Like "when this method is called first time we know the second argument bogus". Or "the software is not capable of deleting all objects as there are pointers somewhere so make them kinda-static" (which leads to a very slightly erroneous behaviour rather than full collapse). Or "let's remove that mutex as it causes deadlock".
My point is that *IF* there will be several different mobile phones with incompatible APIs, Android will fail (at least in my eyes).
I do not care whether Android team says "we do not support that" - actually this just worsens the situation.
Apple's "3.0" is very, very close to previous, it is almost fully backward compatible and fully upward. That took over a year.
*IF*, as I read the comments, there are already lot of (enough to worry) incompatible API changes within first few months and *IF* the situation continues like that for a year then even Symbian will look very good indeed, in comparision.
Sure my text is 100% FUD, but I have learned to fear...
Actually I think that *IF* (huge if, btw) Java can pull it all together it has a possibility of being winner. It is interpreted so it does not have as big binary incompatibility issues as, say, C. Anyway it is already in most phones in the world, although I do admit some (most?) implementations are "criminally" lacking.
The poster I replied said "earlier / later version of the API". Sure, he was referring to documentation, but it is enough to show API is unstable.
Note: there already are Android phones around. They must use some API, or they would not function. If the API is still unstable as you say they have already failed: you cannot develop so that it works on *EVERY* Android phone, past, present and (near) future.
No, I have not developed myself (for several reasons), but this does not look good. Anyway I leave my judgement for 2010 - now it is too early to claim anything.
So maybe the cameras are positioned wrong or are not of high quality enough?
I would like to see a study why the cameras did not help. Too high in the ceiling so baseball cap obscures too much? Analog or low resolution so the picture is a mess? Couldn't catch license plate of run-away car?
Sorry to take one word out of the context but this might be the key issue. If there are several, incompatible versions of the API then there will not be a platform. Exactly this is one of the complaints of JavaME and Symbian and one big strength of Apple.
Unfortunately FOSS, and especially Linux, does not have a culture of keeping APIs unchanged, so whether Android can do it or not is yet to be seen.
Also, does anyone know what the long-term prospects are for the typical European medical system? Aren't they expecting to have solvency issues similar to medicare?
Short answer: no.
Long answer: "we" (not all Europeans are on par) do expect that the price will go up as more and more expensive treatments are developed. Sooner or later we have to deal with that, but at the moment it is not the biggest problem. Neither do we expect diabetes to skyrocket the costs.
I do not know why medicaid/medicare does not work in USA nor why the costs are hugely bigger than anywhere else in the world but it certainly is not "lavish lifestyle".
You should search for the kernel developer comment on why they do - purposefully - break drivers. Anyway you cannot copy driver from 2.6.y to 2.6.y+1, it will not be loaded.
I really have to make a web page: MTP008 - was in 2.4 was dropped in 2.6. Peracom USB ethernet adapter - stopped working while in kernel tree (i.e. was untested by the developers - probably did not give a shit) DiBcom 7070 based DVB-T tuners - manufacturer helped to make device driver ~5 years ago. Not in kernel tree. Various gspcav based USB video cameras - not in kernel tree. Several WiFi cards (e.g. EeePCs). Now in 9.04 works - years later.
I want you to tell me how I should do in the (hypothetical) case I want to make Linux laptop, say with a 3G chip (or touchscreen or...). Let's assume the chip is in testing phase and there is no support for it anywhere. I start the development soon and plan on releasing, say christmas 2010. Obviously I do not want to release specs to the wild as competitors would get me. I probably cannot release the specs especially if the chip is in development.
So what is the procedure to get the laptop out?
Wishes: I'd like to keep the source closed at least for a while so that big competitors could not immediately cut me. I'd like to support as many distributions as possible - after all people do have preferences and I cannot afford to do my own distribution anyway. Neither can I afford to keep compiling & testing the driver for every distribution and kernel version.
Hmm... maybe we should disallow people from saying "actor/actress in someobscurefilm sucks bad"!? Gimme a break! Saying only "skank" over anybody is so lame that the judge should have thrown the case out immediately.
Some units, like "mile" vary depending where you use it - nautical, aerial, british, official and for measuring gold. Or something like that, I really don't know all of them.
Actually here in north Celsius is better - crossing the zero degree does have some consequences which you are better to be aware. Not a big difference but still.
Comparing the peak power of GSM to average power of this system is misleading at best (gsm transmits only 1/8 of the time due to TDM) and a lie at worst (GSM power control almost always pushes the level 10-20dB below that).
Then 2W max is on 850/900MHz, 1800/1900 has max 1W.
if it blocks the radiation effectively enough to prevent reception sufficiently
I would not wear even if it did do that.
Reason: when phone is about to lose (or loses) contact with BTS the phone will increase TX power in order to avoid that. This can easily make situation (radiation dose) even worse.
$155/4h is exactly the reason companies cannot/will not do it - it is not profitable enough.
I myself have gotten some freebies from my ex-company, installed Linux on them and gave them away (for free) as net computers. Problem: display was usually bad. Only problem is that hard disk policy was getting too strict (it is not enough that *I* wipe the disk, it should have been wiped by the IT guys).
Well, anyway, that was the ex-company, in current I do not get to the "scrap yard" as I know nobody there.
The drivers did not work with people's favorite distro because the hardware Asus chose wasn't compatible with Linux.
I think the opposite: Linux was/is not compatible with the hardware and device drivers. This is the thing Linux does wrong. There is absolutely no good reason to purposefully break device drivers.
"Slows down the development" is bullshit argument - it does not slow down OS-X, Windows, OpenSolaris, *BSD,... As is "forces companies to open source drivers" - it affects *more* open source drivers than closed ones, it is more PITA for users than companies, and most importantly, open sourcing the drivers does not help at all (there are drivers which has been open sourced ~5 years ago and still they pose problems to users).
Slashdot Mirror
Exactly.
Like "when this method is called first time we know the second argument bogus".
Or "the software is not capable of deleting all objects as there are pointers somewhere so make them kinda-static" (which leads to a very slightly erroneous behaviour rather than full collapse).
Or "let's remove that mutex as it causes deadlock".
They are trying to get the courts to rule on whether Global Warming will be _harmful_ to humans.
Which will be exactly as useful as rulings which said smoking is harmless to humans.
My point is that *IF* there will be several different mobile phones with incompatible APIs, Android will fail (at least in my eyes).
I do not care whether Android team says "we do not support that" - actually this just worsens the situation.
Apple's "3.0" is very, very close to previous, it is almost fully backward compatible and fully upward. That took over a year.
*IF*, as I read the comments, there are already lot of (enough to worry) incompatible API changes within first few months and *IF* the situation continues like that for a year then even Symbian will look very good indeed, in comparision.
Sure my text is 100% FUD, but I have learned to fear ...
Actually I think that *IF* (huge if, btw) Java can pull it all together it has a possibility of being winner. It is interpreted so it does not have as big binary incompatibility issues as, say, C. Anyway it is already in most phones in the world, although I do admit some (most?) implementations are "criminally" lacking.
They place them at intersections to catch red light runners and in parked cars to catch speeders.
Good, they deserve to be fined.
Yours, "many times almost hit by a speeding red running car".
we have the technical know-how and experience.
To go to Mars? I welcome whatever-country-you-are-from as superior overlords.
I think I used too many negatives. Let me rephrase myself: "I think there are only few who think current copyright law is good".
PiratPartiet wants all creative works to go "public domain" as long as it is used non-commercially.
This is either untenable for entrepreneurs or double standards - I think latter.
The poster I replied said "earlier / later version of the API". Sure, he was referring to documentation, but it is enough to show API is unstable.
Note: there already are Android phones around. They must use some API, or they would not function. If the API is still unstable as you say they have already failed: you cannot develop so that it works on *EVERY* Android phone, past, present and (near) future.
No, I have not developed myself (for several reasons), but this does not look good. Anyway I leave my judgement for 2010 - now it is too early to claim anything.
Hint: creating information is a service people will gladly pay for...
Really? Every newspaper I know has dropped charges for the net version as far too few subscribe.
I do not think there are many individuals who do not think current copyright law "is wrong".
Still I do not think there are many who think that law where released music/computer programs/... go immediately to public domain is better.
So ThePirateBay, I cannot support them (anyway I do not warez). Nor can I support PiratPartiet. They are just too ... anarchy.
Because they are not accused of hosting the files.
They are accused of "helping in copyright infringement". In Scandinavia helping is often a lesser crime compared to the act, but it is a crime.
So maybe the cameras are positioned wrong or are not of high quality enough?
I would like to see a study why the cameras did not help. Too high in the ceiling so baseball cap obscures too much? Analog or low resolution so the picture is a mess? Couldn't catch license plate of run-away car?
I have no clue what such a study would reveal.
platform
Sorry to take one word out of the context but this might be the key issue. If there are several, incompatible versions of the API then there will not be a platform.
Exactly this is one of the complaints of JavaME and Symbian and one big strength of Apple.
Unfortunately FOSS, and especially Linux, does not have a culture of keeping APIs unchanged, so whether Android can do it or not is yet to be seen.
You mean mobile phone, don't you?
Sorry NextPark does not have English pages, but here we go :-) http://www.nextpark.com/
Also, does anyone know what the long-term prospects are for the typical European medical system? Aren't they expecting to have solvency issues similar to medicare?
Short answer: no.
Long answer: "we" (not all Europeans are on par) do expect that the price will go up as more and more expensive treatments are developed. Sooner or later we have to deal with that, but at the moment it is not the biggest problem. Neither do we expect diabetes to skyrocket the costs.
I do not know why medicaid/medicare does not work in USA nor why the costs are hugely bigger than anywhere else in the world but it certainly is not "lavish lifestyle".
You should search for the kernel developer comment on why they do - purposefully - break drivers.
Anyway you cannot copy driver from 2.6.y to 2.6.y+1, it will not be loaded.
I really have to make a web page:
MTP008 - was in 2.4 was dropped in 2.6.
Peracom USB ethernet adapter - stopped working while in kernel tree (i.e. was untested by the developers - probably did not give a shit)
DiBcom 7070 based DVB-T tuners - manufacturer helped to make device driver ~5 years ago. Not in kernel tree.
Various gspcav based USB video cameras - not in kernel tree.
Several WiFi cards (e.g. EeePCs). Now in 9.04 works - years later.
I want you to tell me how I should do in the (hypothetical) case I want to make Linux laptop, say with a 3G chip (or touchscreen or ...). Let's assume the chip is in testing phase and there is no support for it anywhere. I start the development soon and plan on releasing, say christmas 2010. Obviously I do not want to release specs to the wild as competitors would get me. I probably cannot release the specs especially if the chip is in development.
So what is the procedure to get the laptop out?
Wishes: I'd like to keep the source closed at least for a while so that big competitors could not immediately cut me. I'd like to support as many distributions as possible - after all people do have preferences and I cannot afford to do my own distribution anyway. Neither can I afford to keep compiling & testing the driver for every distribution and kernel version.
So what is your advice?
If buyer can get money back by claiming "empty box" the system must go down fast.
There are just too many people on earth willing to do that repeatedly - especially when they do not know the seller.
Hmm ... maybe we should disallow people from saying "actor/actress in someobscurefilm sucks bad"!?
Gimme a break!
Saying only "skank" over anybody is so lame that the judge should have thrown the case out immediately.
It is worse that that.
Some units, like "mile" vary depending where you use it - nautical, aerial, british, official and for measuring gold. Or something like that, I really don't know all of them.
Actually here in north Celsius is better - crossing the zero degree does have some consequences which you are better to be aware. Not a big difference but still.
But it does have one distinct advantage: it is now "cool" to take iPhone from your pocket, put it on a table and pretend-to-do-something with it.
I saw quite a few examples in bars and cafes.
Comparing the peak power of GSM to average power of this system is misleading at best (gsm transmits only 1/8 of the time due to TDM) and a lie at worst (GSM power control almost always pushes the level 10-20dB below that).
Then 2W max is on 850/900MHz, 1800/1900 has max 1W.
if it blocks the radiation effectively enough to prevent reception sufficiently
I would not wear even if it did do that.
Reason: when phone is about to lose (or loses) contact with BTS the phone will increase TX power in order to avoid that. This can easily make situation (radiation dose) even worse.
$155/4h is exactly the reason companies cannot/will not do it - it is not profitable enough.
I myself have gotten some freebies from my ex-company, installed Linux on them and gave them away (for free) as net computers. Problem: display was usually bad.
Only problem is that hard disk policy was getting too strict (it is not enough that *I* wipe the disk, it should have been wiped by the IT guys).
Well, anyway, that was the ex-company, in current I do not get to the "scrap yard" as I know nobody there.
The drivers did not work with people's favorite distro because the hardware Asus chose wasn't compatible with Linux.
I think the opposite: Linux was/is not compatible with the hardware and device drivers. This is the thing Linux does wrong. There is absolutely no good reason to purposefully break device drivers.
"Slows down the development" is bullshit argument - it does not slow down OS-X, Windows, OpenSolaris, *BSD, ... As is "forces companies to open source drivers" - it affects *more* open source drivers than closed ones, it is more PITA for users than companies, and most importantly, open sourcing the drivers does not help at all (there are drivers which has been open sourced ~5 years ago and still they pose problems to users).