I have a 28/1.8 Sigma for my 20D, and I have to say that for a long time it was my favourite lens.
Since then I've gotten a 10-22mm which is just *insane*. It's not as fast though, so indoors you tend to need a flash. And it's by no stretch a cheap lens either.
But in many ways it's exemplifies why you should go with the dSLR route. Begin with a cheap(-ish) 350D/D50 and kit lens. After a while trade up and get some better Sigma/Tamron lenses (they are quite good compared to the first party ones). In my experience the 3rd party lenses are quite comparable wrt optical quality, but the autofocus tend to lag and be a bit noisy.
If you find Laszlo interesting I recommend you listen to an interview with one creators. That talk was the first time I heard about the system and it seems really interesting. Particularly how they decided to make it a "proper" compiler so you can stick on different backends depending on what you want to target. Ie you could theoretically stick on a full SVG backend if you want to target Firefox 1.5 and later (as it has full SVG support).
I posted a link above about this. Apparently the experiment has been reproduced, on site, using several hundred people and polished shields. This was done on the BBC show "What the ancients did for us".
In previous Slashdot articles there has been a lot of comments about weather/temperature differences. Typically it goes something like first someone points out that SF weather may not be a good test. After this, typically a person from SF or at least California state that it representative since it's approx on the same parallell. Finally every now and then you get someone who has actually been to both SF and Greece who say that the weather isn't really similar and that the test isn't valid. (For the record I've never been to neither SF or Greece, though I've been in southern Europe a lot during the summers.)
However, I think that what I find that they fail at is to even recognise these errors. (Sure they have "rematches" but they seem to fail to adress the errors there too.)
In this case the myth was that Arcamedies had an army of soliders with highly polished shields. On command the soliders directed the sunlight at the advancing ships. And burned to dust an entire fleet. No not one boat, but an entire advancing fleet.
Well, turning one ship on fire at a time would work as well.
I think if you were sitting in a ship about to go to war and all of a sudden there's a flash of light from the city walls you are about to invade, and the ship next to you is set ablase. I don't know about you but that would freak me out.
In fact I can't even remember that even they claimed that the entire fleet was turned to dust in a single second. I think the original myth is more along the lines of "roman fleet being scared away" or "invasion stopped" or something along those lines. (Or it may be even more accurte to say that the myth is that Arcimedes created a "weapon" that could focus the sun and turn ships on fire from a distance.)
What sucks the most about this (IMHO) is that the myth has been confirmed, in Greece, with polished shields and about 500 people. This was demonstrated in the BBC program What the ancients did for us. (If anyone can find that program I'd like to see it though. It seems hard to find references to the experiment online.)
It was a really really big anti armour sniper rifle. I imagine it would do squat against a tank.
FYI a real anti tank RPG works by having a core of eg copper. At impact this core is heated and forced into the point of impact by the cone shaped mantle. The result is a "spear" of molten copper which can penetrate the tanks armour and enter the tank. (The actual entry hole is quite small though.) After it enters the tank it will scatter and quite probably significantly burn personell at the same time as it creates a significant over preassure inside the tank. This preassure causes the atmosphere inside the tank to explode out creating a under preassure which afterwards leads to implosion as preassures stabilise.
Basically if you're in a tank that's hit you're screwed.
The frontend needs to be powerful if you want to decode HDTV. 1080p requires a really good CPU/GPU combo to playback without framedrops.
The reason you would want a backend with CPU power instead of hardware encoders is because software encoders are typically higher quality. It's hard to do multi-pass encoding and similar if you are recording real-time.
Naturally these don't apply to everyone, but there are reasons for getting a high powered HTPC.
When you sharpen an image in Photoshop (or Gimp) you are only increasing the contrast. While this has the effect of making the image "pop out" more which looks sharp it can't extract more detailed information from the image.
I have RGB inputted from my Xbox to my TV. I also have a GameCube hooked up with a different Scart using composite connections. Point is, the Scart doesn't really specify the signal any more than saying it's "RCA" (those are the connectors you usually get on composite/component wires).
Now I bet that you can get a description which just say "scart" and then it's probably because it's by marketing.;-)
How about if you try to look into facts before shooting your mouth off?
While Java and real-time may seem like hard to integrate it is possible and it has been done (successfully). Eg I've seen demonstrations of RT systems runnning on industrial robots which balanced inverted pendulums without a problem.
Now naturally timing may differ between architectures, that's to be expected. What's important is that if you have a specific architecture you need to know, exactly, what delays occur and that they have upper bounds. The same would be true with a Asm or C RT system.
The main reason you'd want to run Java RT is because first off, it's faster to write correctly. It also makes it possible to have eg dynamic reprogramming on a factory floor. (To the point where new control data is updated while the welder robot switches between weld spots.)
Achieving Zen may not be for everybody, but an irrational possessiveness, it seems, is the source of much unhappiness in this world.
I agree fully.
That is why I support developers who aid others in their quest for achieving Zen by releasing the code as GPL and thus ensuring that more people will stop hoarding knowledge in their irrational possessiveness. Hopefully this will bring about a better world filled with cuddly teddybears for everyone.
I agree with you that it would be nice with games that are online capable across architectures. However...
Multi-console games will always suffer. The reason being that it takes time for the dev team to make it work across consoles. And then the design team has to alter the game so that it can be played even on the slowest ones. In the end you end up with a game that plays about equally well (or poorly) on every console.
Personally I prefer it when the dev team spends their time optimising the heck out of the game. Preferrably they should also make good use of any specific features of a certain platform (eg Live on XBox). This does pose a bigger risk for the title though as the potential market is smaller.
But I'd rather get a top of the line "XBox only" (Or GC only, or PS2 only) game than a mediocre "on all platforms" game. And rather than a top of the line "on all platforms" game I'd like to have a "best game of console" game.
Of course, I (now) have all current gen consoles.
Re:Since the submitter didn't bother to explain...
on
IBM Releases Cell SDK
·
· Score: 1
I don't really think that idea would work. It could work for a super scalar processor like the Itanium where they do this (but on the same die). But in that case you already have the data loaded locally and it's a lot easier to roll back the changes when one branch is rejected. I'm sure that just rejecting and rolling back changes with multiple SPEs would kill any benefit from running multiple branches in different SPEs.
They just hope that the compiler will fix as much as possible for them. And other than that they hope that other problems will be fixed by brute force.
It's a really interesting design though, it will be interesting to see what will be done with it.
Listen to this speech Ontology is Overrated by Clay Shirky (from ITConversations creat site!).
Then if you look at the 200 Religion category in the Dewey decimal system you will most likely see a tad bit of unbalancedness. Out of 100 classification slots there are more that are unused than that are allocated to non-Christian religions. A quick overview suggests that a majority of them actually has "Christianity" explicitely in the name or implicitely by being about eg Jesus. Now this may work well in the US, but it sure doesn't work very well in the rest of the world.
Furthermore what Shirky comments on in the speech above is that the entire idea of dividing something into smaller topics is really only needed when you have a physical limitation. (Ie you can't put a physical book in two different bookshelves.) But for all intents and purposes all ontology systems fail in more or less spectacular ways. Furthermore they are always made by someone with (typically) only a vague idea of the topics being ordered.
He also talks about how free form tagging systems like on Flickr and Delicious allow for a more true categorisation to form freely. Similarly to how comments can spread information just the study of how people group items together can give us a lot of knowledge about the items. (This could give the recommendations on Amazon a huge boost in relevance.)
I agree with you that it's quite easy to monitor data, even on the scale we see here. And let's be clear, there is no need to actually monitor the data. You only need to monitor the amount of data in order to find these darknets. Furthermore, even if you sometimes fail to achieve 100% of data logging that doesn't matter since you'll randomly drop packets from your data loggers and thus the darknets will still shine like beacons of bandwidth. One trick is to look for bandwidth during the off-hours. Typically traffic is low during the night, but people that download a lot typically maintain this during all hours. So during the day there might well be way too much information to parse, but during the nights only the "usual suspects" are working.
However, as I was previously involved with such logging (as an admin for a small student network) I toyed with the idea of making a more advanced darknet. Usually logging and tracking are based on the assumption that the darknets are operating on usual IP adresses, naturally this isn't necessary. Eg you could let multiple computers on one subnet create a new virtual host together. You would then load balance the darknet over all of these hosts. And this loadbalancing could be made using non-standard IP packets. The idea is that if you were to look at any specific stream of traffic it wouldn't make sense. Only when you correctly put them together are they correct.
I mentioned previously that you only log the actual data lengths. Theoretically you can make communication channels using elaborate port knocking which would circumvent this. Eg you could use port knocking to transmitt data, in a similar manner as morse code. So you are not really sending any data in packets, you are just "knocking on the ports" of the other computer in a manner which signifies a message. Inefficient as hell, that's for sure, but hard to detect unless you know what to look for.
The extreme version of this would be to hook up one computer to a central switch on a logging port and hide it. The idea is that this computer would be able to intercept all network transmission on a network and furthermore to allow it to send data on all these ports. That would allow for a pretty extreme variant of the above "multiple computers on a subnet" as suddenly you really have one computer that is hooked in on the entire subnet. This allows it to loadbalance over all of the subnet essentially making it "invicible" to most basic data logging analysis.
I'm sure people who actually spend a lot of time analysing IP data can think of even more subtle and hard to find ways of creating hidden communication channels.
If you don't want to learn the language of a country then don't move there! Hello, McFly!
The discussion is regarding people who flee their home and go to a different country out of necessity. In some cases it is to flee powerty (as in Mexico) but mostly it is people that flee because they risk being imprisoned or killed in their native country.
It is not so much what they are running towards as what they are running from.
Here in Canada there are actually many inexpensive (and free) classes on learning Engligh (and French I believe).
Of course you can learn Swedish for free as an immigrant here too. However since many immigrants end up in the same geographical place they don't "need" to learn Swedish as they can get around just fine talking in their native language. The downside is of course that this makes it harder for them to communicate outside their own community which further closes them off.
Personally I think that where the Swedish integration politics really fail is ensuring that people are integrated into the country. Not only wrt learning the language but also that many immigrants are not given the jobs they are trained for, which further alienates them and ultimately wastes their talents.
Slashdot Mirror
I have a 28/1.8 Sigma for my 20D, and I have to say that for a long time it was my favourite lens.
Since then I've gotten a 10-22mm which is just *insane*. It's not as fast though, so indoors you tend to need a flash. And it's by no stretch a cheap lens either.
But in many ways it's exemplifies why you should go with the dSLR route. Begin with a cheap(-ish) 350D/D50 and kit lens. After a while trade up and get some better Sigma/Tamron lenses (they are quite good compared to the first party ones). In my experience the 3rd party lenses are quite comparable wrt optical quality, but the autofocus tend to lag and be a bit noisy.
If you find Laszlo interesting I recommend you listen to an interview with one creators. That talk was the first time I heard about the system and it seems really interesting. Particularly how they decided to make it a "proper" compiler so you can stick on different backends depending on what you want to target. Ie you could theoretically stick on a full SVG backend if you want to target Firefox 1.5 and later (as it has full SVG support).
What can I say, I'm practicing my double talk.
Woooooosh.
Considering you are also currently posting on Slashdot using HTML lists no less, I don't think you're doing a very good job at pleasing her.
Unless she is very nerdy I doubt she'll consider this neither foreplay nor pillow talk.
I posted a link above about this. Apparently the experiment has been reproduced, on site, using several hundred people and polished shields. This was done on the BBC show "What the ancients did for us".
In previous Slashdot articles there has been a lot of comments about weather/temperature differences. Typically it goes something like first someone points out that SF weather may not be a good test. After this, typically a person from SF or at least California state that it representative since it's approx on the same parallell. Finally every now and then you get someone who has actually been to both SF and Greece who say that the weather isn't really similar and that the test isn't valid. (For the record I've never been to neither SF or Greece, though I've been in southern Europe a lot during the summers.)
However, I think that what I find that they fail at is to even recognise these errors. (Sure they have "rematches" but they seem to fail to adress the errors there too.)
Well, turning one ship on fire at a time would work as well.
I think if you were sitting in a ship about to go to war and all of a sudden there's a flash of light from the city walls you are about to invade, and the ship next to you is set ablase. I don't know about you but that would freak me out.
In fact I can't even remember that even they claimed that the entire fleet was turned to dust in a single second. I think the original myth is more along the lines of "roman fleet being scared away" or "invasion stopped" or something along those lines. (Or it may be even more accurte to say that the myth is that Arcimedes created a "weapon" that could focus the sun and turn ships on fire from a distance.)
What sucks the most about this (IMHO) is that the myth has been confirmed, in Greece, with polished shields and about 500 people. This was demonstrated in the BBC program What the ancients did for us. (If anyone can find that program I'd like to see it though. It seems hard to find references to the experiment online.)
I think that's more a topic for Penn&Teller's Bullshit. And they've already done it.
They were not impressed.
It was a really really big anti armour sniper rifle. I imagine it would do squat against a tank.
FYI a real anti tank RPG works by having a core of eg copper. At impact this core is heated and forced into the point of impact by the cone shaped mantle. The result is a "spear" of molten copper which can penetrate the tanks armour and enter the tank. (The actual entry hole is quite small though.) After it enters the tank it will scatter and quite probably significantly burn personell at the same time as it creates a significant over preassure inside the tank. This preassure causes the atmosphere inside the tank to explode out creating a under preassure which afterwards leads to implosion as preassures stabilise.
Basically if you're in a tank that's hit you're screwed.
The frontend needs to be powerful if you want to decode HDTV. 1080p requires a really good CPU/GPU combo to playback without framedrops.
The reason you would want a backend with CPU power instead of hardware encoders is because software encoders are typically higher quality. It's hard to do multi-pass encoding and similar if you are recording real-time.
Naturally these don't apply to everyone, but there are reasons for getting a high powered HTPC.
When you sharpen an image in Photoshop (or Gimp) you are only increasing the contrast. While this has the effect of making the image "pop out" more which looks sharp it can't extract more detailed information from the image.
Whichever you connect to it.
;-)
I have RGB inputted from my Xbox to my TV. I also have a GameCube hooked up with a different Scart using composite connections. Point is, the Scart doesn't really specify the signal any more than saying it's "RCA" (those are the connectors you usually get on composite/component wires).
Now I bet that you can get a description which just say "scart" and then it's probably because it's by marketing.
How about if you try to look into facts before shooting your mouth off?
While Java and real-time may seem like hard to integrate it is possible and it has been done (successfully). Eg I've seen demonstrations of RT systems runnning on industrial robots which balanced inverted pendulums without a problem.
Now naturally timing may differ between architectures, that's to be expected. What's important is that if you have a specific architecture you need to know, exactly, what delays occur and that they have upper bounds. The same would be true with a Asm or C RT system.
The main reason you'd want to run Java RT is because first off, it's faster to write correctly. It also makes it possible to have eg dynamic reprogramming on a factory floor. (To the point where new control data is updated while the welder robot switches between weld spots.)
I agree fully.
That is why I support developers who aid others in their quest for achieving Zen by releasing the code as GPL and thus ensuring that more people will stop hoarding knowledge in their irrational possessiveness. Hopefully this will bring about a better world filled with cuddly teddybears for everyone.
SCART is just a type of connector like RCA. You can have a SCART which carries a component/composite and/or S-video as well as stereo audio.
I agree with you that it would be nice with games that are online capable across architectures. However...
Multi-console games will always suffer. The reason being that it takes time for the dev team to make it work across consoles. And then the design team has to alter the game so that it can be played even on the slowest ones. In the end you end up with a game that plays about equally well (or poorly) on every console.
Personally I prefer it when the dev team spends their time optimising the heck out of the game. Preferrably they should also make good use of any specific features of a certain platform (eg Live on XBox). This does pose a bigger risk for the title though as the potential market is smaller.
But I'd rather get a top of the line "XBox only" (Or GC only, or PS2 only) game than a mediocre "on all platforms" game. And rather than a top of the line "on all platforms" game I'd like to have a "best game of console" game.
Of course, I (now) have all current gen consoles.
I don't really think that idea would work. It could work for a super scalar processor like the Itanium where they do this (but on the same die). But in that case you already have the data loaded locally and it's a lot easier to roll back the changes when one branch is rejected. I'm sure that just rejecting and rolling back changes with multiple SPEs would kill any benefit from running multiple branches in different SPEs.
They just hope that the compiler will fix as much as possible for them. And other than that they hope that other problems will be fixed by brute force.
It's a really interesting design though, it will be interesting to see what will be done with it.
Best AD evar as well!
Double page filled with red and two windshield wipers, and the slogan "If you've got the RAM, we've got the pedestrians."
But otherwise I think Carmageddon was better.
From what I've seen this DOES NOT work with video though. It will only stream video from a MCE computer.
It does however work with music and photos.
I didn't preview and then Slashcode seems to have swallowed my reply.
m l.
Anyways the link to the speech is http://www.itconversations.com/shows/detail470.ht
Listen to this speech
Ontology is Overrated by Clay Shirky (from ITConversations creat site!).
Then if you look at the 200 Religion category in the Dewey decimal system you will most likely see a tad bit of unbalancedness. Out of 100 classification slots there are more that are unused than that are allocated to non-Christian religions. A quick overview suggests that a majority of them actually has "Christianity" explicitely in the name or implicitely by being about eg Jesus. Now this may work well in the US, but it sure doesn't work very well in the rest of the world.
Furthermore what Shirky comments on in the speech above is that the entire idea of dividing something into smaller topics is really only needed when you have a physical limitation. (Ie you can't put a physical book in two different bookshelves.) But for all intents and purposes all ontology systems fail in more or less spectacular ways. Furthermore they are always made by someone with (typically) only a vague idea of the topics being ordered.
He also talks about how free form tagging systems like on Flickr and Delicious allow for a more true categorisation to form freely. Similarly to how comments can spread information just the study of how people group items together can give us a lot of knowledge about the items. (This could give the recommendations on Amazon a huge boost in relevance.)
I agree with you that it's quite easy to monitor data, even on the scale we see here. And let's be clear, there is no need to actually monitor the data. You only need to monitor the amount of data in order to find these darknets. Furthermore, even if you sometimes fail to achieve 100% of data logging that doesn't matter since you'll randomly drop packets from your data loggers and thus the darknets will still shine like beacons of bandwidth. One trick is to look for bandwidth during the off-hours. Typically traffic is low during the night, but people that download a lot typically maintain this during all hours. So during the day there might well be way too much information to parse, but during the nights only the "usual suspects" are working.
However, as I was previously involved with such logging (as an admin for a small student network) I toyed with the idea of making a more advanced darknet. Usually logging and tracking are based on the assumption that the darknets are operating on usual IP adresses, naturally this isn't necessary. Eg you could let multiple computers on one subnet create a new virtual host together. You would then load balance the darknet over all of these hosts. And this loadbalancing could be made using non-standard IP packets. The idea is that if you were to look at any specific stream of traffic it wouldn't make sense. Only when you correctly put them together are they correct.
I mentioned previously that you only log the actual data lengths. Theoretically you can make communication channels using elaborate port knocking which would circumvent this. Eg you could use port knocking to transmitt data, in a similar manner as morse code. So you are not really sending any data in packets, you are just "knocking on the ports" of the other computer in a manner which signifies a message. Inefficient as hell, that's for sure, but hard to detect unless you know what to look for.
The extreme version of this would be to hook up one computer to a central switch on a logging port and hide it. The idea is that this computer would be able to intercept all network transmission on a network and furthermore to allow it to send data on all these ports. That would allow for a pretty extreme variant of the above "multiple computers on a subnet" as suddenly you really have one computer that is hooked in on the entire subnet. This allows it to loadbalance over all of the subnet essentially making it "invicible" to most basic data logging analysis.
I'm sure people who actually spend a lot of time analysing IP data can think of even more subtle and hard to find ways of creating hidden communication channels.
That's becuase they also have seminars on how to work with British people. Particularly those in the toilet industry.
If you don't want to learn the language of a country then don't move there! Hello, McFly!
The discussion is regarding people who flee their home and go to a different country out of necessity. In some cases it is to flee powerty (as in Mexico) but mostly it is people that flee because they risk being imprisoned or killed in their native country.
It is not so much what they are running towards as what they are running from.
Here in Canada there are actually many inexpensive (and free) classes on learning Engligh (and French I believe).
Of course you can learn Swedish for free as an immigrant here too. However since many immigrants end up in the same geographical place they don't "need" to learn Swedish as they can get around just fine talking in their native language. The downside is of course that this makes it harder for them to communicate outside their own community which further closes them off.
Personally I think that where the Swedish integration politics really fail is ensuring that people are integrated into the country. Not only wrt learning the language but also that many immigrants are not given the jobs they are trained for, which further alienates them and ultimately wastes their talents.
And since Diplomacy relies so heavily on discussions you could make it a plug-in for GoogleTalk or GMail.