But, then, who said real-time programming when millions of lives are on the line was meant to be fun?
This hypothetical machine that has a non-MS Word word processor installed... Does it have all the screen-reader and other accessibility software installed?
As I understood it, the Amiga 500 could gen-lock, and was used for countless video-titling tasks. Not the full up Video Toaster, but sufficient for basic NLE and titling with an external VCR controller.
I did go looking for links but it's been hard going. One reference at Linux Weekly News that Linus wanted the 2.3.x dev cycle to be "half the length or less" of the 2.1.x dev cycle. (In reality, 2.3 took 18 months vs. 2.1's 27 months--shorter, but only by a third.)
No, the 2.6.x.y are patch releases of 2.6.x. The development releases are 2.6.x-preY. The release candidates are 2.6.x-rcY.
Makes sense to me at least.
--Joe
Re:History should be written by those who remember
on
Gadgets, Then & Now
·
· Score: 1
I'm just saying that the most popular media wasn't necessarily 3.5" media at that point. From 89-90, there is a strong argument that 3.5" was the up and comer, but not dominant yet amongst the bulk of users.
Let's extend your car analogy for a moment. Miles per gallon is easy. But if you just gave your mom a description of the car in terms of # of cylinders, total displacement, final drive ratio and drag coefficient, don't you think it'd be a tad overwhelming as compared to "20 MPG city, 30 MPG highway?"
Slashdot Mirror
Like I said... you don't decaffeinate the beer, you decaffeinate whatever it is you're adulterating^Wbrewing the beer with. Either you add caffeine to beer or you don't. It's not a naturally occuring ingredient in beer.
And yes, being a USian, I haven't hit too many beer festivals in the UK. I have tried well over 200 different kinds of beer at the Flying Saucer, though.
--JoeI've heard of caffeinated beer, but removing the caffeine from beer is kinda like removing the lead from gasoline—you just don't put it in to begin with. That's why they call it unleaded, not deleaded.
--JoeDecaf beer?
Ok, so then in those fake elections, don't send your vote rigging crew in, and it looks like they're fine. Then, in November, rig it. How have you solved anything?
Voting machines should have a paper trail. I personally would do this in triplicate: A receipt handed to the voter, a log in the machine, and then a combined log of all the machines at a polling place. In the event of a recount, you could then go back to the paper logs. Without paper logs, you can only ask the machine to give you the number it already gave you--not very useful.
--JoePerhaps. My main point, though I could've done better articulating it, is that the only "news" here is that "Some web search results are dangerous; don't expect (Google|MSN|Ask|whoever) to be your nanny." The fact that most of the search engines had differences that were near the noise margin means there's no real strong conclusions you can draw.
MSN's number is an obvious anomaly that could be explained multiple ways: Microsoft heavily vets its advertisers; advertisers don't think MSN's worth advertising on; and Microsoft doesn't seek advertisers as aggressively because it doesn't need them. All three are plausible, and perhaps the truth is a mixture of the three.
Those are the most annoying ads: Meta ads that claim to have "everything you're looking for" and end up just being some crappy sort of meta page that has nothing immediately useful.
Ebay's pretty obvious when it comes up. There are others that are more annoying.
--JoeAnd, the results aren't too surprising. MSN came in lowest, but at the same time, it's probably the least dependent on generating ad revenue. Google's the biggest, but has a repuation to maintain, so it probably does at least some filtering of advertisers. Ask keeps tring to reinvent itself back to relevance...
--JoeOr it could just be a baffled reporter writing what they thought they understood, or Boyd dumbed it down too much for the reporter, or both. Notice it wasn't a direct quote?
Electromigration can cause wires to "pull apart." This happens when the switching current generates too strong of an electromagnetic field, and it literally pulls the wire along.
Also, you can punch through the gate oxide and dielectric layers, either causing a short or destroying the transistors.
--JoeOn reading the WikiPedia page more closely, I see where we're talking past each other. You're referring to a mandatory precedent, and I'm thinking of the broader concept. An initial decision of a lower court can be construed as a persuasive or advisory precedent.
At any rate, to have this case kicked out of the appellate court would establish a negative mandatory precedent. With a vacated conviction, even the persuasive precedent doesn't stand. The prosecution can wait for a stronger case to come by to establish the desired mandatory precedent.
Learn something new every day. :-)
--JoeConvictions don't set precedents? I thought that was the basis of common law. Are you saying precedents only apply in civil procedings?
--JoeOh, certainly not! But what happens to any precedent established by the initial conviction? Does that evaporate, or does it remain based on the merits of the case and available evidence at the time it was decided?
The reason I raise the question is that one possible motivation for vacating a conviction is that it doesn't set the precedent you want, or that it establishes the precendent weakly and you'd prefer a stronger case to underpin it. I suspect the latter was the motivation here.
--JoeWhat effect does vacating a conviction like this have on precedent? That is, if the appeal proceded and the original conviction was overturned, the precedent would clearly side with McDanel, under some legal theory to be articulated in the judgment handed down by the appellate court. But, given that the conviction was vacated, does that mean the case sets no precedent whatsoever? How does this work?
(It should be clear IANAL.)
--JoeThat's what I was thinking. Also, notice a lot of bots and viruses and such come from Eastern Europe, especially Russia.... What was that saying about idle hands again?
--Joe
Basically, what you're saying is that Parnas' calculus was only strong enough to determine if program X correctly produces the output expected by algorithm Y, without respect to the actual complexity. Is that correct? I'm curious: Does it prove that it actually implements algorithm Y, or just the same outputs as algorithm Y? For example, bubble sort and stabilized quicksort both produce the same sorted lists, but with very different different time complexity. OTOH, two different quicksort implementations might vary in execution cost due to quality of implementation, even if they go through all the same intermediate states.
If it proves program X implements algorithm Y (and not just a result-quivalent fascimile), then you can make broad statements about run time, but to only slightly better granularity than O() notation. It doesn't sound strong enough to predict anything about the actual execution time of the program.
Now, if you can analyze the algorithms involved in terms of "# of instances of each of these primitives" (where the primitives might be things like "read sensor", "compare string", "dereference structure pointer") such that you could write complexity equations for each algorithm that were more detailed than O(), you could then limit much of your analysis to measuring the costs of the primitives and then plugging-and-chugging on the complexity equations. Still not much fun.
But, then, who said real-time programming when millions of lives are on the line was meant to be fun?
--JoeThey gave up the standing seating because this approach looked more promising. The Beastie Boys have seen the future of air travel.
--JoeThis hypothetical machine that has a non-MS Word word processor installed... Does it have all the screen-reader and other accessibility software installed?
As I understood it, the Amiga 500 could gen-lock, and was used for countless video-titling tasks. Not the full up Video Toaster, but sufficient for basic NLE and titling with an external VCR controller.
--JoeDon't forget the intermediate upgrade to 'your hands.'
Oops, botched the LWN link.
I did go looking for links but it's been hard going. One reference at Linux Weekly News that Linus wanted the 2.3.x dev cycle to be "half the length or less" of the 2.1.x dev cycle. (In reality, 2.3 took 18 months vs. 2.1's 27 months--shorter, but only by a third.)
Then, when 2.5 kicked off, they said they wanted to shorten 2.5.x over 2.3.x, and, well, 2.5.x took 25 months.
BTW, you can look at kernel.org if you don't beleive me on the timeline.
--JoeI think that's been tried, or at the very least discussed and ruled out.
--JoeNo, the 2.6.x.y are patch releases of 2.6.x. The development releases are 2.6.x-preY. The release candidates are 2.6.x-rcY.
Makes sense to me at least.
--JoeI'm just saying that the most popular media wasn't necessarily 3.5" media at that point. From 89-90, there is a strong argument that 3.5" was the up and comer, but not dominant yet amongst the bulk of users.
Let's extend your car analogy for a moment. Miles per gallon is easy. But if you just gave your mom a description of the car in terms of # of cylinders, total displacement, final drive ratio and drag coefficient, don't you think it'd be a tad overwhelming as compared to "20 MPG city, 30 MPG highway?"
--Joe