Exactly, so Apple's model is to make money form IPOD sales, not from distribution of songs.... thats just an additional service to push the sales of IPODs... I was just responsding to the parent, and a few others who were speculating whether there is money in distribution or not, and the answer is that no, unless you can use distribution to push a hardware device, or maybe a DRM technology(something like what sony is now trying to with with Sony Connect) there is not a lot of money in distribution
Well, it has always been and will always be players... I chatted with the MTV CEO about why they don't enter the music distribution business, considering that they already have the relationships in place. And then he brought me down to earth. How much do you think does apple earn from selling one song? About 20-30 cents. The rest goes to music companies and artists. So how many songs do you have to sell in order to make 50 million dollars, an amount which is fairly trivial actually for these companies. The answer is around 250 million... so there is no way that a company can make a lot of money here.. in fact apple pissed off a lot of people by carging 99 cents per song, because they set the industry standard, which no one can deviate from now.
Marketcap is only inflated or deflated during certain periods (bubbles). Under normal circumstances, it is an extremely good indication of how much the company is worth and what are its future growth prospects. Agreed that marketcap can be inflated, but not for a long time (basic finance, can't elaborate much here). And contrary to popular belief, Mutual funds don't have that big an impact on marketcap. The point is that mutual funds buy if the company is attractive, not the other way around. And the factors that you mentioned (revenue, assets, profit, etc), all go in the calculation of marketcap. So for all practical purposes, market is always right.
LMAO... and when was that last time? Twenty years ago? According to Indian labour laws, if you are working in the private sector, there is no such thing as permanent job. So wake up and smell the coffee.
Yeah, it has a static overhead per task. As the number of tasks increase, teh overhead will increase as well. It no doubt will be able to maintain a lower latency but if teh CPU load is 100%, then every task switch becomes expensive, so I am not sure about the throughput part in such a case.
That was kinda interesting though it doesn't really explain how they achieve the O(1)ness, and for example how did kernel 2.4 fare in comparison.
Anyway, an O(1) scheduler is nice and all under low to moderate loads, but at high loads, when a lot of threads are running, the O(1) behaviour should enusre that the tasks are switching extremely fast, and as every context switch has some related overhead in terms of saving the context etc., that means that performance of the O(1) schedular should be worse than that of 2.4(where tasks are not switching as fast) in theory atleast. So is this really true?
Hmm, I agree that it is a bit hard to edit the files by hand, but Mandrake comes with extensive online documentation and if you had looked, you would have found how to change the settings manually.
As far as applying the patch to the kernel, well mandrake adds some features like supermount to the kernel they ship which sometimes means that you can't apply the patches but, one mandrake does give you linus kernel in their urpmi repository and secondly, it is equally easy to download a stock kernel and then apply the patch. So I don't think that is a good criteria to judge a kernel. The fact of the matter is, Mandrake philosophy has always been to make it easy to use and so they add features like supermount. But that doesn't take away the power from you to replace the kernel alltogether. For example, I personally run 2.6 test 11, but then I don't use supermount. How does that make a distro bad is beyond me.
RPM hell? When was the last time you used mandrake? You do know that you have to configure urpmi so that it can download latest software from mirrors, which by the way can easily be found on this site: Easy Urpmi
Subscription service is a value added service for club members. There you can get the latest test software and then club members test them before they are released to rest of the public. So get a clue before starting to flame.
Even the US army officials conceded that he wasn't controlling anything. In fact there were hardly any people protecting him. The people running the resistance are different. Saddam wasn't controlling them.
What it does though is that probably it will remove the fear from people that what will happen to them if he comes back to power, but frankly, that was only a remote possibilty to say the least.
Anyway, I am happy that a rutheless dictator like him as been captured. But I am unhappy that this will make Dubya win the next election. I mean why only Saddam? What about North Korea and Mayanmaar, which are also dictatorships? North Korea declares that they have WMDs and Mayanmaar is involved big time in drug trafficking, but China won't let US come so close to it. So why these double standards? I mean happy as I may be at Saddam's capture, the point still remains that the basis of the war was still wrong.
Well, no it is not noticeably faster than 2.4.22-mm2 (multimedia patch), so in case you are scared to move on to a test kernel, you can try the mm patches for 2.4 (though they themselves are test patches). Theese include Ingo's O1 scheduler which is the biggest improvement in 2.6. But hey, for the sake of the community, why not move on to 2.6. I personally run both on my laptop and both seem to work fine. I had some problems with ACPI on 2.4 but 2.6 fixes that and also hotplug really works in 2.6.
Lol, well mostly they are stupid:-)
But many times we develop some inhouse hardware, especially in the area of RF chips. The hardware could be really stupid but the point is that at the time it is developed, the competitors haven't thought about it as yet. And RF chips are relatively easy to reverse engineer (hardware design vise) once you know the instruction set. That is one of the biggest reasons of this secrecy. Mostly you can get away with getting the patents for such a thing but sometimes it is not possible to do so, especially in Europe, and some other times, it is possible to circumvent the patent once you know how the device exactly works (We do this all the time).
As far as having a bigger market, honestly speaking hackers are a minority and ppl who blindly flash their devices are a majority. One such example is making the DVD drives region free. The amount which companies spend in supporting such guys is not very small. I mean just imagine that you update the firware and you screw the device. IN such a case, most ppl come to the company hoping that no one will notice what they tried to do. And frankly verifying every such claim costs money. So these are the two main reasons.
I know that such projects are quite good if you really want to understand hardware and maybe write drivers for such devices to work on you favorite Operating systems, but I am an embedded developer myself and he point is that in case you need to do something like this, as in you can't make it work with you r computer without reverse engineering the device, such projects make a lot of sense. Otherwise it is fairly simple for a device manufacturer to simply burn the JTAG pins. Only thing they need to do is to create a in program bootloader which can proigram the flash either over tftp or over serial interface and you can easily protect such things with public and private keying mechanism, which is much harder to crack. The reason companies don't do that is because JTAG is more flexible and the fastest way to program a device but even my company is contemplating burning the JTAG chips once the device is programmed, so I guess other companies would be thinking in the same direction as well.
IIRC, Mandrakesoft did make a deal with VmWare to provide a trial version of the software to club members. What I am not sure though is f it is included in the distro. I would rather doubt that. Maybe Wine is included?
Precisely. Atleast you run a dell. I run a no name laptop, made by a small Taiwanese company. Redhat and Suse have trouble with configuring the graphics card, even though its a Radeon derivate but with Mandrake, everything runs smooth as butter:-)
Not to mention urpmi rocks, and supermount is really cool too. And to top it all, there is the community. I am a club memeber and no other company lets you have so much say in the final distro as mandrake does. It almost functions like a democracy. You vote for RPMs, you zourself package the RPMs, you vote for features which go in the release and you test the RPMs.
So overall, Mandrake is atleast my top choice.
This is a bit strange but club memebers are normally sent a mail informing about important things but somehow they were haven't been informed about the torrents so far. Also the club website doesn't say anything about torrent links (atleast not teh memebers site). Which is probably why the download speed is still very low on bittorrent.
And I was surprised at the voice quality which I got with a simple dialup connection. It wasn't better than phone as skype claims but it sure was good enough. I could hear everything clearly and the lag was negligible. The only problem was that the connection kept gettting dropped, no idea why, but I still managed to get about 15 minutes call between each drop so it wasn't that bad. So yes this technology does have a future.
As far as inability to snoop, I don't think it is relevant. I mean it can only matter if the company is based in US. It is almost like saying that ban freenet in US because we can't see whats going on there. So I don't really see a problem with that.
It reminds of of the article a few days back that Indian government had asked its ISPs to block access to a particular groups's yahoo group. As ISPs couldn't implement that, they blocked all the yahoo groups. Anyway, the question is that are such type of actions completely unjustified? I mean I know they are against freedom of expression, but what are the alternatives for a government when faced with such a situation. Take the example of SCO. Even though they emanate a lot of FUD, many people still believe that they are right (What was that group which bought SCOX shares?) The point is most people don't research the claims made by others. They believe what is aligned with their existing belief system and reject the rest (the reason for all the religious wars I think). So now faced with such a situation, should the govermnent, with the knowledge that this may lead to some other people joining the movement without knowing all the facts about it, be justified in imposing some kind of restrictions? I mean I know it will make many people extremely mad, and maybe make them join the ranks of such an organisation, even though they might not have under normal circumstances, but is this number is smaller than the total number of people who would have been affected otheriwse, is the government justified?
As the earth is still coming out of its last ice age, we shouldn't be too concerned about global warming.
Actually, according to theory, we should have already come out of the last ice age. Ice ages occur every 20,000 years and the last one occured more than 10,000 years ago. In fact this is one of the biggest arguments climate scientists give as proof of global warming. According to them, we should have reached the peak of the temperature curve and now should be on the downward slope, but temperature keeps rising each year, which means that something is changing the natural balance.
The result of this growing disparity between the haves and have nots. I mean everyone acknowledges that brain drain happens because the conditions in some other country are much better than conditions in one's home country, which used to be the case in India up until 90s, but now I think the process has slowed.
I know that there are a lot of slashdotters who oppose Indians taking their jobs, but the point is that this is the only area where Indians were able to compete with US, in the face of such a huge disparity. Did you know that US pays a 3 Billion dollars subsidy to its cotton farmers every year. And do you know the number of cotton farmers in US? 25000. Which means a subsidy of 120,000 USD per farmer per year, enough to hire two software engineers. These farmers then compete with farmers of countries like India in the international market whose per capita income is 500 USD per year . That is the irony of the situation that these poaching practices killed almost all the industries of the developing countries, and now the only capital they are left with is their people. (India used to be the biggest producer of cotton once upon a time btw). So now we are seeing them fighting back with the only resource they have. How come slashdotters can make societies to ban H1Bs but can't make societies to ask their sentors to cut down the subsidies being given to already rich farmers and maybe invest this money to make education cheaper or start some other development activity? That is the tragedy of US, that every economist says these policies are bad, every senator knows that as well, but majority of the people are not aware because it doesn't affect them directly. All I am saying is don't fight what you see in front. Spare some thought for the causes behind the problem as well.
First, there are people from US working in India already in many companies and it is very much possible to do that. But the company should be willing to sponsor your visa. This chap just decided that no one is sponsoring just because someone told him so. What more can I say. Another thing which I don't understand is that when the whole US manufacturing industry was moved to China, no one said anything, so why now this hue and cry about India? The principals are the same. If you can get something chepaer somewhere else, you will get it from there. Its India today, will be somewhere else tomorrow.
It is not unusual for Indian companies not to advertise on their websites. Bulk of thier recruitment is done from college campuses where they hire students fresh out of college. For example, a company like Infosys will approximately hire around 2000 people per year from campuses.
I suppose google can afford to offer such serivices for free. Just look at google groups. But I won't be very surprised to see context specific ads on the blogs as well. The strategy google is following is targeted advertising. So if some blogger writes about say IBM Vs SCO, you can expect to see an ad of some Linux solution on top of that blog (Or worse, an MS ad saying you won't have any IP problems with MS). I think its a good idea because like search engine, you know who your target customer is for blogs. So there is indeed a business plan behind this.
Slashdot Mirror
Exactly, so Apple's model is to make money form IPOD sales, not from distribution of songs.... thats just an additional service to push the sales of IPODs... I was just responsding to the parent, and a few others who were speculating whether there is money in distribution or not, and the answer is that no, unless you can use distribution to push a hardware device, or maybe a DRM technology(something like what sony is now trying to with with Sony Connect) there is not a lot of money in distribution
Well, it has always been and will always be players... I chatted with the MTV CEO about why they don't enter the music distribution business, considering that they already have the relationships in place. And then he brought me down to earth. How much do you think does apple earn from selling one song? About 20-30 cents. The rest goes to music companies and artists. So how many songs do you have to sell in order to make 50 million dollars, an amount which is fairly trivial actually for these companies. The answer is around 250 million... so there is no way that a company can make a lot of money here.. in fact apple pissed off a lot of people by carging 99 cents per song, because they set the industry standard, which no one can deviate from now.
Marketcap is only inflated or deflated during certain periods (bubbles). Under normal circumstances, it is an extremely good indication of how much the company is worth and what are its future growth prospects. Agreed that marketcap can be inflated, but not for a long time (basic finance, can't elaborate much here). And contrary to popular belief, Mutual funds don't have that big an impact on marketcap. The point is that mutual funds buy if the company is attractive, not the other way around. And the factors that you mentioned (revenue, assets, profit, etc), all go in the calculation of marketcap. So for all practical purposes, market is always right.
LMAO... and when was that last time? Twenty years ago? According to Indian labour laws, if you are working in the private sector, there is no such thing as permanent job. So wake up and smell the coffee.
Yeah, it has a static overhead per task. As the number of tasks increase, teh overhead will increase as well. It no doubt will be able to maintain a lower latency but if teh CPU load is 100%, then every task switch becomes expensive, so I am not sure about the throughput part in such a case.
That was kinda interesting though it doesn't really explain how they achieve the O(1)ness, and for example how did kernel 2.4 fare in comparison. Anyway, an O(1) scheduler is nice and all under low to moderate loads, but at high loads, when a lot of threads are running, the O(1) behaviour should enusre that the tasks are switching extremely fast, and as every context switch has some related overhead in terms of saving the context etc., that means that performance of the O(1) schedular should be worse than that of 2.4(where tasks are not switching as fast) in theory atleast. So is this really true?
Hmm, I agree that it is a bit hard to edit the files by hand, but Mandrake comes with extensive online documentation and if you had looked, you would have found how to change the settings manually.
As far as applying the patch to the kernel, well mandrake adds some features like supermount to the kernel they ship which sometimes means that you can't apply the patches but, one mandrake does give you linus kernel in their urpmi repository and secondly, it is equally easy to download a stock kernel and then apply the patch. So I don't think that is a good criteria to judge a kernel. The fact of the matter is, Mandrake philosophy has always been to make it easy to use and so they add features like supermount. But that doesn't take away the power from you to replace the kernel alltogether. For example, I personally run 2.6 test 11, but then I don't use supermount. How does that make a distro bad is beyond me.
RPM hell? When was the last time you used mandrake? You do know that you have to configure urpmi so that it can download latest software from mirrors, which by the way can easily be found on this site:
Easy Urpmi
Subscription service is a value added service for club members. There you can get the latest test software and then club members test them before they are released to rest of the public. So get a clue before starting to flame.
Even the US army officials conceded that he wasn't controlling anything. In fact there were hardly any people protecting him. The people running the resistance are different. Saddam wasn't controlling them. What it does though is that probably it will remove the fear from people that what will happen to them if he comes back to power, but frankly, that was only a remote possibilty to say the least. Anyway, I am happy that a rutheless dictator like him as been captured. But I am unhappy that this will make Dubya win the next election. I mean why only Saddam? What about North Korea and Mayanmaar, which are also dictatorships? North Korea declares that they have WMDs and Mayanmaar is involved big time in drug trafficking, but China won't let US come so close to it. So why these double standards? I mean happy as I may be at Saddam's capture, the point still remains that the basis of the war was still wrong.
Well, no it is not noticeably faster than 2.4.22-mm2 (multimedia patch), so in case you are scared to move on to a test kernel, you can try the mm patches for 2.4 (though they themselves are test patches). Theese include Ingo's O1 scheduler which is the biggest improvement in 2.6. But hey, for the sake of the community, why not move on to 2.6. I personally run both on my laptop and both seem to work fine. I had some problems with ACPI on 2.4 but 2.6 fixes that and also hotplug really works in 2.6.
This seems to work fine for me.
Lol, well mostly they are stupid :-)
But many times we develop some inhouse hardware, especially in the area of RF chips. The hardware could be really stupid but the point is that at the time it is developed, the competitors haven't thought about it as yet. And RF chips are relatively easy to reverse engineer (hardware design vise) once you know the instruction set. That is one of the biggest reasons of this secrecy. Mostly you can get away with getting the patents for such a thing but sometimes it is not possible to do so, especially in Europe, and some other times, it is possible to circumvent the patent once you know how the device exactly works (We do this all the time).
As far as having a bigger market, honestly speaking hackers are a minority and ppl who blindly flash their devices are a majority. One such example is making the DVD drives region free. The amount which companies spend in supporting such guys is not very small. I mean just imagine that you update the firware and you screw the device. IN such a case, most ppl come to the company hoping that no one will notice what they tried to do. And frankly verifying every such claim costs money. So these are the two main reasons.
I know that such projects are quite good if you really want to understand hardware and maybe write drivers for such devices to work on you favorite Operating systems, but I am an embedded developer myself and he point is that in case you need to do something like this, as in you can't make it work with you r computer without reverse engineering the device, such projects make a lot of sense. Otherwise it is fairly simple for a device manufacturer to simply burn the JTAG pins. Only thing they need to do is to create a in program bootloader which can proigram the flash either over tftp or over serial interface and you can easily protect such things with public and private keying mechanism, which is much harder to crack. The reason companies don't do that is because JTAG is more flexible and the fastest way to program a device but even my company is contemplating burning the JTAG chips once the device is programmed, so I guess other companies would be thinking in the same direction as well.
Try openoffice 1.1. Your slides work with that.
I am already using 2.4.22.mm2 with Ingo's O(1) scheduler and mm patches from Andrew Morton. So what other benifits can I expect by upgrading to 2.6?
IIRC, Mandrakesoft did make a deal with VmWare to provide a trial version of the software to club members. What I am not sure though is f it is included in the distro. I would rather doubt that. Maybe Wine is included?
Precisely. Atleast you run a dell. I run a no name laptop, made by a small Taiwanese company. Redhat and Suse have trouble with configuring the graphics card, even though its a Radeon derivate but with Mandrake, everything runs smooth as butter :-)
Not to mention urpmi rocks, and supermount is really cool too. And to top it all, there is the community. I am a club memeber and no other company lets you have so much say in the final distro as mandrake does. It almost functions like a democracy. You vote for RPMs, you zourself package the RPMs, you vote for features which go in the release and you test the RPMs.
So overall, Mandrake is atleast my top choice.
This is a bit strange but club memebers are normally sent a mail informing about important things but somehow they were haven't been informed about the torrents so far. Also the club website doesn't say anything about torrent links (atleast not teh memebers site). Which is probably why the download speed is still very low on bittorrent.
And I was surprised at the voice quality which I got with a simple dialup connection. It wasn't better than phone as skype claims but it sure was good enough. I could hear everything clearly and the lag was negligible. The only problem was that the connection kept gettting dropped, no idea why, but I still managed to get about 15 minutes call between each drop so it wasn't that bad. So yes this technology does have a future. As far as inability to snoop, I don't think it is relevant. I mean it can only matter if the company is based in US. It is almost like saying that ban freenet in US because we can't see whats going on there. So I don't really see a problem with that.
It reminds of of the article a few days back that Indian government had asked its ISPs to block access to a particular groups's yahoo group. As ISPs couldn't implement that, they blocked all the yahoo groups.
Anyway, the question is that are such type of actions completely unjustified? I mean I know they are against freedom of expression, but what are the alternatives for a government when faced with such a situation. Take the example of SCO. Even though they emanate a lot of FUD, many people still believe that they are right (What was that group which bought SCOX shares?) The point is most people don't research the claims made by others. They believe what is aligned with their existing belief system and reject the rest (the reason for all the religious wars I think). So now faced with such a situation, should the govermnent, with the knowledge that this may lead to some other people joining the movement without knowing all the facts about it, be justified in imposing some kind of restrictions? I mean I know it will make many people extremely mad, and maybe make them join the ranks of such an organisation, even though they might not have under normal circumstances, but is this number is smaller than the total number of people who would have been affected otheriwse, is the government justified?
As the earth is still coming out of its last ice age, we shouldn't be too concerned about global warming.
Actually, according to theory, we should have already come out of the last ice age. Ice ages occur every 20,000 years and the last one occured more than 10,000 years ago. In fact this is one of the biggest arguments climate scientists give as proof of global warming. According to them, we should have reached the peak of the temperature curve and now should be on the downward slope, but temperature keeps rising each year, which means that something is changing the natural balance.
The result of this growing disparity between the haves and have nots. I mean everyone acknowledges that brain drain happens because the conditions in some other country are much better than conditions in one's home country, which used to be the case in India up until 90s, but now I think the process has slowed. I know that there are a lot of slashdotters who oppose Indians taking their jobs, but the point is that this is the only area where Indians were able to compete with US, in the face of such a huge disparity. Did you know that US pays a 3 Billion dollars subsidy to its cotton farmers every year. And do you know the number of cotton farmers in US? 25000. Which means a subsidy of 120,000 USD per farmer per year, enough to hire two software engineers. These farmers then compete with farmers of countries like India in the international market whose per capita income is 500 USD per year . That is the irony of the situation that these poaching practices killed almost all the industries of the developing countries, and now the only capital they are left with is their people. (India used to be the biggest producer of cotton once upon a time btw). So now we are seeing them fighting back with the only resource they have. How come slashdotters can make societies to ban H1Bs but can't make societies to ask their sentors to cut down the subsidies being given to already rich farmers and maybe invest this money to make education cheaper or start some other development activity? That is the tragedy of US, that every economist says these policies are bad, every senator knows that as well, but majority of the people are not aware because it doesn't affect them directly. All I am saying is don't fight what you see in front. Spare some thought for the causes behind the problem as well.
First, there are people from US working in India already in many companies and it is very much possible to do that. But the company should be willing to sponsor your visa. This chap just decided that no one is sponsoring just because someone told him so. What more can I say. Another thing which I don't understand is that when the whole US manufacturing industry was moved to China, no one said anything, so why now this hue and cry about India? The principals are the same. If you can get something chepaer somewhere else, you will get it from there. Its India today, will be somewhere else tomorrow.
It is not unusual for Indian companies not to advertise on their websites. Bulk of thier recruitment is done from college campuses where they hire students fresh out of college. For example, a company like Infosys will approximately hire around 2000 people per year from campuses.
I suppose google can afford to offer such serivices for free. Just look at google groups. But I won't be very surprised to see context specific ads on the blogs as well. The strategy google is following is targeted advertising. So if some blogger writes about say IBM Vs SCO, you can expect to see an ad of some Linux solution on top of that blog (Or worse, an MS ad saying you won't have any IP problems with MS). I think its a good idea because like search engine, you know who your target customer is for blogs. So there is indeed a business plan behind this.