The most bootloader "locking" I've seen with Samsungs is the "custom binary count" counter/warning in newer phones. It'll rat on you that you flashed a custom kernel via ODIN, but it will happily boot it.
This goes for every GalaxyS variant (including the Epic 4G) - although most of these don't have the "new" custom binary count feature, all interim devices (like the Infuse and Droid Charge), and all GS2 variants (which all have CBC counters - but happily boot custom kernels.)
So if anyone is eating your foot, it's you. It's clear that you either don't have a damn clue or you're trolling.
As to my credentials in this regard - I maintain custom kernels for the AT&T Infuse and AT&T GSII. I know for a fact that the original Epic 4G had unlocked bootloaders and PLENTY of custom kernels because I based some of the features in my Infuse kernel series on an Epic 4G kernel series.
They want people to blame someone other than them, when it is CLEARLY *their* fault.
Just look at it this way: Samsung - no locked bootloaders on any carrier (The Tab 10.1 is semi-locked, but that's even the wifi-only versions - and it's pretty light locking.) HTC - Heavy bootloader locking only recently, and across all carriers. The Thunderbolt is just a bit higher profile, as it was the first one. Motorola - All locked, all the time, with very rare exceptions, on all carriers - even European ones! In fact, with the original Droid, the Verizon variant was the only one WITHOUT a locked bootloader (the internation GSM variant, the Milestone, had a locked bootloader.)
Huh? How would Intel pull something like that with GCC - next to impossible.
Also the stated "crippling" you talk about would make the P4 look significantly WORSE than the P3, not better. The P4 is the first chip in which x87 instructions got crippled in favor of using SSE. (In the P4, the recommended method for floating point was SSE - even for non-SIMD work.)
And key aspects of that major disaster: 1) A fundamentally dangerous reactor core design that has never been legal to use for power generation in the United States. Positive void coefficient = no NRC approval. Period. (Exception: Possibly some military reactors, but no civilian ones.) 2) They were running a dangerous experiment and overrode multiple safety protocols - they were under pressure to achieve Great Success - OR ELSE. The guys at the controls wanted to SCRAM it and be done, but the shift supervisor overrode them because he was a good Party man.
Technically, Fukushima is the first civilian power reactor to release more than a few bananas' worth of radiation. Chernobyl may have been a civilian reactor officially - but if you look at its design, it was clearly intended to be suitable for weapons production and had its safety features compromised as a result. You only build a reactor like that if you want online refueling - and the primary benefit of online refueling is to reduce Pu-240 production in favor of more "boom-friendly" plutonium isotopes.
So, since people fight the construction of new plants tooth and nail - how do you propose they do this?
All of the design improvements you state have already been made. But people fight construction of new plants (even if it is to replace old ones), resulting in old clunkers like Fukushima staying in service.
You do realize that Unit 1 at Fukushima was one of the oldest operating reactors on the planet, and was originally supposed to start the decommissioning process days before the quake but received a service life extension because no one would allow modernized plants to be built?
ESBWRs would have shrugged off the tsunami without problems (fully passive cooling - provisions for 3 days of cooling with no intervention, and all you need beyond that is a fire truck to refill the isolation condenser pools). Even ABWRs likely would have been fine, as they have a backup gas turbine in one of the buildings in addition to the outdoor diesels.
There's also the issue of cooling water. If you build inland, you need large cooling towers. If you don't - just dump waste heat into the ocean.
In France, they're a little more responsible - even plants with plenty of river cooling water have towers in order to reduce thermal impacts on the rivers.
I think that's why the Fukushima plants were fairly low, even though if you look at the plant layout - Just a few hundred feet back would've put them on the order of 5-10 meters higher at least - that 5-10m extra elevation would have required more powerful coolant pumps.
Problem is, this particular project wastes the power of ARM by using it to run a BASIC intepreter.
Target a 1768 directly with a C compiler and you have a very powerful chip that is still possible to solder yourself at home (e.g. you can design your own custom board around it.)
Good luck soldering the Package-on-Package BGA stack (yes, that's multiple chips stacked on top of each other) in the BeagleBoard.
I'm not sure if HC source is guaranteed when that happens, as it isn't their full internal source repo that's being made public.
However, I wouldn't be surprised if Google does this out of goodwill, especially considering that their reasoning for withholding the source will no longer be valid.
In fact, I think at one point I saw Google claim that not only would ICS source be released, but Honeycomb source would be released - but NOT until after ICS source went live.
Google has stated many times, over and over again, why HC sources were withheld.
Android source releases in the past have NEVER been provided on announcement day.
As you've pointed out, it usually happens after a device with the new version is released, and don't expect it on device launch day either - usually it takes a couple of weeks, but it's rarely withheld longer than that (Honeycomb being the exception, an exception which Google explicitly stated was an explanation and why it was an exception.)
Complicating the issue is the fact that the AOSP repos have traditionally been hosted on kernel.org - which is STILL at least partially hosed. As the kernel.org saga goes forwards, more and more people that were hoping for it to return quickly have migrated - see Linus moving to github, and as of 2-3 weeks ago, the AOSP tress finally got mirrored over to github, but I'm not sure if that was an official move. (It was impossible to sync a full Cyanogenmod tree from scratch for at least a month due to the fact that the components of Android CM doesn't modify get synced directly from the original repos.)
Was it one of their early shutter-glasses implementations, or the newer "3D Vision" stuff.
Early shutter-glasses implementations were often card-manufacturer-specific, used 60Hz monitors (reducing per-eye refresh to 30 Hz), and were driver hell.
3D Vision is, based on all I've read, majorly improved. The only reason I haven't tried it is due to the lack of 3D Vision-capable monitors - I'm NOT dropping to a 20-22" monitor for 3D! It looks like supposedly monitor selection is improved, but it bothers me that the list of compatible monitors didn't change for at least three years.
Yup. OpenOffice was dead once it forked hard, XFree86 style.
Oracle separating themselves from OO was too little, too late - by the time Oracle stopped meddling, the project was already dead.
OO being dead doesn't really matter that much other than the fact that LibreOffice is a rather lame name which will probably inhibit corporate acceptance in some organizations. LibreOffice just has too many idealistic/propaganda connotations in the name - it makes it sound like it came from a bunch of RMS-style nutjobs (even if it didn't).
What I read indicates he's retired from physical acting but not voice acting... Which is consistent with him having a major role in a summer blockbuster movie this year, however in voice-only form. (Sentinel Prime in Transformers 3.)
I found it amusing how they worked a Spock quote into Nimoy's lines in Transformers 3.
You're an idiot. It's silly to think that someone would sign a contract for more than 2 years for a device that will be obsolete in 2, and it's silly to think that a person isn't going to go for an upgrade at the end of those 2 years. (Again - product obsolescence.)
Also, your $105/month estimate is WAY high. This is Sprint we're talking about, typically one of the cheaper carriers. Data plans are usually $20-30/month. Voice plan shouldn't be counted, since the assumption is that someone getting a smartphone probably had voice service already.
I think they're talking about custom Linux distros for low-end phones for those who have gone with solutions other than Android at the high end of the market segment.
Slashdot Mirror
Don't forget Heimdall for Mac/Linux users. Heck - I know a lot of ROM devs use it on Windows too because it's just plain more consistent.
Um, it isn't?
The most bootloader "locking" I've seen with Samsungs is the "custom binary count" counter/warning in newer phones. It'll rat on you that you flashed a custom kernel via ODIN, but it will happily boot it.
This goes for every GalaxyS variant (including the Epic 4G) - although most of these don't have the "new" custom binary count feature, all interim devices (like the Infuse and Droid Charge), and all GS2 variants (which all have CBC counters - but happily boot custom kernels.)
So if anyone is eating your foot, it's you. It's clear that you either don't have a damn clue or you're trolling.
As to my credentials in this regard - I maintain custom kernels for the AT&T Infuse and AT&T GSII. I know for a fact that the original Epic 4G had unlocked bootloaders and PLENTY of custom kernels because I based some of the features in my Infuse kernel series on an Epic 4G kernel series.
Moto is just trying to find an excuse...
They want people to blame someone other than them, when it is CLEARLY *their* fault.
Just look at it this way:
Samsung - no locked bootloaders on any carrier (The Tab 10.1 is semi-locked, but that's even the wifi-only versions - and it's pretty light locking.)
HTC - Heavy bootloader locking only recently, and across all carriers. The Thunderbolt is just a bit higher profile, as it was the first one.
Motorola - All locked, all the time, with very rare exceptions, on all carriers - even European ones! In fact, with the original Droid, the Verizon variant was the only one WITHOUT a locked bootloader (the internation GSM variant, the Milestone, had a locked bootloader.)
Huh? How would Intel pull something like that with GCC - next to impossible.
Also the stated "crippling" you talk about would make the P4 look significantly WORSE than the P3, not better. The P4 is the first chip in which x87 instructions got crippled in favor of using SSE. (In the P4, the recommended method for floating point was SSE - even for non-SIMD work.)
And key aspects of that major disaster:
1) A fundamentally dangerous reactor core design that has never been legal to use for power generation in the United States. Positive void coefficient = no NRC approval. Period. (Exception: Possibly some military reactors, but no civilian ones.)
2) They were running a dangerous experiment and overrode multiple safety protocols - they were under pressure to achieve Great Success - OR ELSE. The guys at the controls wanted to SCRAM it and be done, but the shift supervisor overrode them because he was a good Party man.
Technically, Fukushima is the first civilian power reactor to release more than a few bananas' worth of radiation. Chernobyl may have been a civilian reactor officially - but if you look at its design, it was clearly intended to be suitable for weapons production and had its safety features compromised as a result. You only build a reactor like that if you want online refueling - and the primary benefit of online refueling is to reduce Pu-240 production in favor of more "boom-friendly" plutonium isotopes.
So, since people fight the construction of new plants tooth and nail - how do you propose they do this?
All of the design improvements you state have already been made. But people fight construction of new plants (even if it is to replace old ones), resulting in old clunkers like Fukushima staying in service.
You do realize that Unit 1 at Fukushima was one of the oldest operating reactors on the planet, and was originally supposed to start the decommissioning process days before the quake but received a service life extension because no one would allow modernized plants to be built?
ESBWRs would have shrugged off the tsunami without problems (fully passive cooling - provisions for 3 days of cooling with no intervention, and all you need beyond that is a fire truck to refill the isolation condenser pools). Even ABWRs likely would have been fine, as they have a backup gas turbine in one of the buildings in addition to the outdoor diesels.
There's also the issue of cooling water. If you build inland, you need large cooling towers. If you don't - just dump waste heat into the ocean.
In France, they're a little more responsible - even plants with plenty of river cooling water have towers in order to reduce thermal impacts on the rivers.
I think that's why the Fukushima plants were fairly low, even though if you look at the plant layout - Just a few hundred feet back would've put them on the order of 5-10 meters higher at least - that 5-10m extra elevation would have required more powerful coolant pumps.
Y U NO FIX? requests are pretty useless without knowing what's broken.
Even older versions of Linux (such as the kernels included with Ubuntu 11.04) work just fine on Sandy Bridge - I just upgraded and it's great.
So whatever's broken for you is obviously some specific corner case which you haven't bothered to specify.
Problem is, this particular project wastes the power of ARM by using it to run a BASIC intepreter.
Target a 1768 directly with a C compiler and you have a very powerful chip that is still possible to solder yourself at home (e.g. you can design your own custom board around it.)
Good luck soldering the Package-on-Package BGA stack (yes, that's multiple chips stacked on top of each other) in the BeagleBoard.
Most of these devices are targeted natively by C compilers instead of interpreting BASIC.
You can target even Atmel's ATTinys with C. (2k flash, I think 512 bytes RAM, 8 bit core for the Tiny25.)
I'm not sure if HC source is guaranteed when that happens, as it isn't their full internal source repo that's being made public.
However, I wouldn't be surprised if Google does this out of goodwill, especially considering that their reasoning for withholding the source will no longer be valid.
In fact, I think at one point I saw Google claim that not only would ICS source be released, but Honeycomb source would be released - but NOT until after ICS source went live.
Google has stated many times, over and over again, why HC sources were withheld.
Note that Google has done nothing of the sort - anything that's GPL has been released in full compliance.
People seem to forget that the majority of Android userland is Apache-licensed.
Technically the master mirrors were all taken down temporarily - not by Google but by the asshat that compromised kernel.org
Android source releases in the past have NEVER been provided on announcement day.
As you've pointed out, it usually happens after a device with the new version is released, and don't expect it on device launch day either - usually it takes a couple of weeks, but it's rarely withheld longer than that (Honeycomb being the exception, an exception which Google explicitly stated was an explanation and why it was an exception.)
Complicating the issue is the fact that the AOSP repos have traditionally been hosted on kernel.org - which is STILL at least partially hosed. As the kernel.org saga goes forwards, more and more people that were hoping for it to return quickly have migrated - see Linus moving to github, and as of 2-3 weeks ago, the AOSP tress finally got mirrored over to github, but I'm not sure if that was an official move. (It was impossible to sync a full Cyanogenmod tree from scratch for at least a month due to the fact that the components of Android CM doesn't modify get synced directly from the original repos.)
They reduced the on/off transition times significantly - not an easy thing to do.
Was it one of their early shutter-glasses implementations, or the newer "3D Vision" stuff.
Early shutter-glasses implementations were often card-manufacturer-specific, used 60Hz monitors (reducing per-eye refresh to 30 Hz), and were driver hell.
3D Vision is, based on all I've read, majorly improved. The only reason I haven't tried it is due to the lack of 3D Vision-capable monitors - I'm NOT dropping to a 20-22" monitor for 3D! It looks like supposedly monitor selection is improved, but it bothers me that the list of compatible monitors didn't change for at least three years.
Because you shouldn't be pushing an agenda when you name your software.
OpenOffice wasn't too bad because Open is such an overloaded and diluted word that no one cares about it.
Free has negative connotations of "crap"
Libre has negative connotations of "those people are probably a bunch of zealots like RMS".
Yup. OpenOffice was dead once it forked hard, XFree86 style.
Oracle separating themselves from OO was too little, too late - by the time Oracle stopped meddling, the project was already dead.
OO being dead doesn't really matter that much other than the fact that LibreOffice is a rather lame name which will probably inhibit corporate acceptance in some organizations. LibreOffice just has too many idealistic/propaganda connotations in the name - it makes it sound like it came from a bunch of RMS-style nutjobs (even if it didn't).
I think the problem is that Oracle has already shown a long history of NOT cooperating with the open source community.
The thought process was likely,
"Damn that driver is crap."
"That driver is coming from Oracle - so upstream will be nonresponsive."
What I read indicates he's retired from physical acting but not voice acting... Which is consistent with him having a major role in a summer blockbuster movie this year, however in voice-only form. (Sentinel Prime in Transformers 3.)
I found it amusing how they worked a Spock quote into Nimoy's lines in Transformers 3.
Plus keep in mind that the focus on iD games is typically multiplayer and not single player. (Doom 3 being a bit of an exception.)
Look at Quake 3 - it was pretty much 100% a multiplayer game, where the SP storyline was just multiplayer games against bots. It still did VERY well.
If the SP storyline sucks, I don't really care, if the MP is good.
Betting on carrier-jumping with the original iPhone was a safe bet.
Now - not so safe, especially considering that iPhone market share is stalled and Android is growing.
You're an idiot. It's silly to think that someone would sign a contract for more than 2 years for a device that will be obsolete in 2, and it's silly to think that a person isn't going to go for an upgrade at the end of those 2 years. (Again - product obsolescence.)
Also, your $105/month estimate is WAY high. This is Sprint we're talking about, typically one of the cheaper carriers. Data plans are usually $20-30/month. Voice plan shouldn't be counted, since the assumption is that someone getting a smartphone probably had voice service already.
I think they're talking about custom Linux distros for low-end phones for those who have gone with solutions other than Android at the high end of the market segment.
The torque curve of electric motors (effectively constant torque from stop up to peak horsepower) makes CVTs irrelevant in all-electric cars.
In fact, Toyota's Hybrid Synergy Drive effectively implements a CVT using the electric motors and a fixed planetary gearset.