It's not like devs are putting their code under BSD with the expectation that companies will share back, only to be in total shock and feeling scorned when a company doesn't. If they cared about that issue as much as the loud people on/. did, they'd be using GPL.
/. can keep waging the license war, while devs will continue to ignore it and license their code in whatever way they're comfortable with.
Clang aims for more integration than "compile & debug". IDEs can directly use Clang libraries for syntax highlighting and error diagnostics. It was my understanding that GCC lacked this functionality and an old, large, and complex code base made it unrealistic for such a large feature addition to be worked on by mere mortals.
I'm sure not having to work with GPL is a definite plus for Apple, but this is the first time I've heard of licensing being a big reason for Clang's existence.
Have you tried x264 with --preset veryfast? My experience is that x264 is able to match a GPU encoder's speed while still giving significantly better quality. I'd only bother with a GPU encoder if I had a terrible CPU (netbook, phone?).
I'm a working software developer and I have no degree, only a hell of a passion for coding and the ability to learn on my own.
It's possible. There are plenty of places who will be looking for cheap junior software devs. The work won't be terribly interesting, but they're excellent places to jump start a career.
A recruiting agency like Volt is an easy place to start. You don't pay them a dime. I recommend grabbing some certifications – with a lack of experience or degree, anything can help get your foot in the door to an in-person interview. Even ones from Brainbench will help, and they're pretty cheap.
It's not the music that matters so much as the mixer/DAC.
High bit-depth playback matters greatly for any system that controls volume at the mixer stage rather than the amp stage. This is very important for PCs, where it is common to keep the amp (speakers) at a fixed volume and control the actual listening volume from the operating system's mixer.
If you keep the volume all the way up on your mixer, controlling your listening only at the amp stage, then a 16-bit pipeline is plenty. Highly integrated hardware such as an MP3 player can easily get away with this.
Otherwise, a 24-bit mixer and DAC can be very useful as it allows you to turn the volume down really far on the mixer while still retaining all the detail of your 16-bit music for the amp to boost back to listenable levels. It's still not perfect (such a low line level will inevitably be noisier) but it's still much better than a 16-bit pipeline.
E-books are not made by readers.
on
The eBook Backlash
·
· Score: 5, Insightful
The only problem with e-books and e-readers is that they're clearly not made by readers.
Books, the good ones at least and most of the bad ones too, pay attention to typography. Paragraph-optimized justification, hyphenation, hanging punctuation, ligatures, etc. All these little things that you take for granted with a dead-tree book, but without them it's a significantly poorer experience.
You find books with left-aligned text, an ugly and jagged right edge carving out a large chunk of empty space on the right. Or worse, you get one that is justified. This is bottom-of-the-barrel justification, without hyphenation and very commonly leaving huge spaces between words.
I've owned a Nook since launch day. I've read a large number of books on it, and I love it. But there is still a lot of room for improvement. I shouldn't need to import my ebooks into Adobe InDesign to make a PDF with proper typography.
They can put whatever positive, law-cooperating spin on it that they want, but the real reason is simply that Rapidshare no longer has any competition to drive up their free download speeds.
Their single largest competitor (Megaupload) just dropped off the map and all their other competitors have either blocked the US or dropped any free support whatsoever.
"We also show that as far as true performance is based - real-world performance as seen by the end-user- 2.4 is as fast, and even faster than some of the servers who may be "better" known as being "fast", like nginx," Jagielski said.
What's with the quotes? Other servers have proven to be faster, lighter weight, and more scalable than Apache for a long time. Don't be bitter because you fell behind. Be happy that you're finally catching up.
They mean to replace multiple per-thread copies of data with one single copy of it, accessible by all threads. No doubt part of Mozilla's latest push to reduce memory consumption.
One feature of x86 is that, save for a specialized SSE streaming store instruction, any store made by one core is immediately visible to all the other cores—even when the old value was already in a core's cache.
Maintaining such cache coherency involves a lot of overhead, so to get better scalability multi-threaded apps will sometimes adopt a "share-nothing" model: all threads get their own copy of the data, and no other threads will ever get to touch it. You trade memory and complexity for speed.
It sounds like Mozilla has decided this trade off is no longer worth it, and so has done away with multi-threading all together. Perhaps they will use green threads instead of native threads, though that brings along its own bag of complexities.
If you've got a Samsung phone, head on over to WindowBreak. It'll give you developer access and native execution abilities, even starting from a locked-down 7.5 (Mango).
It certainly stresses the video card, but I don't know anyone who acquired it because of this.
Personally I think Crysis and Crysis 2 were quite fun and even replayable. The story is forgettable, but the gameplay is great. A while after beating it playing full stealth, I revisited Crysis and attempted to rambo using a mix of all the suit abilities -- speed in, leap into the air, and knock some enemies out with all your strength.
Lots of my friends have now taken to waiting for GOTY editions to release so they can get the full game and all expansions/DLC for one sane price. People with this attitude may account for less initial sales, helping piracy appear disproportionate.
Still, 3.9 million is quite an impressive number. It will no doubt fuel Ubisoft's want for even more retarded DRM.
These days, I could. Because the 80s and 90s were something of a fluke in which hardware was progressing at a rapid rate, it coincided with the growth of the video game industry and attracted a lot of hardware geeks. But that era is gone, and hardware has stabilized to the point where new games are coming out targeting five year old hardware, and most people are okay with it. Skyrim runs on my first-generation Intel iMac from 2006.
For better or worse, there's a reason PC games have no issues on five year old hardware, and it's not because hardware has slowed down.
99% of PC games released now are ports from the consoles. Usually the only extras you get on PC are the trivially easy ones: support for higher resolutions, higher fps, and larger textures. If games didn't need to worry about running on ancient console hardware, they'd be able to look a lot more stunning on the PC.
There's also the problem of Windows' graphics stack introducing a huge amount of overhead to keep things safe, whereas consoles give more direct access to the hardware. Carmack guessed that GPUs in PCs are at least 10x more powerful than the ones in consoles, but are severely limited in how much new data you can give them every frame due to this overhead.
I've been a long-time DD-WRT user, but its development seemed to stagnate. I recently put TomatoUSB on my Linksys WRT160N v1, and it is working wonderfully. The interface is much nicer, and exposes more QoS and bandwidth management features which I've found useful.
Check out the TomatoUSB website for a list of routers it supports.
Take a short miniseries and stretch it out to a full series to milk it and make the plot run like molasses. Add in a lot of mindless action with big explosions and helicopter chases, because that's what American shows look like, right?
I'm afraid of what they're going to do to Doctor Who, but if Torchwood was any example, keep Hollywood's dirty hands off it.
WinRT doesn't replace.NET or the Win32 API, nor is it a second-class citizen when used from C#.
Metro apps are intended to be small—about the size of a cell phone app. They aren't made to replace full desktop apps meant to get work done. The.NET APIs you get for Metro apps have been scrubbed, getting rid of "insecure" APIs as well as ones that WinRT made redundant. A lot of the.NET API you've become used to is still available, though.
It uses COM behind the scenes, but there's no hint of COM in the API -- in fact, people in the Mono project have already begun work on a purely managed WinRT implementation. It definitely doesn't feel like a second-class citizen while using it. I'll be curious to see how much overhead it introduces.
WinRT in C++ is a little weird. There are two ways to use it: WinRTL, which is written in standard C++ but has all the verbosity of COM, or C++/CX, some extensions to C++ that make using it a breeze if you don't care about portability. It doesn't replace the Win32 API either. They encourage you to use WinRT when possible, but the entire Win32 API is still available. It's not clear if they'll be filtering the APIs you're allowed to use during the review process.
The biggest problem with Metro apps is that you can't just distribute an exe—all distribution needs to go through the Windows app store, complete with $100/year fees and review process. I'm sure a hack will come out, but this might make life a bit more difficult for many hobby developers trying to reach the common user.
Not quite. Seagate's tech is a simple block cache, where the most frequently accessed blocks get thrown on the SSD portion. It is no doubt quite effective, but we should be able to do better if we move the logic into the OS.
The OS has intimate knowledge of the filesystem, and can easily profile its use. Files that are small or randomly accessed should get put on the SSD, while large sequentially accessed files should get put on the HDD.
The Raspberry Pi isn't exactly the same thing -- it does not include a case, keyboard, LCD, or speakers. But, you can probably get all that stuff for another $25. So maybe the OLPC has a new partner.
I've had a Windows Phone 7 device since day one, and it asks at every turn before doing something that would collect location (or any other) data. If it's true and not just a misconfigured device or data being poorly interpreted, I'd be surprised if it was intentional.
But, I know I know. Always assume the worst yada yada. Microsoft is evil, right?
I regretted mentioning realism the moment I clicked submit, because I knew some people would latch onto that word and forget everything else I said.
I only meant that, because arcade-style games are much easier to make good looking than realistic games, Quake 3 can easily "look" like a relevant benchmark despite not properly using a modern GPU.
Modern engines people develop new games with, like Unity or Unreal 3, work very different from Quake 3 even if you're not going for realism. These would make a much more relevant benchmark.
(And I still play Quakeworld Team Fortress to this day, so don't lump me in with the graphics-are-everything crowd. Thank you.)
Slashdot Mirror
It's not like devs are putting their code under BSD with the expectation that companies will share back, only to be in total shock and feeling scorned when a company doesn't. If they cared about that issue as much as the loud people on /. did, they'd be using GPL.
/. can keep waging the license war, while devs will continue to ignore it and license their code in whatever way they're comfortable with.
Clang aims for more integration than "compile & debug". IDEs can directly use Clang libraries for syntax highlighting and error diagnostics. It was my understanding that GCC lacked this functionality and an old, large, and complex code base made it unrealistic for such a large feature addition to be worked on by mere mortals.
I'm sure not having to work with GPL is a definite plus for Apple, but this is the first time I've heard of licensing being a big reason for Clang's existence.
Have you tried x264 with --preset veryfast? My experience is that x264 is able to match a GPU encoder's speed while still giving significantly better quality. I'd only bother with a GPU encoder if I had a terrible CPU (netbook, phone?).
I'm a working software developer and I have no degree, only a hell of a passion for coding and the ability to learn on my own.
It's possible. There are plenty of places who will be looking for cheap junior software devs. The work won't be terribly interesting, but they're excellent places to jump start a career.
A recruiting agency like Volt is an easy place to start. You don't pay them a dime. I recommend grabbing some certifications – with a lack of experience or degree, anything can help get your foot in the door to an in-person interview. Even ones from Brainbench will help, and they're pretty cheap.
It's not the music that matters so much as the mixer/DAC.
High bit-depth playback matters greatly for any system that controls volume at the mixer stage rather than the amp stage. This is very important for PCs, where it is common to keep the amp (speakers) at a fixed volume and control the actual listening volume from the operating system's mixer.
If you keep the volume all the way up on your mixer, controlling your listening only at the amp stage, then a 16-bit pipeline is plenty. Highly integrated hardware such as an MP3 player can easily get away with this.
Otherwise, a 24-bit mixer and DAC can be very useful as it allows you to turn the volume down really far on the mixer while still retaining all the detail of your 16-bit music for the amp to boost back to listenable levels. It's still not perfect (such a low line level will inevitably be noisier) but it's still much better than a 16-bit pipeline.
The only problem with e-books and e-readers is that they're clearly not made by readers.
Books, the good ones at least and most of the bad ones too, pay attention to typography. Paragraph-optimized justification, hyphenation, hanging punctuation, ligatures, etc. All these little things that you take for granted with a dead-tree book, but without them it's a significantly poorer experience.
You find books with left-aligned text, an ugly and jagged right edge carving out a large chunk of empty space on the right. Or worse, you get one that is justified. This is bottom-of-the-barrel justification, without hyphenation and very commonly leaving huge spaces between words.
I've owned a Nook since launch day. I've read a large number of books on it, and I love it. But there is still a lot of room for improvement. I shouldn't need to import my ebooks into Adobe InDesign to make a PDF with proper typography.
They can put whatever positive, law-cooperating spin on it that they want, but the real reason is simply that Rapidshare no longer has any competition to drive up their free download speeds.
Their single largest competitor (Megaupload) just dropped off the map and all their other competitors have either blocked the US or dropped any free support whatsoever.
"We also show that as far as true performance is based - real-world performance as seen by the end-user- 2.4 is as fast, and even faster than some of the servers who may be "better" known as being "fast", like nginx," Jagielski said.
What's with the quotes? Other servers have proven to be faster, lighter weight, and more scalable than Apache for a long time. Don't be bitter because you fell behind. Be happy that you're finally catching up.
They mean to replace multiple per-thread copies of data with one single copy of it, accessible by all threads. No doubt part of Mozilla's latest push to reduce memory consumption.
One feature of x86 is that, save for a specialized SSE streaming store instruction, any store made by one core is immediately visible to all the other cores—even when the old value was already in a core's cache.
Maintaining such cache coherency involves a lot of overhead, so to get better scalability multi-threaded apps will sometimes adopt a "share-nothing" model: all threads get their own copy of the data, and no other threads will ever get to touch it. You trade memory and complexity for speed.
It sounds like Mozilla has decided this trade off is no longer worth it, and so has done away with multi-threading all together. Perhaps they will use green threads instead of native threads, though that brings along its own bag of complexities.
HDCP supported DVI before it supported HDMI, and has been available on graphics cards for years. This won't be closing any holes.
Windows 8 Metro apps will use C++, C#, or HTML5/JavaScript. I suspect HTML5 will get the largest use, though I'm sure C# will be a close second.
If you've got a Samsung phone, head on over to WindowBreak. It'll give you developer access and native execution abilities, even starting from a locked-down 7.5 (Mango).
It certainly stresses the video card, but I don't know anyone who acquired it because of this.
Personally I think Crysis and Crysis 2 were quite fun and even replayable. The story is forgettable, but the gameplay is great. A while after beating it playing full stealth, I revisited Crysis and attempted to rambo using a mix of all the suit abilities -- speed in, leap into the air, and knock some enemies out with all your strength.
Lots of my friends have now taken to waiting for GOTY editions to release so they can get the full game and all expansions/DLC for one sane price. People with this attitude may account for less initial sales, helping piracy appear disproportionate.
Still, 3.9 million is quite an impressive number. It will no doubt fuel Ubisoft's want for even more retarded DRM.
These days, I could. Because the 80s and 90s were something of a fluke in which hardware was progressing at a rapid rate, it coincided with the growth of the video game industry and attracted a lot of hardware geeks. But that era is gone, and hardware has stabilized to the point where new games are coming out targeting five year old hardware, and most people are okay with it. Skyrim runs on my first-generation Intel iMac from 2006.
For better or worse, there's a reason PC games have no issues on five year old hardware, and it's not because hardware has slowed down.
99% of PC games released now are ports from the consoles. Usually the only extras you get on PC are the trivially easy ones: support for higher resolutions, higher fps, and larger textures. If games didn't need to worry about running on ancient console hardware, they'd be able to look a lot more stunning on the PC.
There's also the problem of Windows' graphics stack introducing a huge amount of overhead to keep things safe, whereas consoles give more direct access to the hardware. Carmack guessed that GPUs in PCs are at least 10x more powerful than the ones in consoles, but are severely limited in how much new data you can give them every frame due to this overhead.
I've been a long-time DD-WRT user, but its development seemed to stagnate. I recently put TomatoUSB on my Linksys WRT160N v1, and it is working wonderfully. The interface is much nicer, and exposes more QoS and bandwidth management features which I've found useful. Check out the TomatoUSB website for a list of routers it supports.
Take a short miniseries and stretch it out to a full series to milk it and make the plot run like molasses. Add in a lot of mindless action with big explosions and helicopter chases, because that's what American shows look like, right?
I'm afraid of what they're going to do to Doctor Who, but if Torchwood was any example, keep Hollywood's dirty hands off it.
Typically when doing this kind of statistical analysis, one uses half the data for training and half the data for accuracy tests.
I haven't RTFA though, so I don't know what they've done.
WinRT doesn't replace .NET or the Win32 API, nor is it a second-class citizen when used from C#.
Metro apps are intended to be small—about the size of a cell phone app. They aren't made to replace full desktop apps meant to get work done. The .NET APIs you get for Metro apps have been scrubbed, getting rid of "insecure" APIs as well as ones that WinRT made redundant. A lot of the .NET API you've become used to is still available, though.
It uses COM behind the scenes, but there's no hint of COM in the API -- in fact, people in the Mono project have already begun work on a purely managed WinRT implementation. It definitely doesn't feel like a second-class citizen while using it. I'll be curious to see how much overhead it introduces.
WinRT in C++ is a little weird. There are two ways to use it: WinRTL, which is written in standard C++ but has all the verbosity of COM, or C++/CX, some extensions to C++ that make using it a breeze if you don't care about portability. It doesn't replace the Win32 API either. They encourage you to use WinRT when possible, but the entire Win32 API is still available. It's not clear if they'll be filtering the APIs you're allowed to use during the review process.
The biggest problem with Metro apps is that you can't just distribute an exe—all distribution needs to go through the Windows app store, complete with $100/year fees and review process. I'm sure a hack will come out, but this might make life a bit more difficult for many hobby developers trying to reach the common user.
Carmack touched on this during his QuakeCon keynote, saying it is possible but not currently planned.
I don't have any 55-gallon trash bags in my house, nor the helium it would take to fill even one of them.
Cool high-school project, though. Or should that be "high school-project".
Not quite. Seagate's tech is a simple block cache, where the most frequently accessed blocks get thrown on the SSD portion. It is no doubt quite effective, but we should be able to do better if we move the logic into the OS.
The OS has intimate knowledge of the filesystem, and can easily profile its use. Files that are small or randomly accessed should get put on the SSD, while large sequentially accessed files should get put on the HDD.
(13) Has proper typography—kerning, ligatures, hanging puncutation, and paragraph-optimized justification.
(14) Is easier to read having a higher DPI, better contrast ratio, and less reflectivity.
(15) Has cheaper books.
(16) Can't be taken away from you at the flip of a switch.
(17) Requires a trip to a store, or a lengthy ship time.
(18) Can only contain a limited number of pages.
I own a Nook and love it, but am waiting very excitedly for this tech to evolve.
The Raspberry Pi isn't exactly the same thing -- it does not include a case, keyboard, LCD, or speakers. But, you can probably get all that stuff for another $25. So maybe the OLPC has a new partner.
I've had a Windows Phone 7 device since day one, and it asks at every turn before doing something that would collect location (or any other) data. If it's true and not just a misconfigured device or data being poorly interpreted, I'd be surprised if it was intentional.
But, I know I know. Always assume the worst yada yada. Microsoft is evil, right?
I regretted mentioning realism the moment I clicked submit, because I knew some people would latch onto that word and forget everything else I said.
I only meant that, because arcade-style games are much easier to make good looking than realistic games, Quake 3 can easily "look" like a relevant benchmark despite not properly using a modern GPU.
Modern engines people develop new games with, like Unity or Unreal 3, work very different from Quake 3 even if you're not going for realism. These would make a much more relevant benchmark.
(And I still play Quakeworld Team Fortress to this day, so don't lump me in with the graphics-are-everything crowd. Thank you.)