With ATL 3.0, MS introduced "import" (sort of like include). Also some special pseudo-template stuff to provide better smart-pointers was introduced. This was in VS 6.0.
I don't know, if the book uses this stuff, though.
Index operations are several orders of magnitude faster than procedure calls.
Actually a table based solution like bison is slower than implementing the parser directly in code (like recursive descent).
But giving up on LALR parsing is not necessary. You can use recursive ascent. See this article that claims a speedup of 2.5 to 6.5 compared to table based LALR(1).
The standard C++ grammar is highly ambiguous. So not even infinite lookahead can be used to parse it. The ambiguities require context-information to be resolved, and in many cases even given this context information (eg. is this identifier a type, variable or template) infinite lookahead is still required. Actually only very few problems are solved with finite (aka k) lookahead.
Also the standard C++ grammar is not free of left-recursion, so LL (or recursive descent) parsing is not well-suited for this grammar at all.
ANTLR uses infinite lookahead (he calls it predicates), and therefore in theory the parsers constructed using this are non-linear.
Either Mark Mitchell of CodeSourcery has transformed the grammar into some unrecognizable gob of rules, or he uses backtracking and infinite look-ahead. In the first case some serious restructuring of the syntax tree is necessary.
I have just finished a thesis on the problems of parsing C++, and I would not have chosen recursive descent. I would have chosen automatically generated recursive acent (LALR(1)) instead, with backtracking and dependency on context information (aka lexical feedback). But I hope the best for their approach.
I just hope that the parser can also be used for other purposes than the GCC-compiler, ie. that it is a separate module, because a decent, free C++ frontend is needed by many projects (editors, code analyzers).
The average user can get a techi friend or friend of a friend to install it. That is where most people get the cracked games from, anyway.
Another possibility is to go to an electronic store to get it done. One such store in by neighborhood has a sign in the window:"mod chips for PS1/PS2 installed". (It's a little, local non-chain, one person store)
Basically, It might be MS's strategy, but I don't think it will prevent 90% from modding their X-box. More like 30%.
These incompatibilities are part of the whole computer industry. People don't buy and install RAM for this reason:)
Swedish researchers recently found traces of triclosan in human breast milk. On that grounds I'm pretty sceptical about handsoap and deodorants containing it (triclosan, not breast milk - yet:) ).
True, the EU is banning hormone-pumped meat from the US (and getting penalized by the WTO). But when it comes to antibiotics they are no saints either. In my country Denmark, Europe we've had big discussions about this. Veterinarians are no longer allowed to feed an animal antibiotics unless it is sick, but some farmers don't care, they have their own supply of drugs and injects all their animals with antibiotics two weeks after birth as it makes them gain more weight - just like hormones.
Is this legal to do in the US ? Anybody know if it is illegal in all European contries ?
Multiply all work estimates by a factor of 4, then tell them it's 'iffy'. The problem is that they will not believe you. A realistic estimate sounds horrendous enough. If you multiply it by four, it simply will not be approved.
Ignore their spec, write your own as generic as possible. When they change it 1/2 way through, you bitch and moan, then write the gui that will give them what they want I actually tried this, except I didn't bitch and moan. Let me tell you, they were really impressed when I used the last month just pulling extensions out of the sleeves, and implementing things they were only dreaming about. I told them about the lazy programmer, and they just kept repeating: "not lazy! smart!".
You are right. A kernel with DRM and freely available source code, so that you can remove DRM from it doesn't make sense.
However, OS X being "Open Source", doesn't help here. The BSD license used does not prevent Apple from closing the source in the future. If Apple incorporated DRM into OS X.2, they could close the source as well and the "million of people" able to make that modification, would shrink to the few hackers who could disassemble the kernel. As Apple also control the hardware side of their product, A modded BIOS (or whatever) would probably also be needed.
If that scenario unfolds, you would have two choices:
go with Apple, assimilate and eat DRM
Avoid DRM and miss major hardware upgrades, stick to a forked kernel developed only by the OS community and slowly but surely turning legacy. There are lots of tricks that Apple can use if they want to make new stuff available only to the DRM customers.
When the pain becomes too much, most people will just go with DRM. Mind you, most businesses don't mind and don't care about DRM.
Excellent point. E.g. Some people are still using KDE 1.x (otherwise it wouldn't still be bundled with SuSE). I for one am using it until I get a faster machine. There are lots of flaws, but I seriously disbelieve anything will be fixed.
The reason people are appalled is that MS might be doing this to gain a profit. They might be open about though, and state for how long time they will fix their products. Preferably in advance, not just suddenly dropping support. Then people could compare the cost of subscription vs. one-time-purchase. I would be surprised (MS is pushing subscription rather hard) if you were not better of subscribing to the software, if you wanted critical updates.
Today people gamble on how long they could use an unsupported product, mostly thinking at the time of purchase that this will be forever. It would be interesting to measure the minimal cost of a MS installation based on the history of unfixed security bugs. I bet you could still run windows NT 4.0 SP x, and you would now have to upgrade from Office 97 to Office XP. How does that compare to the subscription model, now and in the future ?
A big problem with the Kyoto agreement is that the worst polluters are not included.
What pollutes more: destroying lakes, local agriculture, inhabitat etc. by dumping toxic waste or intense emission of CO2 ? The answer is blowing in the wind so to speak.
If you think toxines are the biggest pollution problem, I guess the US aren't that bad. And the countries you mentioned probably are. On the other hand, when it comes to CO2, the US is one of the major polluters. They emission 23% of the world total, and they account for only 5% of the worlds population. To my knowledge the main issue in Kyoto was CO2. The problem with CO2 emission is that it pollutes globally.
The US govt would likely show more support, or even sign such an agreement, if there weren't some apparent economic motivations behind the requirements.
First of, as I stated in an earlier post the US themselves put in the apparent economic motivations.
Secondly, Anyone agrees that there is some relationship between economic growth and the amount of CO2 emission. I think it would be fair if we let the developing countries pollute a little more than they do today, maybe just 10% of what the USA is doing today, and gain some economic growth. Russia had an excuse for not reducing to 1992 level, because in 1992 Russia was at an economic colapse. Well, unfortunately the US was not as lucky;)
The other industrial countries which signed the treaty were under the same requirements as the US, and it should cost them as much money. They signed it to prevent pollution. Nobody stood to be rich from the (volountary) donations.
As for China, I definitely think they should have signed, but who can expect anything from an arrogant, communist dictatorship. Would you only agree to sign a treaty, if Iraq did ?
Sorry for bashing the US. The reason, I think, many Europeans (like me) do it, is because we are disapointed with the US. The USA is the world's only superpower; they are the world leader militarily, monetarily and culturally (to many). We admire you. But why is it that in everything you do, you only consider your own interest. America, what happened to the New World Order, when you won the cold war ? Why can't you be the world's policeman like we envisioned ?
Each time the US administration does something selfish, shortsighted and damaging, I tell myself that only a majority of US citizens are to blame. A lot of Americans are probably intelligent, well-meaning, responsible people. There just aren't enough.
This wasn't my point. I was never talking about donation to the third world in general. I was speaking specifically of Kyoto, and I though I made that very clear.
You misunderstood, so I will try to elaborate:
The poster insinuated that third world countries backed Kyoto to make easy money. Originally there was NOTHING in the Kyoto treaty which implied that money should be given to anyone. All countries just had to cut back the pollution to 1990 level before 2008 (or something). USA could have agreed to this and they wouldn't have had to pay any money to the third world. But USA themselves changed this, so that it was possible to buy the right to pollute from other countries. If then buying the right to pollute, makes the seller of the right greedy, and therefore you will not sign the treaty, it's like saying:
"Let's make a deal: I will pay you to slap yourself. But I don't think it is a fair deal, because you are so greedy to take my money."
Now, there could be a number of other reasons for the US not to sign Kyoto. Personally I probably wouldn't agree with them, though.
On a side note: The large majority of households are not heated by electricity, but by wasted heat (hot water running in tubes, called "remote heating") from garbage/coal burning power plants. Some have actually expressed concern that if too big a share of the Danish energy comes from another source than these power plants, there will not be enough heat, or electricity would be wasted.
Well, as far as I remember USA was the one who was in favour of the part of the treaty that allowed a country to PAY developing countries for the right to pollute instead of them. Most countries thought that it would be best if you HAD to pollute less.
Please don't accuse the third world of being greedy on account of this.
As I see it, the GPL doesn't protect us from this , anyway.
If a market leader with a GPL'ed installer owns enough copyright, they could simply re-license the installer as closed source, including a major rewrite and all the customers would be forced to switch or enter the same path as you describe assuming the vendor is that evil/stupid.
What is preventing Red Hat from doing this, except for the fact that they have a history of being open source ? A (virtual) monopoly is dangerous everywhere, mainly because it can dictate how things should be (aka. defacto standards), and which projects should receive support. I like the fact that SuSE supports the KDE project. If there were only Red Hat, KDE would not have received any support. I don't blame Red Hat for this, they just have other interests. Some of which I benefit from, some of which I don't.
Besides, if the installer is such a big problem, then why is it that there are so many of them ? Even a handfull of which is GLP'ed. I think the installer is only a small part of which makes a distribution. What matters more is testing of packages and integration. Having the installer be closed source software is a nuissance, because you cannot fix it, though.
When an ISV want to write software he should use the LSB as a
reference, which should provide some compatibility. However he cannot
test against LSB, he has to test against a number of concrete
distributions. Bringing that number down is important, maybe not to
you, but to the ISV.
I am very pleased to hear that not only SuSE and UL but also Debian
and Mandrake are committed to it.
As usual a lot of people here cry wolf at the slightest sign of
distribution consolidation.
UL is not forcing anyone to do anything. It isn't even threatening to
stifle competition, as there are plenty still. A non-dominating number
of players decided that ISV are important, and to ensure that they get
ISV support they form a group and standardize some of their
binaries. I have no problem with the fact that Mandrake are not a
member of UL. Fine! More competition. I just hope UL is big enough to
push some standards, but not so big as to create de-facto standards
like Red Hat.
In addition they get the advantage of additional combined
testing. This is what I, as a user, consider extremely important. I
have yet to try installing Linux on a nontrivial box without
problems. There has always been something wrong, and if the next
version fixed it, there was sure to be another problem with that. I am
talking soundcard, 3D, scanner, printer, X in various resolutions,
monitor frequenzy, CD-burner, IDE-DMA, netscape plugins, java. I won't
even begin to mention upgrading versions. If anyone new to Linux read
this: Always re-install Linux from scratch! Never upgrade!
Debians notion of stable and unstable is great and I hope one of the
UL distributions will focus on being stable as opposed to the current
obsession with being current.
The difference between OS/2 and Linux/wine is that there are infinitely more applications for Linux. You don't use linux/wine to run only Windows apps. You run all the linux apps and then maybe a few windows apps.
In OS/2 there was only very basic stuff. So you basically needed windows apps for everything. And there was no signs of a change to the better.
Another difference is that the Linux apps are almost guaranteed to become better over time due to the GPL (and to some extent other opensource licenses). In the end a native linux application will catch up at leat to a level, where you don't need the features of the windows app.
Well well,
And I guess You are using Galeon, Mozilla or Konqueror. Wow, you are so cool. Unfortunately they don't display some IE-only sites pretty well. So you also have to ditch YOUR shitty browser.
Slashdot Mirror
With ATL 3.0, MS introduced "import" (sort of like include). Also some special pseudo-template stuff to provide better smart-pointers was introduced. This was in VS 6.0.
I don't know, if the book uses this stuff, though.
Index operations are several orders of magnitude faster than procedure calls.
Actually a table based solution like bison is slower than implementing the parser directly in code (like recursive descent).
But giving up on LALR parsing is not necessary. You can use recursive ascent. See this article that claims a speedup of 2.5 to 6.5 compared to table based LALR(1).
The standard C++ grammar is highly ambiguous. So not even infinite lookahead can be used to parse it. The ambiguities require context-information to be resolved, and in many cases even given this context information (eg. is this identifier a type, variable or template) infinite lookahead is still required. Actually only very few problems are solved with finite (aka k) lookahead.
Also the standard C++ grammar is not free of left-recursion, so LL (or recursive descent) parsing is not well-suited for this grammar at all.
ANTLR uses infinite lookahead (he calls it predicates), and therefore in theory the parsers constructed using this are non-linear.
Either Mark Mitchell of CodeSourcery has transformed the grammar into some unrecognizable gob of rules, or he uses backtracking and infinite look-ahead. In the first case some serious restructuring of the syntax tree is necessary.
I have just finished a thesis on the problems of parsing C++, and I would not have chosen recursive descent. I would have chosen automatically generated recursive acent (LALR(1)) instead, with backtracking and dependency on context information (aka lexical feedback). But I hope the best for their approach.
I just hope that the parser can also be used for other purposes than the GCC-compiler, ie. that it is a separate module, because a decent, free C++ frontend is needed by many projects (editors,
code analyzers).
The average user can get a techi friend or friend of a friend to install it. That is where most people get the cracked games from, anyway.
:"mod chips for PS1/PS2 installed". (It's a little, local non-chain, one person store)
:)
Another possibility is to go to an electronic store to get it done. One such store in by neighborhood has a sign in the window
Basically, It might be MS's strategy, but I don't think it will prevent 90% from modding their X-box. More like 30%.
These incompatibilities are part of the whole computer industry. People don't buy and install RAM for this reason
Swedish researchers recently found traces of triclosan in human breast milk. On that grounds I'm pretty sceptical about handsoap and deodorants containing it (triclosan, not breast milk - yet:) ).
True, the EU is banning hormone-pumped meat from the US (and getting penalized by the WTO). But when it comes to antibiotics they are no saints either. In my country Denmark, Europe we've had big discussions about this. Veterinarians are no longer allowed to feed an animal antibiotics unless it is sick, but some farmers don't care, they have their own supply of drugs and injects all their animals with antibiotics two weeks after birth as it makes them gain more weight - just like hormones.
Is this legal to do in the US ? Anybody know if it is illegal in all European contries ?
Multiply all work estimates by a factor of 4, then tell them it's 'iffy'.
The problem is that they will not believe you. A realistic estimate sounds horrendous enough. If you multiply it by four, it simply will not be approved.
Ignore their spec, write your own as generic as possible. When they change it 1/2 way through, you bitch and moan, then write the gui that will give them what they want
I actually tried this, except I didn't bitch and moan. Let me tell you, they were really impressed when I used the last month just pulling extensions out of the sleeves, and implementing things they were only dreaming about. I told them about the lazy programmer, and they just kept repeating: "not lazy! smart!".
However, OS X being "Open Source", doesn't help here. The BSD license used does not prevent Apple from closing the source in the future. If Apple incorporated DRM into OS X.2, they could close the source as well and the "million of people" able to make that modification, would shrink to the few hackers who could disassemble the kernel. As Apple also control the hardware side of their product, A modded BIOS (or whatever) would probably also be needed.
If that scenario unfolds, you would have two choices:
- go with Apple, assimilate and eat DRM
- Avoid DRM and miss major hardware upgrades, stick to a forked kernel developed only by the OS community and slowly but surely turning legacy. There are lots of tricks that Apple can use if they want to make new stuff available only to the DRM customers.
When the pain becomes too much, most people will just go with DRM. Mind you, most businesses don't mind and don't care about DRM.Excellent point. E.g. Some people are still using KDE 1.x (otherwise it wouldn't still be bundled with SuSE). I for one am using it until I get a faster machine. There are lots of flaws, but I seriously disbelieve anything will be fixed.
The reason people are appalled is that MS might be doing this to gain a profit. They might be open about though, and state for how long time they will fix their products. Preferably in advance, not just suddenly dropping support. Then people could compare the cost of subscription vs. one-time-purchase. I would be surprised (MS is pushing subscription rather hard) if you were not better of subscribing to the software, if you wanted critical updates.
Today people gamble on how long they could use an unsupported product, mostly thinking at the time of purchase that this will be forever. It would be interesting to measure the minimal cost of a MS installation based on the history of unfixed security bugs. I bet you could still run windows NT 4.0 SP x, and you would now have to upgrade from Office 97 to Office XP. How does that compare to the subscription model, now and in the future ?
A big problem with the Kyoto agreement is that the worst polluters are not included.
;)
What pollutes more: destroying lakes, local agriculture, inhabitat etc. by dumping toxic waste or intense emission of CO2 ? The answer is blowing in the wind so to speak.
If you think toxines are the biggest pollution problem, I guess the US aren't that bad. And the countries you mentioned probably are. On the other hand, when it comes to CO2, the US is one of the major polluters. They emission 23% of the world total, and they account for only 5% of the worlds population. To my knowledge the main issue in Kyoto was CO2. The problem with CO2 emission is that it pollutes globally.
In 1995 every US citizen emissioned 23 tons of CO2. In India it was 1 ton. In China 3 ton. In Japan 11 ton. In Russia 13 ton. The worst polluters are not included ? Japan and Russia are.
The US govt would likely show more support, or even sign such an agreement, if there weren't some apparent economic motivations behind the requirements.
First of, as I stated in an earlier post the US themselves put in the apparent economic motivations.
Secondly, Anyone agrees that there is some relationship between economic growth and the amount of CO2 emission. I think it would be fair if we let the developing countries pollute a little more than they do today, maybe just 10% of what the USA is doing today, and gain some economic growth. Russia had an excuse for not reducing to 1992 level, because in 1992 Russia was at an economic colapse. Well, unfortunately the US was not as lucky
The other industrial countries which signed the treaty were under the same requirements as the US, and it should cost them as much money. They signed it to prevent pollution. Nobody stood to be rich from the (volountary) donations.
As for China, I definitely think they should have signed, but who can expect anything from an arrogant, communist dictatorship. Would you only agree to sign a treaty, if Iraq did ?
Sorry for bashing the US. The reason, I think, many Europeans (like me) do it, is because we are disapointed with the US. The USA is the world's only superpower; they are the world leader militarily, monetarily and culturally (to many). We admire you. But why is it that in everything you do, you only consider your own interest. America, what happened to the New World Order, when you won the cold war ? Why can't you be the world's policeman like we envisioned ?
Each time the US administration does something selfish, shortsighted and damaging, I tell myself that only a majority of US citizens are to blame. A lot of Americans are probably intelligent, well-meaning, responsible people. There just aren't enough.
This wasn't my point. I was never talking about donation to the third world in general. I was speaking specifically of Kyoto, and I though I made that very clear.
You misunderstood, so I will try to elaborate:
The poster insinuated that third world countries backed Kyoto to make easy money. Originally there was NOTHING in the Kyoto treaty which implied that money should be given to anyone. All countries just had to cut back the pollution to 1990 level before 2008 (or something). USA could have agreed to this and they wouldn't have had to pay any money to the third world. But USA themselves changed this, so that it was possible to buy the right to pollute from other countries. If then buying the right to pollute, makes the seller of the right greedy, and therefore you will not sign the treaty, it's like saying:
"Let's make a deal: I will pay you to slap yourself. But I don't think it is a fair deal, because you are so greedy to take my money."
Now, there could be a number of other reasons for the US not to sign Kyoto. Personally I probably wouldn't agree with them, though.
Plutonium halves every 24000 years. That's a lot of man hours (21 millions) before you can go from two guards to one guard.
On a side note: The large majority of households are not heated by electricity, but by wasted heat (hot water running in tubes, called "remote heating") from garbage/coal burning power plants. Some have actually expressed concern that if too big a share of the Danish energy comes from another source than these power plants, there will not be enough heat, or electricity would be wasted.
You guessed it, I live there.
Well, as far as I remember USA was the one who was in favour of the part of the treaty that allowed a country to PAY developing countries for the right to pollute instead of them. Most countries thought that it would be best if you HAD to pollute less.
Please don't accuse the third world of being greedy on account of this.
As I see it, the GPL doesn't protect us from this , anyway.
If a market leader with a GPL'ed installer owns enough copyright, they could simply re-license the installer as closed source, including a major rewrite and all the customers would be forced to switch or enter the same path as you describe assuming the vendor is that evil/stupid.
What is preventing Red Hat from doing this, except for the fact that they have a history of being open source ? A (virtual) monopoly is dangerous everywhere, mainly because it can dictate how things should be (aka. defacto standards), and which projects should receive support. I like the fact that SuSE supports the KDE project. If there were only Red Hat, KDE would not have received any support. I don't blame Red Hat for this, they just have other interests. Some of which I benefit from, some of which I don't.
Besides, if the installer is such a big problem, then why is it that there are so many of them ? Even a handfull of which is GLP'ed. I think the installer is only a small part of which makes a distribution. What matters more is testing of packages and integration. Having the installer be closed source software is a nuissance, because you cannot fix it, though.
tru64unix manpages
There are no frames, javascript or the like.
When an ISV want to write software he should use the LSB as a reference, which should provide some compatibility. However he cannot test against LSB, he has to test against a number of concrete distributions. Bringing that number down is important, maybe not to you, but to the ISV.
I am very pleased to hear that not only SuSE and UL but also Debian and Mandrake are committed to it.
As usual a lot of people here cry wolf at the slightest sign of distribution consolidation.
UL is not forcing anyone to do anything. It isn't even threatening to stifle competition, as there are plenty still. A non-dominating number of players decided that ISV are important, and to ensure that they get ISV support they form a group and standardize some of their binaries. I have no problem with the fact that Mandrake are not a member of UL. Fine! More competition. I just hope UL is big enough to push some standards, but not so big as to create de-facto standards like Red Hat.
In addition they get the advantage of additional combined testing. This is what I, as a user, consider extremely important. I have yet to try installing Linux on a nontrivial box without problems. There has always been something wrong, and if the next version fixed it, there was sure to be another problem with that. I am talking soundcard, 3D, scanner, printer, X in various resolutions, monitor frequenzy, CD-burner, IDE-DMA, netscape plugins, java. I won't even begin to mention upgrading versions. If anyone new to Linux read this: Always re-install Linux from scratch! Never upgrade!
Debians notion of stable and unstable is great and I hope one of the UL distributions will focus on being stable as opposed to the current obsession with being current.
In OS/2 there was only very basic stuff. So you basically needed windows apps for everything. And there was no signs of a change to the better.
Another difference is that the Linux apps are almost guaranteed to become better over time due to the GPL (and to some extent other opensource licenses). In the end a native linux application will catch up at leat to a level, where you don't need the features of the windows app.
And I guess You are using Galeon, Mozilla or Konqueror. Wow, you are so cool. Unfortunately they don't display some IE-only sites pretty well. So you also have to ditch YOUR shitty browser.
Go troll somewhere else