I probably wouldn't mind the new discussion system so much if it weren't so broken on konqueror. Just use IE6 on Microsoft Windows whatever and be happy. You Will Be Assimilated.
I'm sorry - I shouldn't need to use a specific web browser to view a particular website. I agree, but that argument was lost a long time ago. return -ETOOMANYIDIOTS. The system I have to use to access mail at work since I refuse to use MS Windows on my desktop fails to load when accessed via Firefox 2 or higher with an error message that says "This application requires Firefox 1.5 or better." Whee!
I cant see any reasonable judge being able to see this in the patent holder's favor. Unless his bank account is also padded. I'm not that cynical yet, but I'm getting close. Sometime I'm going to have to write up a journal entry on my Ninong's mayorial election that was stolen and his thoughts on how computerized voting machines would make things perfect, Just Like They Are in The United States. I tried to set him straight...
The United States is indistinguishable from a 3rd world country sometimes and unfortunately it takes actually living in one to understand it.
Backwash can be a son of a bitch, so I hope the folks who been swinging things their way are ready for it. Is there an economist out there in SlashDot land who can argue an economic case against the current patent system as well as our own NYC Lawyer is doing against the..AAs? Please step forward.
That's mentioned in TFA:
Also, as noted in the comments to the link above, it would appear that there's a fair amount of prior art. In fact, Apple even sent over some prior art concerning the patent just before it was originally supposed to be issued last summer -- but somehow patent holder's lawyers talked their way around it. If this is true, this is not going to be an easy patent to defeat.
... lose's it's... it's word's... reader's, it's... seem's... becau'se... pay's... word's. If I had a nickel for every time a slashdotter misused an apostrophe, I'd be richer than Bill Gates.
We are watching your every word. Be afraid, be very afraid.
-Your Friendly Neighborhood Spelling and Grammar Nazi
RTFA. The Javascript injection is separate from the Apache process, so it neither appears in the logs nor requires Apache to be running as root.
The virus patches itself into a running kernel via/dev/mem. It probably requires a specific kernel version to do the patching, but that's a completely different issue.
The extra wieght of the Higgs particle is what makes the change fall out of my pocket! Perhaps, but more importantly, it is not symmetric and has been known to attract left or right socks more strongly than the other. This explains the dryer effect.
However, I don't agree with the first constraint. Backwards compatibility should not be an issue here. Period. Agreed. This only makes sense when thinking about IE as a critical O/S component.
And don't give me that crap about old websites that are no longer maintained. Agreed. This only makes sense when you think about API between various O/S components (which are static and cannot be changed).
Movement towards standards compliance is a good thing, but Microsoft is not doing that. This is a thinly veiled way to keep every web developer coding specifically for IE, every time they write a website. Microsoft is the center of the universe, and all that. Agreed. Disagreed. Disagreed. Agreed.
This is all about keeping Microsoft Windows itself afloat. The decision to make a browser (and a non-standard one at that) a vital O/S component was not a good engineering decision, but it made sense to management at a time when Netscape Must Die. I posted a longer analysis earlier.
I think you're missing the point somewhat here, as are a lot of people. The core issue is that Microsoft have painted themselves into a corner by not following the specifications in the past. I think you have part of it right, just complete the thought.
First, Microsoft only does things to further their own goals. This is not a matter for the convenience of non-Microsoft developers. This is an internal engineering decision for the benefit of internal Microsoft Windows developers same as the paragraph-spacing-for-word96-for-macos "tags" in MSOOXML.
Second, Internet Explorer is a critical operating system component. Moreover, it functionally has a similar role to that of the Unix shell. As such, its defaults for how it renders HTML is a fundamental system interface. Nobody is perfect here. As an analogy, consider how on Linux systems,/bin/sh is usually a link to bash, then people write shell scripts with #!/bin/sh and then use bash extensions, which of course breaks when a Posix compliant/bin/sh exists. See we can make the same mistake too. The difference is that the Microsoft engineering team can't just tell apps to fix themselves (or have the users recompile the now-broken app). They must support the old interface in the same way that they have to keep other ABIs compatible.
Do you see how this makes sense? IE8 can't break IE6 stuffs because that's a critical O/S interface that too many internal components rely upon. We can just say "You're an idiot. If you must use bashisms in scripts, you must also use #!/bin/bash #!/bin/bash-4.0.6 or whatever." Microsoft doesn't have the same luxury.
Defaulting to IE6 behavior makes total sense if you think of it this way. So what would you rather have, standards compliance when viewing web pr0n and a non-functional MS Windows O/S (which probably cannot boot far enough to even get to the point where you can look at pr0n in the first place) or having your O/S continue to work as it has?
As a general engineering solution, relying on specific version numbers to control behavior went out of fashion a couple decades ago. I'm not surprised other browser developers are not enthusiastic about the idea. However, I'm also not surprised to see this kind of thing in an end-to-end controlled environment where you can trace specific desired "non-standard" behavior to a specific version number and act accordingly.
I will now present the source code to Internet Explorer 8:
#include <ieglobals.h> extern int getDesiredIEVersion(int, char **, char **);
void main(argc, argv, envp) int argc; char **argv, **envp; {
switch (getDesiredIEVersion(argc, argv, envp)) {
case 6:
default:
ie6_main(argc, argv, envp);
break;
case 7:
ie7_main(argc, argv, envp);
break;
case 8:
ie8_main(argc, argv, envp);
break;
} }
Unfortunately that means they are also able to exert an influence (large, although I'd hesitate to say disproportionate) on other games - I for one believe that WotLK (the next expansion) has been done or nearly done since before the end of the year, and that they are waiting to unleash it a month or so before the 'next big competitor' (I believe Age of Conan) is released. For what it's worth, I was in the local GameStop a couple weeks ago and asked if I could sign up for a WotLK waiting list. The clerk showed me his list of upcoming games with scheduled release dates and it was way down at the bottom with scheduled release date 30-Nov-2008.
They have to balance between people anxious for new content and have either finished or don't care about the end game and people who want a decent chance to play the end game stuffs before moving on. With the demographic data they have, they should be able to get a pretty good estimate of who is whom. They also have to keep releasing expansions to continue drawing new players. They've done a great job so far and the only real mistakes[1] they've made were underestimating popularity.
[1] Killing Mac sound a couple of patches ago to cater to crappy MS Windows notebook computers wasn't very nice, but to Blizzard's credit, they already had a "fix" posted in their support forum before I logged in for the evening.
Yes, they do - the Warden thingy must be able to figure that out. Just as interesting would be a breakdown of accounts by level, number of alts, average level, total time played, player age, gender etc. I don't think they would ever publish it[1], but I'm sure they have enough revenue to do extensive analysis on all their account data.
I was planning to subscribe again to WoW when I heard that they increased the rate at witch you gain XP. Does it really make a difference? You have to do quests to get the full leveling benefit, but the nastiest group-only quests have been nerfed (most of the outdoor elites are gone). Yes, it's much faster, easier and balanced towards solo play now.
Quest givers show up on the minimap, reducing another source of tedium. Get a friendly mage or warlock to teleport you to Shattrath so you can set your hearthstone there. That makes it easy to get to just about anywhere you need to go fairly quickly.
New stuff that is exactly like the old stuff, and does nothing but kill time while they work on the next expansion, which is the only part that actually advances the plot. The thing is that that's not true. There are ten million active accounts, certainly not all of the last three or four million are advanced Black Temple raiders who have completed all their epic sets.
What has impressed me over the time I've been playing, since December 2006 (about a month before BC was released) is how much attention to detail Blizzard pays to every facet of the game.
To name one recent example, they changed the rules for leveling recently to make experience gain higher from completed quests, experience required for a new level lower (one of those is between levels 20 to 60 and the other 30 to 60, I forget which) and at the same time changed the vendor discount for reputation - revered and exalted now have bigger discounts and removed most of the outdoor elite monsters in the Old World. What do those changes mean to the game play?
Leveling to 20 remained unchanged. It's quite difficult to avoid learning how to do that already. The worst grinding was nerfed out of the game. It's now possible to do most of the quests solo (because finding someone to level with you has become all but impossible), the experience gain is rapid enough to not be particularly painful (in my n00b opinion) and between the added vendor discount for rep and added experience, you want to and can do most of the quests quickly by yourself. The main side effect of this change has been that "leveling services" are out of business. Good going Blizzard. They also want to get most of the more recent players into Outlands before the next expansion. That will happen.
Sadly, another side effect of BC is that there's a relatively huge grind to get the epic flying mount due to the amount of gold involved. You can either grind for it or purchase it from a gold seller. The grinding has led to the absurd situation that crafting seriously sucks due to the high price of raw mats in the auction house that are being sold at prices higher than the finished goods they can be used to make which just makes it all the more worthwhile to buy gold for the few items you need to craft along the way. I expect Blizzard to attempt to balance the economy, though I don't know how they're going to do it.
If you spend any time reading the WoW forums and don't play the game yourself, you would get the impression that people are quitting in droves. Obviously they are not as the community continues to grow. Certainly I will be one of those purchasing the next expansion on the day it's released.
Who are these people? Are they smart enough to breed? I don't know, but I'm so happy to be typing from a Mac where I'm completely immune to this kind of stuff. As a matter of fact, hold my beer and watch th
Windows 7 -- if it's true -- sounds like it could be what Vista was supposed to be. Of course, by now no one will care. It doesn't matter whether people care or not so long as it's the only O/S you can buy preinstalled on a computer. Microsoft wins no matter what they do. Ah, the joys of being a monopoly.
However, in the non-academic world it's pretty obvious that there's far more Java than LISP code in production. At one time, everyone had a mood ring. So what?
LISP has a half century under its belt. It is widely used today in conjunction with XEmacs/Emacs.
Let's have this discussion after the end of (signed 32 bit) time in the 2040s and we'll see how many people are still actively developing in Java. I'll bet anything that LISP will still be in use then.
A fad is a fad because it's popular. A fad isn't a fad when it develops the kind of mindshare that the LISP language has gotten[1]. COBOL, like Java, was a fad. The fact that poor, lost souls are still maintaining COBOL code today doesn't make it any less a fad. But pay close attention Java lovers, that is your future, if that's the only computer language you can deal with.
[1] In some respects, Java is a descendant of LISP, at least with respect to how it deals with byte code.
Aside from "Geek Value", there doesn't seem to be a compelling reason to me for having access to the OS/2 code (and for real geek value, I would prefer being able to read System V sources by Thompson and Ritchie (though, for all I know that is available somewhere and I just don't know where to look)). Maybe you're confusing Unix V[1] with System V. The likelihood that commercial System V had any significant Thompson and Ritchie code is near zero. Ritchie's original C compiler[2] was never ported and Steve Johnson's PCC was the one that was widely used well before then.
By 1983, Ken Thompson had moved on to Plan 9 which not only has considerable geek value[3], but was later open sourced so you can read that source code if you're so inclined. For System V internals, the Bach Book http://portal.acm.org/citation.cfm?id=8570 contains sufficient detail to write your own kernel.
6th Edition Unix has considerable geek value and it was "open sourced" - the so-called Lion's Book with the famous long comment in the scheduler regarding some context switching magic and ending with "you are not expected to understand this".
The one piece of source code I'd love to be able to find and read again was an include file of VM and swapping constants that included a discussion of VAX core memory costs as driving selection of some of the constants. That appeared in both of the m68k System V/R2 systems I owned in the 80's.
The OS/2 afficionados should just bite the bullet, try to get complete specs on the system and clean room rewrite it. The value of an open source OS is not the direct cost, it is the value of having a system that can never be taken away from you, as this whole incident amply illustrates.
[1] A dead end Unix fork that had the first real virtual memory implementation.
[2] I was able to read through some of it and alas, I did not come away enlightened. For compilers, I recommend Davie and Morrison, Recursive Descent Compiling - http://portal.acm.org/citation.cfm?id=1098737 the error recovery algorithm they describe is priceless.
[3] For True Believers that the One True O/S is the one running on Ken Thompson's desktop.
FORTRAN (the current version) is still very much alive in academia. You'll find a lot of scientists still write programs in the language. In fact that was what it was designed for, it's NOT a General Purpose Language such as Java, Pascal, C/C++. Historically speaking, that is not true. FORTRAN is important because it proved that compilers could make efficient translations of source code (think of all the strange rules and exceptions in the older versions like the odd restrictions on expressions to compute array indices). Before FORTRAN it was still a wide open debate as to whether "high level" languages were worth anything at all. Calling FORTRAN a specialized language is not fair either as later dialects greatly expanded its scope and at the time it was designed, the only use of computers was number crunching.
Currently, FORTRAN is still meaningful because it defines specific evaluation order of expressions. That is essential for scientific computations. Today's FORTRAN is not your Father's FORTRAN.
I am guessing that the real lesson here is that nothing is infallible, but that the 777 is pretty-darn good. That's what I read out of it too. The track record remains and speaks for itself - those are damn good planes.
They experienced a catastrophic failure losing both engines at low altitude where the plane has all the flight worthiness of a brick and nobody died.
This needs to be fixed now. Embedded systems can have a long lifetime (think building thermostats, alarm systems, industrial control systems and other infrastructure). It's been fixed. Recompile with 64-bit userland and a 64-bit kernel. Coding changes now won't solve anything on existing embedded systems and for near future embedded systems where speed and space is of utmost concern, it makes sense to keep it 32-bit for as long as you can.
Code speaks louder than words, look: $ grep 64.*time_t asm-x86/**/*.h asm-x86/msgbuf.h: * - 64-bit time_t to solve y2038 problem asm-x86/sembuf.h: * - 64-bit time_t to solve y2038 problem asm-x86/shmbuf.h: * - 64-bit time_t to solve y2038 problem
The context is:
* Pad space on i386 is left for:
* - 64-bit time_t to solve y2038 problem
* - 2 miscellaneous 32-bit values
Older flavours of Unix will wrap-around on 32 bit int. More modern systems use time_t instead of int and may or may not be affected by the problem. That's right and wrong. Version 7 (the oldest version of Unix I used) had time(2) returning a signed long integer. I don't think they changed that to time_t until System V. The most common Unix systems have always had a 32 bit long int.
Glibc defines time_t as a signed long integer.
If you've been a megabozo and using int for time(2) your application will lose. But if you've been a bozo and hardcoding long int, a recompile on a 64 bit platform will still fix you.
Slashdot Mirror
OMFG. Now there are Microsoft Windows jokes on Slashdot the rest of us are supposed to understand. Run for the hills!
I chose that word deliberately. Do you aspire to be an idiot or does it just come naturally?
And your point is?
Unless his bank account is also padded. I'm not that cynical yet, but I'm getting close. Sometime I'm going to have to write up a journal entry on my Ninong's mayorial election that was stolen and his thoughts on how computerized voting machines would make things perfect, Just Like They Are in The United States. I tried to set him straight
The United States is indistinguishable from a 3rd world country sometimes and unfortunately it takes actually living in one to understand it.
Backwash can be a son of a bitch, so I hope the folks who been swinging things their way are ready for it. Is there an economist out there in SlashDot land who can argue an economic case against the current patent system as well as our own NYC Lawyer is doing against the
... lose's it'sWe are watching your every word. Be afraid, be very afraid.
-Your Friendly Neighborhood Spelling and Grammar Nazi
RTFA. The Javascript injection is separate from the Apache process, so it neither appears in the logs nor requires Apache to be running as root.
/dev/mem. It probably requires a specific kernel version to do the patching, but that's a completely different issue.
The virus patches itself into a running kernel via
This will be the most secure and bug free Capitalism ever released! How can you not see that?
This is all about keeping Microsoft Windows itself afloat. The decision to make a browser (and a non-standard one at that) a vital O/S component was not a good engineering decision, but it made sense to management at a time when Netscape Must Die. I posted a longer analysis earlier.
First, Microsoft only does things to further their own goals. This is not a matter for the convenience of non-Microsoft developers. This is an internal engineering decision for the benefit of internal Microsoft Windows developers same as the paragraph-spacing-for-word96-for-macos "tags" in MSOOXML.
Second, Internet Explorer is a critical operating system component. Moreover, it functionally has a similar role to that of the Unix shell. As such, its defaults for how it renders HTML is a fundamental system interface. Nobody is perfect here. As an analogy, consider how on Linux systems,
Do you see how this makes sense? IE8 can't break IE6 stuffs because that's a critical O/S interface that too many internal components rely upon. We can just say "You're an idiot. If you must use bashisms in scripts, you must also use #!/bin/bash #!/bin/bash-4.0.6 or whatever." Microsoft doesn't have the same luxury.
Defaulting to IE6 behavior makes total sense if you think of it this way. So what would you rather have, standards compliance when viewing web pr0n and a non-functional MS Windows O/S (which probably cannot boot far enough to even get to the point where you can look at pr0n in the first place) or having your O/S continue to work as it has?
As a general engineering solution, relying on specific version numbers to control behavior went out of fashion a couple decades ago. I'm not surprised other browser developers are not enthusiastic about the idea. However, I'm also not surprised to see this kind of thing in an end-to-end controlled environment where you can trace specific desired "non-standard" behavior to a specific version number and act accordingly.
I will now present the source code to Internet Explorer 8:
#include <ieglobals.h>
extern int getDesiredIEVersion(int, char **, char **);
void main(argc, argv, envp)
int argc;
char **argv, **envp;
{
switch (getDesiredIEVersion(argc, argv, envp)) {
case 6:
default:
ie6_main(argc, argv, envp);
break;
case 7:
ie7_main(argc, argv, envp);
break;
case 8:
ie8_main(argc, argv, envp);
break;
}
}
They have to balance between people anxious for new content and have either finished or don't care about the end game and people who want a decent chance to play the end game stuffs before moving on. With the demographic data they have, they should be able to get a pretty good estimate of who is whom. They also have to keep releasing expansions to continue drawing new players. They've done a great job so far and the only real mistakes[1] they've made were underestimating popularity.
[1] Killing Mac sound a couple of patches ago to cater to crappy MS Windows notebook computers wasn't very nice, but to Blizzard's credit, they already had a "fix" posted in their support forum before I logged in for the evening.
Yes, they do - the Warden thingy must be able to figure that out. Just as interesting would be a breakdown of accounts by level, number of alts, average level, total time played, player age, gender etc. I don't think they would ever publish it[1], but I'm sure they have enough revenue to do extensive analysis on all their account data.
[1] In the right hands, that data is priceless.
Quest givers show up on the minimap, reducing another source of tedium. Get a friendly mage or warlock to teleport you to Shattrath so you can set your hearthstone there. That makes it easy to get to just about anywhere you need to go fairly quickly.
What has impressed me over the time I've been playing, since December 2006 (about a month before BC was released) is how much attention to detail Blizzard pays to every facet of the game.
To name one recent example, they changed the rules for leveling recently to make experience gain higher from completed quests, experience required for a new level lower (one of those is between levels 20 to 60 and the other 30 to 60, I forget which) and at the same time changed the vendor discount for reputation - revered and exalted now have bigger discounts and removed most of the outdoor elite monsters in the Old World. What do those changes mean to the game play?
Leveling to 20 remained unchanged. It's quite difficult to avoid learning how to do that already. The worst grinding was nerfed out of the game. It's now possible to do most of the quests solo (because finding someone to level with you has become all but impossible), the experience gain is rapid enough to not be particularly painful (in my n00b opinion) and between the added vendor discount for rep and added experience, you want to and can do most of the quests quickly by yourself. The main side effect of this change has been that "leveling services" are out of business. Good going Blizzard. They also want to get most of the more recent players into Outlands before the next expansion. That will happen.
Sadly, another side effect of BC is that there's a relatively huge grind to get the epic flying mount due to the amount of gold involved. You can either grind for it or purchase it from a gold seller. The grinding has led to the absurd situation that crafting seriously sucks due to the high price of raw mats in the auction house that are being sold at prices higher than the finished goods they can be used to make which just makes it all the more worthwhile to buy gold for the few items you need to craft along the way. I expect Blizzard to attempt to balance the economy, though I don't know how they're going to do it.
If you spend any time reading the WoW forums and don't play the game yourself, you would get the impression that people are quitting in droves. Obviously they are not as the community continues to grow. Certainly I will be one of those purchasing the next expansion on the day it's released.
LISP has a half century under its belt. It is widely used today in conjunction with XEmacs/Emacs.
Let's have this discussion after the end of (signed 32 bit) time in the 2040s and we'll see how many people are still actively developing in Java. I'll bet anything that LISP will still be in use then.
A fad is a fad because it's popular. A fad isn't a fad when it develops the kind of mindshare that the LISP language has gotten[1]. COBOL, like Java, was a fad. The fact that poor, lost souls are still maintaining COBOL code today doesn't make it any less a fad. But pay close attention Java lovers, that is your future, if that's the only computer language you can deal with.
[1] In some respects, Java is a descendant of LISP, at least with respect to how it deals with byte code.
By 1983, Ken Thompson had moved on to Plan 9 which not only has considerable geek value[3], but was later open sourced so you can read that source code if you're so inclined. For System V internals, the Bach Book http://portal.acm.org/citation.cfm?id=8570 contains sufficient detail to write your own kernel.
6th Edition Unix has considerable geek value and it was "open sourced" - the so-called Lion's Book with the famous long comment in the scheduler regarding some context switching magic and ending with "you are not expected to understand this".
The one piece of source code I'd love to be able to find and read again was an include file of VM and swapping constants that included a discussion of VAX core memory costs as driving selection of some of the constants. That appeared in both of the m68k System V/R2 systems I owned in the 80's.
The OS/2 afficionados should just bite the bullet, try to get complete specs on the system and clean room rewrite it. The value of an open source OS is not the direct cost, it is the value of having a system that can never be taken away from you, as this whole incident amply illustrates.
[1] A dead end Unix fork that had the first real virtual memory implementation.
[2] I was able to read through some of it and alas, I did not come away enlightened. For compilers, I recommend Davie and Morrison, Recursive Descent Compiling - http://portal.acm.org/citation.cfm?id=1098737 the error recovery algorithm they describe is priceless.
[3] For True Believers that the One True O/S is the one running on Ken Thompson's desktop.
Currently, FORTRAN is still meaningful because it defines specific evaluation order of expressions. That is essential for scientific computations. Today's FORTRAN is not your Father's FORTRAN.
A generation ago it was "BASIC causes irreversible and total brain damage."
Nearly all computer languages come and go. Java is a fad. Others like LISP, are timeless. "C"/Unix and "C++"/KDE show no signs of going away either.
I wouldn't worry about it. OO sucks, Real Programmers can code in more languages than they count. Functional languages ftw!
They experienced a catastrophic failure losing both engines at low altitude where the plane has all the flight worthiness of a brick and nobody died.
Good troll. There's an easy answer to the dirty diapers problem - let Mom or the Yaya take care of them. Works for me, YMMV.
...
Now about getting sleep when a screaming infant is in the house
Code speaks louder than words, look:
$ grep 64.*time_t asm-x86/**/*.h
asm-x86/msgbuf.h: * - 64-bit time_t to solve y2038 problem
asm-x86/sembuf.h: * - 64-bit time_t to solve y2038 problem
asm-x86/shmbuf.h: * - 64-bit time_t to solve y2038 problem
The context is:
* Pad space on i386 is left for:
* - 64-bit time_t to solve y2038 problem
* - 2 miscellaneous 32-bit values
Glibc defines time_t as a signed long integer.
If you've been a megabozo and using int for time(2) your application will lose. But if you've been a bozo and hardcoding long int, a recompile on a 64 bit platform will still fix you.