Cloud computing is the client-server model that everyone is used to... but where the "server" is distributed.
There are significant advantages over a more traditional client-server model, even if the "server" is a cluster. Because the cloud is distributed geographically; 1. infrastructure outages are far less damaging to the application, 2. entire data centers can be taken off line and added at will, 3. power and cooling advantages can be used to keep costs lower, 4. Bandwidth utilization are distributed to multiple centers, which may help prevent saturation of regional carrier networks, 5. data may be read from the closest data center reducing bandwidth utilization vs a single data center on the other side of the globe.
Many things don't make sense to have on the cloud... but some things are particularly well suited for it. Email, I believe, is the ultimate cloud use... it needs high availability and fault tolerance. It needs huge capacities. And finally to keep storage utilization to a minimum, you will ideally have a single mail store, otherwise a single email sent to users on several mail stores will result in several copies of the same data. This kills Exchange in a lot of large organizations, a VP sends a large attachment to all of the regional head, now that same attachment exists on every exchange server in the organization, gets backed up with every server backup, and possibly never gets deleted since it came from the boss. A single mail store means a single copy (redundancy excluded, that exists with many mail stores too) of that attachemnt.
Actually... gas taxes are not a percentage of the sale price; instead they are $0.xx/gallon. Unfortunately that means that as gas prices go up, the government actually loses money because people buy less gas. Gas stations are the same way, many of them nearly went broke when prices soared to $3+ just because the credit card companies take their cut as a percentage while the gas station takes a fixed cut per gallon, with high priced gas, the credit card cut was often higher than the stations' cut.
There have also been a number of drugs under development, which were abandoned because they only benefited a small percentage of patients (say, for example, 30% of a hypothetical new drug), while carrying significant side effects. But just as those 10% of cancer patients got no benefit from tamoxifen because of a genetic variation, it might be the case that the 70% who didn't benefit from our hypothetical drug did so because they didn't have the correct genome. If we could identify the 30% that it worked for, what was once viewed as a failed drug could instead be a miracle drug, albeit for a subset of the entire patient population.
Imagine the research opportunities here. The R&D was done, the drug worked wonders in a small subset of patients and failed catastrophically in others... lets find out WHY. This idea could employ tens of thousands of researchers for several years just pouring over old trial data, running new trials, and linking genetic research to drug research.
I know if I were a pharma company I would be ramping up a similar effort. A good example is Elan's Tysabri, which was hailed as a wonder drug for many MS but faced major delays in light of a number of patients dying during trials... which nearly killed the company. A little genetic R&D and they may have been able to pinpoint those people who should not take the drug... allowing it to fly through FDA approvals.
I addressed the availability of "decent schools" in my post... and to suggest that providing access to a variety of reading material is pointless is just false.
I don't care how poor and isolated a person is, if there is a desire to learn, someone with the skill to teach, and books to be read, and time, then literacy will increase. And once literacy is established, the most important thing is access to a wealth of materials to read.
Years ago it was uncommon for even wealthy families to have access to more than a few books, this includes borrowing from family and friends. As publishing technology made books more readily available, literacy increased. The more literate the population, the more that having a large variety of books became valuable.
So essentially, yes there may be issues where many of the recipients of an OLPC cannot read the books initially. But lets say they learn, with help of a teacher, their peers, and their OLPC. As they gain skill, don't you think being able to research and study various fields of interest is a good thing.
Essentially, suggesting that this is a wasted effort is like suggesting that supporting your local library is stupid because only a small part of your community actually uses it.
If you reread my comment... I actually qualified my statement with "those who have had access to decent schools."
What I am suggesting is that the GGP's ignorant statement "can they even read?" was ignorant due to it's assumption that a third world country wouldn't benefit from this program because they must all be illiterate.
I suppose I should have mentioned that availability of a variety of reading materials has always led to increased literacy within a population where education and literacy is valued. Put as many libraries you want here in Detroit, and I doubt there will be more literacy, but put them in most 3rd world countries, and people will learn to read if they have the time.
No, the very definition of cheapest is "the cheapest solution that meets your needs".
Most of my machines are content with little more than enough storage to hold the OS and any data I am currently working with. Beyond that I use a large file server for the bulk of my data.
So in my case, if I can buy a 2TB magnetic drive for $200, or for the same price buy a faster, quieter, lower power solid state drive I would opt for the latter. In my case I am more concerned about price/transfer rate, price/power requirements, and price/noise level.
Finally, I never said that the "cloud" is the best place for data. Only that more and more people are moving to low powered devices that do little more than access internet resources... which further reduces the needs of having large storage devices on their workstations. Instead they want fast, quiet, low power, and portable storage.
Good question. His ignorant comment and the fact that I didn't have any Mod points left me no choice?
I guess I just get tired of ignorant, raciest, bigoted, and hateful people and hope that maybe, just maybe, a reply will convince them to actually reconsider their beliefs.
I just love how we in developed nations assume that those in the 3rd world are stupid. Actually, those who have had access to decent schools are quite likely smarter than you simply due to motivation. This has been proven time after time as students from developing nations visit our Universities and as a whole out perform our students by a tremendous margin, even with the cultural, language, and social barriers that they must overcome.
You do realize that x86 is a processor instruction set... it has been used by Intel, AMD, Cyrix, and many others. It is the instruction set that was first created by Intel with the 8086 processor and based upon other large instruction sets that proceeded it.
I suspect that they chose and x86 processor because there wasn't an ARM processor that was powerful enough to meet their needs. Even today, there isn't an ARM processor that can match even low end x86 processors from Intel or AMD. They are however very low power and pack tremendous performance/watt. Coupled with a good GPU that can be used for things like video decoding, ARM systems are finally becoming competitive for general computing tasks.
There are a number of reasons that the XO didn't take off... few of them were due to hardware decisions. Actually, the XO had some amazing hardware for the time. Unfortunately XO created a market for people wanting small low power portable computers and the netbook craze was started, which caused a huge drop in the prices making the $150 XO seem less appealing. I suspect that if they had just slapped a slightly modified Ubuntu on there as soon as the hardware was ready and started marketing it, they would have had a winner. But they spent months perfecting it, while others developed products that would compete with the XO.
Of course, the XO was never intended to be a hardware platform; it was a hardware/software platform that shipped with its own ideology about how it's users would interact with the computer and each other. Essentially it was as much a research project as it was a product.
No... it takes a year to perform optical character recognition on 1.6 Million Books so that they they only require a few kilobytes to be transmitted and stored rather than several megabytes.
Of course they will be cheapest, but only cheapest per/GB.
As time goes on, SSD's will be the default in desktops and laptops... mostly because these systems don't need very large drives... especially as we move more and more data to the "cloud".
Sure, per GB, magnetic storage will remain king when it comes to capacity, but they will only be used by those with extreme storage needs.
It is indeed double dipping... and it's brilliant.
It's like when you and a sibling or friend fought over a toy and your parents ripped it in half and gave you each a piece. No one wins unless one side relents. That's justice!
To me, a population is too large when its environment can no longer sustain the population. So, in many parts of the world, where people die of starvation especially, they are overpopulated. To argue that its a logistics problem is a fallacy. What if the entire earth were starving, would it be a logistics issue because we are not consuming the resources of a planet in a nearby solar system?
I personally think that we have already significantly overpopulated the earth, because we have modified our environment so greatly in order to sustain our population growth. So though there are resources available, they are not natural resources, and thus our current population exceeds the natural limits.
I don't propose that we put and end to reproduction in an effort to reduce populations to hunter-gatherer levels... but I would like to see a culture of reproductive responsibility encouraged. I did my part, my wife and I stopped after 2 children so we conformed to a zero population growth ideal. I wish everyone could share these ideals.
You missed something... I agree that the term is not clear and is commonly misused, but it does have a meaning.
A cloud service is one that is not provided by A server, but by many servers. Additionally to be considered a cloud service, it must be distributed geographically.
In the beginning computing was centralized, you would use a dumb terminal to access a mainframe system and all of your computing needs were centralized. Then with the PC, computing was distributed. Finally they centralized much of it again with web servers. Now with cloud computing they are actually distributing it again, however they are distributing the servers this time and presentation is still done on a relatively dumb device.
Cloud computing also creates a new networking model. Peer to peer and client-server have been around for ages... but the idea that the "server" can actually be many machines distributed all over the globe working together to provide the service that you perceive as the "server" in the client-server model is a pretty big shift.
Think back just a few years, there were almost no systems of servers that cooperatively worked together to provide a service that weren't within a single data center. Sure there were redundant data centers, but the entirety of your data was stored within a single location. With Gmail for example, your email is distributed across many datacenters. A few years ago the closest we came to distributed processing of a large scale was SETI@home, today they are introducing a product that will render a 3D image that would normally take 10's of minutes in a second or so by leveraging the processing power of thousands of computers around the world.
So "Cloud Computing" definitely has a meaning. Without it you would need to say "global scale distributing computing system".
Actually you couldn't... unless you calculated the acceleration and terminal velocity of the rolling ball and factored that into your time marks. If you spaced your marks equally, your clock would get faster and faster until the ball reached terminal velocity.
If I were on a deserted island, I would simply count time in my head and pace out the distance. So for example if I wanted to figure out how fast a river on my island is, I would pace out and mark two points along the shore... throw a stick in upstream and start counting "one one thousand, two one thousand, three one thousand" as soon as it passes the first and stop when it passes the second.
This velocity might be 100 paces in 20 "one thousands" or about 5 paces per "one thousand". If I had to guess it would be about 20ft per second... not a river I want to wade across.
Imagine a small home/workgroup server like this, but with iPhoto support so that everyone can share a photo database.
OSX server includes an iCalendar server, Address Book server, Mail Server, iChat server... so they have every other server component that a Mac Centric office would need, why no iPhoto server?
Actually, I have heard some great things about OSX server. You do realize at the core its just BSD, which makes some great servers. And if your going to have a GUI, why not the OSX gui rather than xwindows.
I compare OSX server to Windows Server, a bunch of crap running on top of a decent network operating system. If I had to choose between the two, I'd probably choose OSX if my environment needed a single file server for a mixed OS network.
For a small workgroup, it has a lot to offer for centralization and system management, similar to having windows machines in an AD environment. See http://www.apple.com/server/macosx/features/
I am certainly no Mac fanboi... In fact my only Mac runs Linux. But I can just imaging a simple plug-and-play file server that damn near sets itself up, has some redundancy, has built in monitoring and alerting, and hopefully integrates directly with Time Machine so that you can configure a simple backup system with versioning.
If this works as well as Apple's products have been known to do, it should save a lot of people a lot of headaches when it comes to backup and recovery.
Agreed on all counts. Only please understand that my justifications were more for the manufacturers than the consumers.
Essentially, if Android will do the same things as WinMo and Symbian, what incentive do manufacturers have to continue to invest money and energy in those products. I'm sure they won't disappear overnight, or even completely, but I have a feeling that Android is about to become far more popular that most people give it credit for, only because it will satisfy most users and it does it more cheaply than the competitors.
I am certain there will be some horrible implementations, as there have been with every OS. However if I made handsets and could standardize on a single OS that cost nothing and would run on every phone I produced, no matter how simple or complex that phone was I would use it. Oh and if I could sell mobile pc's running the same OS, and carputers, and GPS's, and.... you get my point.
Up until now, there wasn't an OS for low power embedded processors that specifically addressed small devices with a rich UI... at least not a free one. Sure Linux was there, but you needed to tweak the OS then write all of the presentation stuff, and finally develop an API if you wanted a community to develop around the product. Now all OEMs need to do is slap Android on there, customize it to their liking, maybe write a driver or two for custom hardware, and ship... no licensing, no contracts, no negotiations, and the customer will have a huge library of apps to run on it.
Ever since google formed the Android Handset Alliance, I knew they had revolutionized the portable electronics market... this is just the beginning.
Now that is a good argument in favor of less variety, and I agree that if HTC or Samsung, or whoever were to take their entire R&D budget and put it into a single phone, they may come up with an alternative that would out-sell a 10 phone lineup. Unfortunately, that is a lot of risk to take (what if it bombs) and the savings wouldn't be that substantial (most of the HTC phones have nearly identical circuitry, so they don't reinvent the wheel each time).
As far as compatibility is concerned... yes, more variety means fewer eyes on a single model which means there will be a few compatibility issues because a developer didn't consider that their app would be used with a trackball rather than touchscreen for example. However if a developer designs to the documented Android "Best Practices" http://developer.android.com/guide/practices/ui_guidelines/index.html then their application should be compatible with 99% of the Android phones.
Finally, the 50 handsets is 50 total across all carriers and markets. Many of the phones will have slightly different names and features and only be available with a single carrier. My guess is, if you go into a mobile store you won't find more than 2 Android phones from a single manufacturer for a given carrier.
First of all... some of the phones on that list have already been released, so we can assume they are being replaced with a newer model.
Next, there are more things that differentiate a phone than the fact it has a big screen and/or keyboard. One might have multitouch, another doesn't have a touch screen at all. One may not have a camera, some of us work in places where camera phones are not allowed. One may have a 3.5mm headphone jack, a memory card slot, a larger/smaller battery. And of course, they probably range in cost by a substantial amount.
Variety is good, what is good for you might not even be an option for me (especially if you want a camera on your phone). I couldn't care less about a 3.5mm headphone jack, or bluetooth, but I desperately want WiFi.
I know what your trying to say, but I have to disagree. I fully expect to see an Android phone take #1 in the future. Why? Because once these 50+ phones are out, and the 100's that follow, there will be far more users of Android phones than phones running the Mac or RIM OS.
At some point, the public will consider Android phones to be equal to the iPhone in features and capability, but they will have choice (Querty keyboard, carrier, camera, form factor, size, screen, cost, etc.). To many people that freedom, coupled with the features and usability they want, is more than enough to keep them away from iPhone.
For Android to compete with RIM, it needs to get serious about business. The good news is, that because Android is open source, and most contributors have real jobs, its capabilities in business will quickly surpass the Blackberry. Honestly, I have been with several companies that standardized on Blackberry, and other than mail and policy managment, the phone is a waste. If Android 2.0 gets the mail part right, RIM should be worried. If they introduce a policy management server... then RIM is in trouble.
Slashdot Mirror
Cloud computing is the client-server model that everyone is used to... but where the "server" is distributed.
There are significant advantages over a more traditional client-server model, even if the "server" is a cluster. Because the cloud is distributed geographically; 1. infrastructure outages are far less damaging to the application, 2. entire data centers can be taken off line and added at will, 3. power and cooling advantages can be used to keep costs lower, 4. Bandwidth utilization are distributed to multiple centers, which may help prevent saturation of regional carrier networks, 5. data may be read from the closest data center reducing bandwidth utilization vs a single data center on the other side of the globe.
Many things don't make sense to have on the cloud... but some things are particularly well suited for it. Email, I believe, is the ultimate cloud use... it needs high availability and fault tolerance. It needs huge capacities. And finally to keep storage utilization to a minimum, you will ideally have a single mail store, otherwise a single email sent to users on several mail stores will result in several copies of the same data. This kills Exchange in a lot of large organizations, a VP sends a large attachment to all of the regional head, now that same attachment exists on every exchange server in the organization, gets backed up with every server backup, and possibly never gets deleted since it came from the boss. A single mail store means a single copy (redundancy excluded, that exists with many mail stores too) of that attachemnt.
Actually... gas taxes are not a percentage of the sale price; instead they are $0.xx/gallon. Unfortunately that means that as gas prices go up, the government actually loses money because people buy less gas. Gas stations are the same way, many of them nearly went broke when prices soared to $3+ just because the credit card companies take their cut as a percentage while the gas station takes a fixed cut per gallon, with high priced gas, the credit card cut was often higher than the stations' cut.
From TFA:
There have also been a number of drugs under development, which were abandoned because they only benefited a small percentage of patients (say, for example, 30% of a hypothetical new drug), while carrying significant side effects. But just as those 10% of cancer patients got no benefit from tamoxifen because of a genetic variation, it might be the case that the 70% who didn't benefit from our hypothetical drug did so because they didn't have the correct genome. If we could identify the 30% that it worked for, what was once viewed as a failed drug could instead be a miracle drug, albeit for a subset of the entire patient population.
Imagine the research opportunities here. The R&D was done, the drug worked wonders in a small subset of patients and failed catastrophically in others... lets find out WHY. This idea could employ tens of thousands of researchers for several years just pouring over old trial data, running new trials, and linking genetic research to drug research.
I know if I were a pharma company I would be ramping up a similar effort. A good example is Elan's Tysabri, which was hailed as a wonder drug for many MS but faced major delays in light of a number of patients dying during trials... which nearly killed the company. A little genetic R&D and they may have been able to pinpoint those people who should not take the drug... allowing it to fly through FDA approvals.
I addressed the availability of "decent schools" in my post... and to suggest that providing access to a variety of reading material is pointless is just false.
I don't care how poor and isolated a person is, if there is a desire to learn, someone with the skill to teach, and books to be read, and time, then literacy will increase. And once literacy is established, the most important thing is access to a wealth of materials to read.
Years ago it was uncommon for even wealthy families to have access to more than a few books, this includes borrowing from family and friends. As publishing technology made books more readily available, literacy increased. The more literate the population, the more that having a large variety of books became valuable.
So essentially, yes there may be issues where many of the recipients of an OLPC cannot read the books initially. But lets say they learn, with help of a teacher, their peers, and their OLPC. As they gain skill, don't you think being able to research and study various fields of interest is a good thing.
Essentially, suggesting that this is a wasted effort is like suggesting that supporting your local library is stupid because only a small part of your community actually uses it.
If you reread my comment... I actually qualified my statement with "those who have had access to decent schools."
What I am suggesting is that the GGP's ignorant statement "can they even read?" was ignorant due to it's assumption that a third world country wouldn't benefit from this program because they must all be illiterate.
I suppose I should have mentioned that availability of a variety of reading materials has always led to increased literacy within a population where education and literacy is valued. Put as many libraries you want here in Detroit, and I doubt there will be more literacy, but put them in most 3rd world countries, and people will learn to read if they have the time.
No, the very definition of cheapest is "the cheapest solution that meets your needs".
Most of my machines are content with little more than enough storage to hold the OS and any data I am currently working with. Beyond that I use a large file server for the bulk of my data.
So in my case, if I can buy a 2TB magnetic drive for $200, or for the same price buy a faster, quieter, lower power solid state drive I would opt for the latter. In my case I am more concerned about price/transfer rate, price/power requirements, and price/noise level.
Finally, I never said that the "cloud" is the best place for data. Only that more and more people are moving to low powered devices that do little more than access internet resources... which further reduces the needs of having large storage devices on their workstations. Instead they want fast, quiet, low power, and portable storage.
I'm 32, and my first PC used ROM basic and I saved my programs to an audio cassette.
Good question. His ignorant comment and the fact that I didn't have any Mod points left me no choice?
I guess I just get tired of ignorant, raciest, bigoted, and hateful people and hope that maybe, just maybe, a reply will convince them to actually reconsider their beliefs.
And no... I am not holding my breath.
I just love how we in developed nations assume that those in the 3rd world are stupid. Actually, those who have had access to decent schools are quite likely smarter than you simply due to motivation. This has been proven time after time as students from developing nations visit our Universities and as a whole out perform our students by a tremendous margin, even with the cultural, language, and social barriers that they must overcome.
You do realize that x86 is a processor instruction set... it has been used by Intel, AMD, Cyrix, and many others. It is the instruction set that was first created by Intel with the 8086 processor and based upon other large instruction sets that proceeded it.
I suspect that they chose and x86 processor because there wasn't an ARM processor that was powerful enough to meet their needs. Even today, there isn't an ARM processor that can match even low end x86 processors from Intel or AMD. They are however very low power and pack tremendous performance/watt. Coupled with a good GPU that can be used for things like video decoding, ARM systems are finally becoming competitive for general computing tasks.
There are a number of reasons that the XO didn't take off... few of them were due to hardware decisions. Actually, the XO had some amazing hardware for the time. Unfortunately XO created a market for people wanting small low power portable computers and the netbook craze was started, which caused a huge drop in the prices making the $150 XO seem less appealing. I suspect that if they had just slapped a slightly modified Ubuntu on there as soon as the hardware was ready and started marketing it, they would have had a winner. But they spent months perfecting it, while others developed products that would compete with the XO.
Of course, the XO was never intended to be a hardware platform; it was a hardware/software platform that shipped with its own ideology about how it's users would interact with the computer and each other. Essentially it was as much a research project as it was a product.
No... it takes a year to perform optical character recognition on 1.6 Million Books so that they they only require a few kilobytes to be transmitted and stored rather than several megabytes.
Doesn't the OLPC have a B&W passive mode on it's display... the first version did.
Of course they will be cheapest, but only cheapest per/GB.
As time goes on, SSD's will be the default in desktops and laptops... mostly because these systems don't need very large drives... especially as we move more and more data to the "cloud".
Sure, per GB, magnetic storage will remain king when it comes to capacity, but they will only be used by those with extreme storage needs.
So long as there is a Free (not $$$ free) alternative, all they will do is push users to it.
It is indeed double dipping... and it's brilliant.
It's like when you and a sibling or friend fought over a toy and your parents ripped it in half and gave you each a piece. No one wins unless one side relents. That's justice!
First, define overpopulated.
To me, a population is too large when its environment can no longer sustain the population. So, in many parts of the world, where people die of starvation especially, they are overpopulated. To argue that its a logistics problem is a fallacy. What if the entire earth were starving, would it be a logistics issue because we are not consuming the resources of a planet in a nearby solar system?
I personally think that we have already significantly overpopulated the earth, because we have modified our environment so greatly in order to sustain our population growth. So though there are resources available, they are not natural resources, and thus our current population exceeds the natural limits.
I don't propose that we put and end to reproduction in an effort to reduce populations to hunter-gatherer levels... but I would like to see a culture of reproductive responsibility encouraged. I did my part, my wife and I stopped after 2 children so we conformed to a zero population growth ideal. I wish everyone could share these ideals.
You missed something... I agree that the term is not clear and is commonly misused, but it does have a meaning.
A cloud service is one that is not provided by A server, but by many servers. Additionally to be considered a cloud service, it must be distributed geographically.
In the beginning computing was centralized, you would use a dumb terminal to access a mainframe system and all of your computing needs were centralized. Then with the PC, computing was distributed. Finally they centralized much of it again with web servers. Now with cloud computing they are actually distributing it again, however they are distributing the servers this time and presentation is still done on a relatively dumb device.
Cloud computing also creates a new networking model. Peer to peer and client-server have been around for ages... but the idea that the "server" can actually be many machines distributed all over the globe working together to provide the service that you perceive as the "server" in the client-server model is a pretty big shift.
Think back just a few years, there were almost no systems of servers that cooperatively worked together to provide a service that weren't within a single data center. Sure there were redundant data centers, but the entirety of your data was stored within a single location. With Gmail for example, your email is distributed across many datacenters. A few years ago the closest we came to distributed processing of a large scale was SETI@home, today they are introducing a product that will render a 3D image that would normally take 10's of minutes in a second or so by leveraging the processing power of thousands of computers around the world.
So "Cloud Computing" definitely has a meaning. Without it you would need to say "global scale distributing computing system".
Actually you couldn't... unless you calculated the acceleration and terminal velocity of the rolling ball and factored that into your time marks. If you spaced your marks equally, your clock would get faster and faster until the ball reached terminal velocity.
If I were on a deserted island, I would simply count time in my head and pace out the distance. So for example if I wanted to figure out how fast a river on my island is, I would pace out and mark two points along the shore... throw a stick in upstream and start counting "one one thousand, two one thousand, three one thousand" as soon as it passes the first and stop when it passes the second.
This velocity might be 100 paces in 20 "one thousands" or about 5 paces per "one thousand". If I had to guess it would be about 20ft per second... not a river I want to wade across.
Imagine a small home/workgroup server like this, but with iPhoto support so that everyone can share a photo database.
OSX server includes an iCalendar server, Address Book server, Mail Server, iChat server... so they have every other server component that a Mac Centric office would need, why no iPhoto server?
Actually, I have heard some great things about OSX server. You do realize at the core its just BSD, which makes some great servers. And if your going to have a GUI, why not the OSX gui rather than xwindows.
I compare OSX server to Windows Server, a bunch of crap running on top of a decent network operating system. If I had to choose between the two, I'd probably choose OSX if my environment needed a single file server for a mixed OS network.
For a small workgroup, it has a lot to offer for centralization and system management, similar to having windows machines in an AD environment. See http://www.apple.com/server/macosx/features/
I am certainly no Mac fanboi... In fact my only Mac runs Linux. But I can just imaging a simple plug-and-play file server that damn near sets itself up, has some redundancy, has built in monitoring and alerting, and hopefully integrates directly with Time Machine so that you can configure a simple backup system with versioning.
If this works as well as Apple's products have been known to do, it should save a lot of people a lot of headaches when it comes to backup and recovery.
Agreed on all counts. Only please understand that my justifications were more for the manufacturers than the consumers.
Essentially, if Android will do the same things as WinMo and Symbian, what incentive do manufacturers have to continue to invest money and energy in those products. I'm sure they won't disappear overnight, or even completely, but I have a feeling that Android is about to become far more popular that most people give it credit for, only because it will satisfy most users and it does it more cheaply than the competitors.
I am certain there will be some horrible implementations, as there have been with every OS. However if I made handsets and could standardize on a single OS that cost nothing and would run on every phone I produced, no matter how simple or complex that phone was I would use it. Oh and if I could sell mobile pc's running the same OS, and carputers, and GPS's, and.... you get my point.
Up until now, there wasn't an OS for low power embedded processors that specifically addressed small devices with a rich UI... at least not a free one. Sure Linux was there, but you needed to tweak the OS then write all of the presentation stuff, and finally develop an API if you wanted a community to develop around the product. Now all OEMs need to do is slap Android on there, customize it to their liking, maybe write a driver or two for custom hardware, and ship... no licensing, no contracts, no negotiations, and the customer will have a huge library of apps to run on it.
Ever since google formed the Android Handset Alliance, I knew they had revolutionized the portable electronics market... this is just the beginning.
Now that is a good argument in favor of less variety, and I agree that if HTC or Samsung, or whoever were to take their entire R&D budget and put it into a single phone, they may come up with an alternative that would out-sell a 10 phone lineup. Unfortunately, that is a lot of risk to take (what if it bombs) and the savings wouldn't be that substantial (most of the HTC phones have nearly identical circuitry, so they don't reinvent the wheel each time).
As far as compatibility is concerned... yes, more variety means fewer eyes on a single model which means there will be a few compatibility issues because a developer didn't consider that their app would be used with a trackball rather than touchscreen for example. However if a developer designs to the documented Android "Best Practices" http://developer.android.com/guide/practices/ui_guidelines/index.html then their application should be compatible with 99% of the Android phones.
Finally, the 50 handsets is 50 total across all carriers and markets. Many of the phones will have slightly different names and features and only be available with a single carrier. My guess is, if you go into a mobile store you won't find more than 2 Android phones from a single manufacturer for a given carrier.
Ok... on that point.
First of all... some of the phones on that list have already been released, so we can assume they are being replaced with a newer model.
Next, there are more things that differentiate a phone than the fact it has a big screen and/or keyboard. One might have multitouch, another doesn't have a touch screen at all. One may not have a camera, some of us work in places where camera phones are not allowed. One may have a 3.5mm headphone jack, a memory card slot, a larger/smaller battery. And of course, they probably range in cost by a substantial amount.
Variety is good, what is good for you might not even be an option for me (especially if you want a camera on your phone). I couldn't care less about a 3.5mm headphone jack, or bluetooth, but I desperately want WiFi.
No, Android will never have #1, because it free.
I know what your trying to say, but I have to disagree. I fully expect to see an Android phone take #1 in the future. Why? Because once these 50+ phones are out, and the 100's that follow, there will be far more users of Android phones than phones running the Mac or RIM OS.
At some point, the public will consider Android phones to be equal to the iPhone in features and capability, but they will have choice (Querty keyboard, carrier, camera, form factor, size, screen, cost, etc.). To many people that freedom, coupled with the features and usability they want, is more than enough to keep them away from iPhone.
For Android to compete with RIM, it needs to get serious about business. The good news is, that because Android is open source, and most contributors have real jobs, its capabilities in business will quickly surpass the Blackberry. Honestly, I have been with several companies that standardized on Blackberry, and other than mail and policy managment, the phone is a waste. If Android 2.0 gets the mail part right, RIM should be worried. If they introduce a policy management server... then RIM is in trouble.