and democrats wanting to "help me" by spending (my?) money
And the Bush Administration plus the Republican-controlled Congress that we had from 2001-2007 proved that the Republicans just want to "help you" too, but in an even *bigger* way. [Sigh]
The sad fact is that fiscal conservatives are no longer in control of the Republican party, nor are they even in a position of influence, and since they've never held much influence in the Democratic party, all I can say is "God help us all".
Im still not sure how ATIs driver base stands with linux, could someone fill me in?
AMD now owns ATI, and is opening up their graphics specs in a big way, as well as improving the Linux drivers for their cards. They now release Linux & Windows drivers at the same time, as they've put the development of the Linux drivers on equal footing with the Windows drivers (moving to a unified development process), so the Linux drivers in the last few revs have seen major improvements as they get them caught up to the same feature set of their Windows counterparts (the last rev added CrossfireX and Overclocking support, for example).
Also, they have already released several specs and docs for their older chipsets (RV500), and have promised much more. How far they can go wrt 3D accel is an open question as that will depend on how much 3rd party tech they used in the driver code (this of course is the reason/excuse NV gives for not opening up their drivers), but AMD has a *much* better track record of openness than ATI ever did, so I'm more hopeful.
I still feel the Nvidia is a better card, and is better built
If we're comparing a modern NV card to a modern ATI card, then I'd have to disagree with you that NV is the "better" card.:)
When I add in things like Linux support (NV getting worse on this; ATI getting much better, rapidly), heat generation/efficiency (NV worse; AMD/ATI, with its RV7xx and IGP cores, good and getting better), and cost (I'll let you guess this one!), then AMD/ATI is starting to look really good here, especially their IGP products.
In particular, the combination of an AMD 790GX Northbridge (builtin RadeonHD 3300) + SB750 Southbridge is what will be at the core of my next machine, and it will be my first ATI graphics system ever. Before the AMD takeover, I never even considered them as a possibility due to their abysmal Linux support.
Of course, if you have to have the fastest thing there is, and money is no object, then the high-end NV cards are still the fastest 3D cards there are (however this now even depends on which benchmarks you use), but for the rest of us the newest AMD/ATI cards are so close in performance, but at much less cost ($200+ cheaper for equivalent performance), that its becoming almost a no-brainer decision for the budget conscious. This fact is why NV has been cutting the prices on their high end cards, and why every market needs at least 2 viable competitors, because with competition *every* consumer wins. Even if your next card is from NV, it'll be cheaper for you... thanks to AMD/ATI.
I wonder why Thomson is demanding a jury trial in a technical case like this.
For the same reason that SCO asked for the same thing in their lawsuit against IBM. If your case is *legally* strong you almost always want the judge to decide it directly.
(In the SCO case fortunately, the judge realized their case wasn't just weak, it was *non-existent*, so a jury trial was refused.)
Empirical observation indicates that the opposite is occurring
I really, *really* want to do disagree with this in some substantial way... but can't think of a counter-response. Empirical observation trumps hopeful human optimism, yet again. Damned depressing, that.
The main threats our ancesters faced were natural and largely beyond their control. Now the main threat we face is each other, i.e. the other 7 billion humans on this planet, each with their own superstitions.
Our rapid success as a species is turning this asset into a liability.
So the real question is: can it *devolve* quickly enough before it ends up destroying us?
That's only true for fairly limited definitions of nobody. Otherwise Microsoft would be a hell of a lot smaller than they are.
If MS had to compete on a level playing field with many other vendors, all producing a compatible version of a commodity OS (with that compatibility allowing users to relatively easily switch between vendors), then yes they would be *much* smaller now... probably nonexistent.
RH doesn't have the luxury of controlling 90% of its market.
It's a creative commons attribution-noncommercial-noderivatives license
That's for the contents of the project's website, not the project's source code!
They're in full control.
Not exactly. From Chromium's FAQ:
Q. What license is the source released under? A. Chrome is a collection of a lot of software with a variety of licenses, but the Google-contributed code is BSD. For more info, see the full breakdown.
I'm having difficulty understanding why open-source implies they won't have the power to control the direction in which it's taken. The best FOSS is driven by a clear roadmap which tries to prevent forking (sorry, don't have citation for this). I think google can and will drive this in their direction,
Anyone can control the direction of their project, it all depends on which direction they take and how many others are interested in the same code.
the FOSS community can resist but they have no leverage against this mammoth.
If Google took a very controversial direction and a lot of other coders became interested in the code (look at how many people are involved with Mozilla), the FOSS community could fork the project, or at least maintain a variant of Google's codebase (a large patch on top of their code).
Google's history *suggests* they will not do something that is antagonistic to the FOSS community, more likely, they will want to develop the thing in a cooperative manner. This allows the project to improve at a much faster rate (more hackers, more eyeballs on the code) without Google having to pay a large number of devs to work on it full time.
There have been several examples of situations kind of like this. See the history of WebKit itself, for example. Apple started it by forking an early KHTML from the KDE project and working on it internally. After much friction though, Apple eventually opened it up and even allowed KDE devs write access to the WebKit code repository, allowing others to become 1st-class maintainers alongside the Apple employees who work on it as well.
As long as there is any level of pragmatic desire among all parties, eventually these kind of things stay open and get developed jointly.
You need to keep in mind a key difference between Google and MS: Google doesn't make its money by selling software, its money comes from its search engine and advertising. They simply don't need to make this browser a closed-source, tightly controlled app in order to improve their own bottom line (again, look at what Mozilla does for them).
You can call it a micro-notebook market if you don't like the acronyms.
The computer industry has waaayyy to many acronyms already..:)
AMD is still not a serious threat to anyone in the low power space.
I didn't say they were a threat *now*.
They will have to come out with a revolution to really make a difference to VIA.
and that revolution *might* have a name: "Fusion".:)
Note that the Bobcat version of this processor is the one targeted for UMPC/OLPC systems ("Swift" is for the notebooks).
AMD also needs to get their financial house in order before we can take the possibility of new R&D from them being applied to the market in any significant way.
I think you under-estimate their current fiscal situation. Their R&D never stopped just because Intel retook the lead in the high-end performance market (word is the Japanese fab maker TSMC has already been sampling early Fusion processors - that is who will produce them for AMD, at least initially). AMD has been in a non-stop fight with the market's 800lb gorilla since they released their 286 clone, so they've always been behind the 8-ball, never having more than 25% or so of the market, because they've *never* had the fab capacity to match Intel, yet while still behind that 8-ball they came up with AMD64, something Intel thought couldn't be done.
I'm not suggesting that Fusion is their next home-run, I don't know, what I do know is that they've got a track record of innovation and getting things done even while operating from a weak market position, so they shouldn't be automatically dismissed as already "dead". And they've already demonstrated the best *integrated* graphics chipset (for desktops) on the market so far, with their 790GX (blows anything from Intel out of the water - even beats most of NVIDIA's IG stuff in performance, while being vastly more energy efficient), so they have the GPU tech to make it happen (what will be in the Fusion is a low-power variant of their ATI tech), the question is whether they've succeeded in integrating the CPU and GPU in such a way to achieve a huge power efficiency win.
When you add to this Intel's current problem with power consumption of its own mobo chipsets, and its very poor integrated graphics, AMD certainly has an opening. And if AMD meets their own targets for energy efficiency then VIA will certainly face the brunt of the pressure created by AMD's success since VIA's offerings here are the weakest, but to a lesser extent, so will Intel, since Intel's own performance here (when including graphics capabilities) is not exactly stellar either.
Now, while at the very bottom of the market VIA and Intel's lesser offerings may continue to hold sway, but on the other hand the entire micro-notebook market is pushing for higher graphics performance, and thats what AMD may have right now that neither VIA nor Intel can match.
once they're included in the Java VM it opens doors for developers
Most developers are now either writing to standards, or they're writing IE specific websites because IE controls 90% of the market. Google has no leverage at this point to force developers to do any different. If/When IE's marketshare has been whittled down to 50-60% of the market, *then* something like this might be possible.
all these new approaches (multi-threaded/processed, new VM, new garbage collector, implementing levels of permissions,...) points towards the start of something
Sure, they point to a profoundly better way to make a standards-compliant web browser!
You still have one problem on this issue though: this browser will be open-source. Google will not have the power that Microsoft has traditionally had, because they'll have no ability to lock anyone in, or force anyone to use their browser via bundling. And as long as this browser has just a tiny fraction of the market, all developers will just ignore any nice "goodies" that Google tries to throw in.
Check that comic strip, somewhere they say they aren't "aiming for bells and whistles just rock solid stability/performance" (paraphrasing). I'm betting they'll let the FOSS community, later on, take care of the bells and whistles, just as with the extensions feature of Firefox, but right now Google is just interested in getting the core functionality working, and working faster and better than all current traditional browsers do (which by itself should scare the hell out of MS).
they had enough of IE and this is a declaration of war.
I agree, but the mistake you're making here is assuming that Google will fight this war the same way Microsoft would.
That would be a *really* neat trick for Google to pull off, considering this browser will be open-source and everyone, including Microsoft will be able to see the code...
Google dominates the search engine market like IE dominates the browser market, they know they're being watched by everyone, including the government, and I have seen no evidence so far that indicates Google is that stupid.
Aren't the V8 people also behind the Dalvik Java VM, which is used in Android?
I don't think so. If you compare what they say about V8 in the "comic strip" on page 13-17 to Dalvik's wiki page they sound very different. Dalvik is designed for low memory systems (no JIT!), whereas V8 is designed for full JVM compliance and speed (high performance garbage collection!).
Its hard to tell *where* "V8" comes from though. Does anyone know if Google has a code shop in Denmark, or is this "team in Denmark" a seperate entity? Is the "team in Denmark" *also* named 'V8' in addition to what they've created? The strip is a little misleading here...
FYI to all: You really should read the "comic strip", there is a lot more information there than its method of presentation would normally imply.:)
Yea, my fault, should have separated JS from Sun & Java in my comment, JS has always been an open standard.
Sun did however have *something* to do with it: they let Netscape use "Java" in its name (before this it was called "LiveScript").:) Sun still holds the trademark on that name.
WebKit is dual licensed under LGPL and BSDL, so Google can use just about any license they wish, probably BSDL, same as their Gears stuff, but because of the additional LGPL, there will be no "problem" with the FSF and FOSS community.
However, since Mozilla is also under the LGPL, if Google chooses to use the LGPL for the project they could incorporate code from Mozilla if they wanted any...
True, but that doesn't really matter either. The difference now is that both Java & Javascript have been open-sourced by Sun.
Note to GP: if you RTFA, you'll see that the new "V8 VM" is an already existing open source project (just like WebKit) from some European developers. Google isn't reinventing all the wheels, they're leveraging existing FOSS toolkits and technologies. This is Google we're talking about: they aren't stupid.
Some websites will exploit the full capabilities of the new V8 VM
The "V8 VM" is just a high performance implementation of the Javascript VM which is already an open industry standard (and FOSS software for that matter), and since IE controls 90% of the browser market, the only browser maker that can still get away with breaking standards is Microsoft, not Google.
Or perhaps Google is thinking long term, and this project is still early in its development, in which case Flash will have been upstaged by Silverlight by the time this browser is ready for prime time. By then there may other stable solutions available for Flash anyway, like swfdec or gnash.
However, IIRC Adobe hasn't opened up their Flash plugin code, only the.swf *format*. If they had open-sourced their Flash plugin code, there wouldn't be a need or a reason for either the swfdec or gnash projects (and we'd have relatively stable Flash plugins for all architectures and OSes by now).
So when it comes to openness and support for standards, Google still beats Adobe so bad its basically a mugging.
VIA's cpus can't really compete with normal consumer desktops. Intel's integrated graphics and low power cpus are much more capable
I'm just speculating here, but I don't think Intel is VIA's only concern. Intel is everyone's concern because they're the 800lb gorilla of the market, but in terms of the low-power/notebook market which is VIA's bread-n-butter, there is I think another threat developing. I would bet VIA is even more worried about what AMD is now up to.
AMD has bought ATI and is integrating their GPU technology into their own platform. With ATI being part of AMD, AMD's more open-source friendly philosophy is now taking root, and we're seeing a profound opening up of ATI's GPU technology, and a leveling of the playing field, with AMD promising (and already starting to deliver) equal consideration to the non-Windows world with updates to their GPU drivers eventually occuring simultaneously for both Windows and Linux, and supporting the same features. They've still got some catching up to do because old ATI's Linux support was a bad running joke in the Linux community before AMD bought them, but since then we've had several doc releases, rapid improvements to AMD/ATI's drivers, and features previously only available to Windows users now showing up in the Linux drivers (like Crossfire & overclocking support), and a commitment from AMD to go much further (and given AMD's track record, there is no reason not to believe them at this point).
Add to this AMD's push for efficiency, which shows up most clearly in their upcoming "Fusion" microprocessor (sometime in 2009), which merges the CPU and GPU within the same chip/package, but not (yet?) on the same silicon/die, similar to what Intel is doing with its "Nehalem" architecture. However, AMD's Fusion may be more of a concern for VIA than Intel because AMD is more focused on energy efficiency than either Intel is with its CPUs or NVIDIA is with its GPUs, and AMD's ATI-based integrated graphics capability is far more powerful than anything Intel or VIA currently has, look up the talk about AMD's latest 790GX northbridge + SB750 southbridge chipset combination for motherboards for example (strong integrated graphics performance with excellent overall power efficiency). Of particular concern to VIA, I suspect, is that the 1st implementation of this new Fusion microprocessor (aka "Swift" - see above link) will be for the notebook market, with the likely 2nd implementation of Fusion being the "Bobcat", a low-voltage, low-power processor aimed at the consumer electronics market, which is where VIA lives.
Not only will this thing be low-power and efficient, but with a CPU based on AMD's K10 architecture and a GPU core based on ATI's R7xx technology, we are talking about a CPU/GPU combo that is far more powerful than VIA's C7 CPU, which is based on old Pentium M technology, and VIA's current integrated graphics, which is based on the old S3 Graphics technology.
If this thing turns out to be all that it is claimed to be with respect to its power usage, then VIA could well be in serious trouble at this point, and since Linux is more visible in the low-power market, for VIA to remain closed, it will only hurt them even more as they start to get pressure not only from Intel but especially from AMD's Fusion systems, given that AMD is already relatively more open and Linux-friendly (arguably even more open than Intel).
VIA basically has no choice at this point if it wants to survive the lean times ahead.
It seems to me you've completely missed the point: he's getting complaints & bug reports from folks who think the problem is with standard Perl, when the problem is with RHEL's patched version of Perl. Given that he didn't create this problem, but is getting flack about it nonetheless, I can understand his frustration.
The reason it doesn't work the other way is because the GPL's additional restrictions are incompatible with BSD license since the latter says nothing about derivatives/modifications of the software.
Sigh, I didn't say that right.
GPL'd code can't be used in a 3 clause BSDL project because the BSDL doesn't include all of the restrictions that the GPL has, thus trying to do so would violate the *GPL*.
However BSDL'd code can be incorporated into a GPL'd project without violating either license as long as you abide by the 3 clauses of the BSD: attribution, don't use author for endorsement, and the liability disclaimer. None of those requirements are incompatible with the GPL.
The BSD license has a advertising clause which is fully incompatible with the GPL. There is also a "simplified BSD" license approved by the OSI that is used by projects like FreeBSD which works similar to MIT licensing.
Right, I was thinking of the latter, the 3-clause version without the advert clause. BSD with that clause is incompatible with the GPL.
Additionally because both licenses require that their BSD-centric copyright notice be reproduced with every copy, this adds additional restrictions
No, the BSD has no "extra" restrictions, thats what makes it "more permissive" than the GPL. See GNU's license list, where the 3 clause BSD is listed as the "modified BSD".
preventing the code from mixing with GPL
As long as you abide by the BSD license (which isn't hard since it *is* so permissive), you can always include it (including a copy of the BSD license itself + plus the author's copyright notice) into a GPL'd project and use a statement like "portions of this code are copyright...".
The reason it doesn't work the other way is because the GPL's additional restrictions are incompatible with BSD license since the latter says nothing about derivatives/modifications of the software. This is why GNU lists this license as "compatible" in their list above, it can be used with GPL'd code without either license being violated.
It's patently not about "remaining free". It is about "making free"
That's patently bullshit, because I am not forcing you to use my code. Its my code that I want to remain free, so if you don't want to release your own code that uses mine, then simply do not use my code.
What I cannot appreciate are the GPL proponents who freely take from BSD sources and give nothing back.
If you don't want your code to be used in a GPL'd project, then don't release it with a license that is more permissive than the GPL. You can't have it both ways.
Slashdot Mirror
and democrats wanting to "help me" by spending (my?) money
And the Bush Administration plus the Republican-controlled Congress that we had from 2001-2007 proved that the Republicans just want to "help you" too, but in an even *bigger* way. [Sigh]
The sad fact is that fiscal conservatives are no longer in control of the Republican party, nor are they even in a position of influence, and since they've never held much influence in the Democratic party, all I can say is "God help us all".
We're at the point where a decent video card is cheaper than the operating system
Really? Where can I get my hands on a *free* video card, pray tell!
Note: If you haven't gotten it by this point, then maybe my response should have been:
I run Linux, you insensitive clod!
Im still not sure how ATIs driver base stands with linux, could someone fill me in?
AMD now owns ATI, and is opening up their graphics specs in a big way, as well as improving the Linux drivers for their cards. They now release Linux & Windows drivers at the same time, as they've put the development of the Linux drivers on equal footing with the Windows drivers (moving to a unified development process), so the Linux drivers in the last few revs have seen major improvements as they get them caught up to the same feature set of their Windows counterparts (the last rev added CrossfireX and Overclocking support, for example).
Also, they have already released several specs and docs for their older chipsets (RV500), and have promised much more. How far they can go wrt 3D accel is an open question as that will depend on how much 3rd party tech they used in the driver code (this of course is the reason/excuse NV gives for not opening up their drivers), but AMD has a *much* better track record of openness than ATI ever did, so I'm more hopeful.
I still feel the Nvidia is a better card, and is better built
If we're comparing a modern NV card to a modern ATI card, then I'd have to disagree with you that NV is the "better" card. :)
When I add in things like Linux support (NV getting worse on this; ATI getting much better, rapidly), heat generation/efficiency (NV worse; AMD/ATI, with its RV7xx and IGP cores, good and getting better), and cost (I'll let you guess this one!), then AMD/ATI is starting to look really good here, especially their IGP products.
In particular, the combination of an AMD 790GX Northbridge (builtin RadeonHD 3300) + SB750 Southbridge is what will be at the core of my next machine, and it will be my first ATI graphics system ever. Before the AMD takeover, I never even considered them as a possibility due to their abysmal Linux support.
Of course, if you have to have the fastest thing there is, and money is no object, then the high-end NV cards are still the fastest 3D cards there are (however this now even depends on which benchmarks you use), but for the rest of us the newest AMD/ATI cards are so close in performance, but at much less cost ($200+ cheaper for equivalent performance), that its becoming almost a no-brainer decision for the budget conscious. This fact is why NV has been cutting the prices on their high end cards, and why every market needs at least 2 viable competitors, because with competition *every* consumer wins. Even if your next card is from NV, it'll be cheaper for you... thanks to AMD/ATI.
I wonder why Thomson is demanding a jury trial in a technical case like this.
For the same reason that SCO asked for the same thing in their lawsuit against IBM. If your case is *legally* strong you almost always want the judge to decide it directly.
(In the SCO case fortunately, the judge realized their case wasn't just weak, it was *non-existent*, so a jury trial was refused.)
their only hope is to confuse a jury?
Bingo.
Empirical observation indicates that the opposite is occurring
I really, *really* want to do disagree with this in some substantial way... but can't think of a counter-response. Empirical observation trumps hopeful human optimism, yet again. Damned depressing, that.
The main threats our ancesters faced were natural and largely beyond their control. Now the main threat we face is each other, i.e. the other 7 billion humans on this planet, each with their own superstitions.
Our rapid success as a species is turning this asset into a liability.
So the real question is: can it *devolve* quickly enough before it ends up destroying us?
Unless god(s) or the aliens actually shows up and confirms it of course. Even then I reckon there will be wide-ranging debate on the issue.
Unless the aliens are armed with BFG-9000s, or its equivalent, in which case the debate will be *very* short-lived...
That's only true for fairly limited definitions of nobody. Otherwise Microsoft would be a hell of a lot smaller than they are.
If MS had to compete on a level playing field with many other vendors, all producing a compatible version of a commodity OS (with that compatibility allowing users to relatively easily switch between vendors), then yes they would be *much* smaller now... probably nonexistent.
RH doesn't have the luxury of controlling 90% of its market.
It's a creative commons attribution-noncommercial-noderivatives license
That's for the contents of the project's website, not the project's source code!
They're in full control.
Not exactly. From Chromium's FAQ:
Q. What license is the source released under?
A. Chrome is a collection of a lot of software with a variety of licenses, but the Google-contributed code is BSD. For more info, see the full breakdown.
I'm having difficulty understanding why open-source implies they won't have the power to control the direction in which it's taken. The best FOSS is driven by a clear roadmap which tries to prevent forking (sorry, don't have citation for this). I think google can and will drive this in their direction,
Anyone can control the direction of their project, it all depends on which direction they take and how many others are interested in the same code.
the FOSS community can resist but they have no leverage against this mammoth.
If Google took a very controversial direction and a lot of other coders became interested in the code (look at how many people are involved with Mozilla), the FOSS community could fork the project, or at least maintain a variant of Google's codebase (a large patch on top of their code).
Google's history *suggests* they will not do something that is antagonistic to the FOSS community, more likely, they will want to develop the thing in a cooperative manner. This allows the project to improve at a much faster rate (more hackers, more eyeballs on the code) without Google having to pay a large number of devs to work on it full time.
There have been several examples of situations kind of like this. See the history of WebKit itself, for example. Apple started it by forking an early KHTML from the KDE project and working on it internally. After much friction though, Apple eventually opened it up and even allowed KDE devs write access to the WebKit code repository, allowing others to become 1st-class maintainers alongside the Apple employees who work on it as well.
As long as there is any level of pragmatic desire among all parties, eventually these kind of things stay open and get developed jointly.
You need to keep in mind a key difference between Google and MS: Google doesn't make its money by selling software, its money comes from its search engine and advertising. They simply don't need to make this browser a closed-source, tightly controlled app in order to improve their own bottom line (again, look at what Mozilla does for them).
You can call it a micro-notebook market if you don't like the acronyms.
The computer industry has waaayyy to many acronyms already.. :)
AMD is still not a serious threat to anyone in the low power space.
I didn't say they were a threat *now*.
They will have to come out with a revolution to really make a difference to VIA.
and that revolution *might* have a name: "Fusion". :)
Note that the Bobcat version of this processor is the one targeted for UMPC/OLPC systems ("Swift" is for the notebooks).
AMD also needs to get their financial house in order before we can take the possibility of new R&D from them being applied to the market in any significant way.
I think you under-estimate their current fiscal situation. Their R&D never stopped just because Intel retook the lead in the high-end performance market (word is the Japanese fab maker TSMC has already been sampling early Fusion processors - that is who will produce them for AMD, at least initially). AMD has been in a non-stop fight with the market's 800lb gorilla since they released their 286 clone, so they've always been behind the 8-ball, never having more than 25% or so of the market, because they've *never* had the fab capacity to match Intel, yet while still behind that 8-ball they came up with AMD64, something Intel thought couldn't be done.
I'm not suggesting that Fusion is their next home-run, I don't know, what I do know is that they've got a track record of innovation and getting things done even while operating from a weak market position, so they shouldn't be automatically dismissed as already "dead". And they've already demonstrated the best *integrated* graphics chipset (for desktops) on the market so far, with their 790GX (blows anything from Intel out of the water - even beats most of NVIDIA's IG stuff in performance, while being vastly more energy efficient), so they have the GPU tech to make it happen (what will be in the Fusion is a low-power variant of their ATI tech), the question is whether they've succeeded in integrating the CPU and GPU in such a way to achieve a huge power efficiency win.
When you add to this Intel's current problem with power consumption of its own mobo chipsets, and its very poor integrated graphics, AMD certainly has an opening. And if AMD meets their own targets for energy efficiency then VIA will certainly face the brunt of the pressure created by AMD's success since VIA's offerings here are the weakest, but to a lesser extent, so will Intel, since Intel's own performance here (when including graphics capabilities) is not exactly stellar either.
Now, while at the very bottom of the market VIA and Intel's lesser offerings may continue to hold sway, but on the other hand the entire micro-notebook market is pushing for higher graphics performance, and thats what AMD may have right now that neither VIA nor Intel can match.
We'll know in a couple of months either way...
once they're included in the Java VM it opens doors for developers
Most developers are now either writing to standards, or they're writing IE specific websites because IE controls 90% of the market. Google has no leverage at this point to force developers to do any different. If/When IE's marketshare has been whittled down to 50-60% of the market, *then* something like this might be possible.
all these new approaches (multi-threaded/processed, new VM, new garbage collector, implementing levels of permissions, ...) points towards the start of something
Sure, they point to a profoundly better way to make a standards-compliant web browser!
You still have one problem on this issue though: this browser will be open-source. Google will not have the power that Microsoft has traditionally had, because they'll have no ability to lock anyone in, or force anyone to use their browser via bundling. And as long as this browser has just a tiny fraction of the market, all developers will just ignore any nice "goodies" that Google tries to throw in.
Check that comic strip, somewhere they say they aren't "aiming for bells and whistles just rock solid stability/performance" (paraphrasing). I'm betting they'll let the FOSS community, later on, take care of the bells and whistles, just as with the extensions feature of Firefox, but right now Google is just interested in getting the core functionality working, and working faster and better than all current traditional browsers do (which by itself should scare the hell out of MS).
they had enough of IE and this is a declaration of war.
I agree, but the mistake you're making here is assuming that Google will fight this war the same way Microsoft would.
So Google will work faster on Google browser
That would be a *really* neat trick for Google to pull off, considering this browser will be open-source and everyone, including Microsoft will be able to see the code...
Google dominates the search engine market like IE dominates the browser market, they know they're being watched by everyone, including the government, and I have seen no evidence so far that indicates Google is that stupid.
Aren't the V8 people also behind the Dalvik Java VM, which is used in Android?
I don't think so. If you compare what they say about V8 in the "comic strip" on page 13-17 to Dalvik's wiki page they sound very different. Dalvik is designed for low memory systems (no JIT!), whereas V8 is designed for full JVM compliance and speed (high performance garbage collection!).
Its hard to tell *where* "V8" comes from though. Does anyone know if Google has a code shop in Denmark, or is this "team in Denmark" a seperate entity? Is the "team in Denmark" *also* named 'V8' in addition to what they've created? The strip is a little misleading here...
FYI to all: You really should read the "comic strip", there is a lot more information there than its method of presentation would normally imply. :)
the name of the VM doesn't exactly send the signal that it will be the fastest one around?
Hey, I'm a RoadWarrior driving one of the last of the V8s, you insensitive clod!
Yea, my fault, should have separated JS from Sun & Java in my comment, JS has always been an open standard.
Sun did however have *something* to do with it: they let Netscape use "Java" in its name (before this it was called "LiveScript"). :) Sun still holds the trademark on that name.
http://en.wikipedia.org/wiki/JavaScript_engine
WebKit is dual licensed under LGPL and BSDL, so Google can use just about any license they wish, probably BSDL, same as their Gears stuff, but because of the additional LGPL, there will be no "problem" with the FSF and FOSS community.
However, since Mozilla is also under the LGPL, if Google chooses to use the LGPL for the project they could incorporate code from Mozilla if they wanted any...
JavaScript != Java
True, but that doesn't really matter either. The difference now is that both Java & Javascript have been open-sourced by Sun.
Note to GP: if you RTFA, you'll see that the new "V8 VM" is an already existing open source project (just like WebKit) from some European developers. Google isn't reinventing all the wheels, they're leveraging existing FOSS toolkits and technologies. This is Google we're talking about: they aren't stupid.
Some websites will exploit the full capabilities of the new V8 VM
The "V8 VM" is just a high performance implementation of the Javascript VM which is already an open industry standard (and FOSS software for that matter), and since IE controls 90% of the browser market, the only browser maker that can still get away with breaking standards is Microsoft, not Google.
but I see it as google bullying the webstandards
[citation needed]
Or perhaps Google is thinking long term, and this project is still early in its development, in which case Flash will have been upstaged by Silverlight by the time this browser is ready for prime time. By then there may other stable solutions available for Flash anyway, like swfdec or gnash.
However, IIRC Adobe hasn't opened up their Flash plugin code, only the .swf *format*. If they had open-sourced their Flash plugin code, there wouldn't be a need or a reason for either the swfdec or gnash projects (and we'd have relatively stable Flash plugins for all architectures and OSes by now).
So when it comes to openness and support for standards, Google still beats Adobe so bad its basically a mugging.
VIA's cpus can't really compete with normal consumer desktops. Intel's integrated graphics and low power cpus are much more capable
I'm just speculating here, but I don't think Intel is VIA's only concern. Intel is everyone's concern because they're the 800lb gorilla of the market, but in terms of the low-power/notebook market which is VIA's bread-n-butter, there is I think another threat developing. I would bet VIA is even more worried about what AMD is now up to.
AMD has bought ATI and is integrating their GPU technology into their own platform. With ATI being part of AMD, AMD's more open-source friendly philosophy is now taking root, and we're seeing a profound opening up of ATI's GPU technology, and a leveling of the playing field, with AMD promising (and already starting to deliver) equal consideration to the non-Windows world with updates to their GPU drivers eventually occuring simultaneously for both Windows and Linux, and supporting the same features. They've still got some catching up to do because old ATI's Linux support was a bad running joke in the Linux community before AMD bought them, but since then we've had several doc releases, rapid improvements to AMD/ATI's drivers, and features previously only available to Windows users now showing up in the Linux drivers (like Crossfire & overclocking support), and a commitment from AMD to go much further (and given AMD's track record, there is no reason not to believe them at this point).
Add to this AMD's push for efficiency, which shows up most clearly in their upcoming "Fusion" microprocessor (sometime in 2009), which merges the CPU and GPU within the same chip/package, but not (yet?) on the same silicon/die, similar to what Intel is doing with its "Nehalem" architecture. However, AMD's Fusion may be more of a concern for VIA than Intel because AMD is more focused on energy efficiency than either Intel is with its CPUs or NVIDIA is with its GPUs, and AMD's ATI-based integrated graphics capability is far more powerful than anything Intel or VIA currently has, look up the talk about AMD's latest 790GX northbridge + SB750 southbridge chipset combination for motherboards for example (strong integrated graphics performance with excellent overall power efficiency). Of particular concern to VIA, I suspect, is that the 1st implementation of this new Fusion microprocessor (aka "Swift" - see above link) will be for the notebook market, with the likely 2nd implementation of Fusion being the "Bobcat", a low-voltage, low-power processor aimed at the consumer electronics market, which is where VIA lives.
Not only will this thing be low-power and efficient, but with a CPU based on AMD's K10 architecture and a GPU core based on ATI's R7xx technology, we are talking about a CPU/GPU combo that is far more powerful than VIA's C7 CPU, which is based on old Pentium M technology, and VIA's current integrated graphics, which is based on the old S3 Graphics technology.
If this thing turns out to be all that it is claimed to be with respect to its power usage, then VIA could well be in serious trouble at this point, and since Linux is more visible in the low-power market, for VIA to remain closed, it will only hurt them even more as they start to get pressure not only from Intel but especially from AMD's Fusion systems, given that AMD is already relatively more open and Linux-friendly (arguably even more open than Intel).
VIA basically has no choice at this point if it wants to survive the lean times ahead.
It seems to me you've completely missed the point: he's getting complaints & bug reports from folks who think the problem is with standard Perl, when the problem is with RHEL's patched version of Perl. Given that he didn't create this problem, but is getting flack about it nonetheless, I can understand his frustration.
The reason it doesn't work the other way is because the GPL's additional restrictions are incompatible with BSD license since the latter says nothing about derivatives/modifications of the software.
Sigh, I didn't say that right.
GPL'd code can't be used in a 3 clause BSDL project because the BSDL doesn't include all of the restrictions that the GPL has, thus trying to do so would violate the *GPL*.
However BSDL'd code can be incorporated into a GPL'd project without violating either license as long as you abide by the 3 clauses of the BSD: attribution, don't use author for endorsement, and the liability disclaimer. None of those requirements are incompatible with the GPL.
The BSD license has a advertising clause which is fully incompatible with the GPL. There is also a "simplified BSD" license approved by the OSI that is used by projects like FreeBSD which works similar to MIT licensing.
Right, I was thinking of the latter, the 3-clause version without the advert clause. BSD with that clause is incompatible with the GPL.
Additionally because both licenses require that their BSD-centric copyright notice be reproduced with every copy, this adds additional restrictions
No, the BSD has no "extra" restrictions, thats what makes it "more permissive" than the GPL. See GNU's license list, where the 3 clause BSD is listed as the "modified BSD".
preventing the code from mixing with GPL
As long as you abide by the BSD license (which isn't hard since it *is* so permissive), you can always include it (including a copy of the BSD license itself + plus the author's copyright notice) into a GPL'd project and use a statement like "portions of this code are copyright ...".
The reason it doesn't work the other way is because the GPL's additional restrictions are incompatible with BSD license since the latter says nothing about derivatives/modifications of the software. This is why GNU lists this license as "compatible" in their list above, it can be used with GPL'd code without either license being violated.
It's patently not about "remaining free". It is about "making free"
That's patently bullshit, because I am not forcing you to use my code. Its my code that I want to remain free, so if you don't want to release your own code that uses mine, then simply do not use my code.
What I cannot appreciate are the GPL proponents who freely take from BSD sources and give nothing back.
If you don't want your code to be used in a GPL'd project, then don't release it with a license that is more permissive than the GPL. You can't have it both ways.