Number 23 on the "101 things mozilla can do that IE cannot." list was colored source viewing, with HTML syntax markup. This is NOT a win for mozilla, seeing as how in their attempt to add color highlighting, they screwed up the primary purpose of "view source" which is to try to determine what's wrong when a page isn't displaying right. Their color highlighting algorithm, whatever it is, tends to LIE about what the source looked like omitting things that it didn't understand.
I understand that it would PROBABLY tend to be be more readable, but on what authority can you make the statement that you know this is the case. You will only ever be able to see a very unrepresentative sample of closed source code. You can only see that closed source which is put out by companies you have worked for or are working for. That's what "closed source" means. So what are you comparing with to make the judgement that open source "tends" to be more readable. If you could make the comparasin with it, it wouldn't be closed source.
Like the subject says. Automatic updates are not a feature that will make people love MS over Linux. Even people who like MS would typically still prefer to decide for THEMSELVES when it's a good time to upgrade instead of having no choice over the matter.
Not buying something something you would have otherwise bought can be a boycott. Not buying something because you can't even make use of it in the first place is not a boycott. For example, the fact that I haven't bought any motorcycle helmets isn't that I'm boycotting them - it's that I don't own a motorcycle, so there'd be no point.
And like most/. polls, my answer isn't listed.
I pronounce it "squirrel". I worked in a place where everyone said "sequel" and I hated that name. The strange thing is, I don't know why. I just did.
To the moderator who said "overrated". I hope you realize that by definition it's not actually possible for something that hasn't been rated yet to be "overrated". Seing "overrated" as the first and only moderation so far doesn't make any sense at all. Hopefully a metamoderator will notice this fact.
Viper was awful the last time I tried using it. The real problem is that while someone went and implemented viper in emacs, vi evolved and vi users started using vim, which has a few emacs-like features like split-screen - the problem is that people used to vim will find that their vim keybindings don't work in emacs/viper. viper instead insists that you use the emacs keys for things like, for example, resizing the window. So it's not a simple drop-in-replacement for vim, although it is for old fashioned vanilla vi, which people don't use much anymore.:wq
viper mode doesn't emulate the colon prompt well. It is not a drop-in replacment for using vi, although someone who only uses vi casually wouldn't notice the difference.
It's a Mac Thing. You wouldn't understand the fact that some people think there should be only one text editor. It goes along with the mentality that less choice is a good thing, which is also a Mac thing you won't understand.
M-x viper has some of the keybindings of vi, but misses out on all the colon-prompt commands (actually ex commands). I'll take:%s/findme/replace/g any day over emacs's "Let's bind most of the useful versions of functions to keys your keyboard doesn't have, or better yet not bind them at all and let you try to wade through info pages to figure out how to give them a binding, but the pages that show you the lisp code to write don't have links pointing you to something telling you WHERE these things are supposed to go" mentality.
Configuring a simple keybinding shouldn't require hours of documentation reading.
Who has burden of proof here? Presumably if someone wanted to accuse a company of de-GPL-ing a derived work, it's up to the prosecution to show that this is the case, isn't it? If so, the company doesn't need to be sure it can prove it's work is not derived from GPL work, as the comment claims. It only needs to be sure it cannot be proved that it DID derive the work.
link with program != link with library. Libraries are things explicitly DESIGNED such that the finished product is a pile of functions you want other people to call.
This is one place where the British way is less logical than the American (there aren't many, I'll admit). In American usage, a group that wants [Note, "group that wants", not "group that want"] to be referred to as plural just has to pick a plural noun for their name. So for example, when watching sporting events here you'll hear phrases like "The Dallas Cowboys *are* coming onto the field.", but also phrases like, "The Milwaukee Wave is coming onto the field." - The usage depends on what sort of name the group is using for itself. The British way of sometimes using a singular noun as if it was plural when it refers to a group is a bit illogical to me because it requires knowing contextual information beyond what part of speech the word really is. To me this seems like a breackdown of the well ordered layer between grammar and semantics. Consider: is this grammar correct: "Slipknot are hard to get out."? Well, in British usage you can't tell yet, because I might be talking about it being difficult to remove a certain type of knot, in which case my grammar was wrong, or I might have been talking about the band called "Slipknot", and complaining about how they just won't seem to leave, in which case my grammar was correct.
To my logical mind, this is unneccessarily messy. Just say that plural nouns are plural, and singular ones are singular and be done with it.
Executable binaries *ARE* source code - in machine language. The difference between a *.c file and a list of machine language program code is not one of classification, but of degree. They both require specialized training to understand. They both are completely unfathomable to the casual observer who hasn't studied them. The only difference is in the degree of the obfuscation, not that such obfuscation exists.
And on that topic, There were once legit uses for C code obfuscation - companies would sometimes find themselves in a position where they HAD to release source for a technical reason, but they didn't want that source to be human readable, so they'd run an obfuscater on the code that would do things like remove all the whitespace, break lines at random places, transform variable names into gibberish strings, and so on.
Is that really enough obfuscation to legally count as not being source code anymore?
Where is the legal definition of that line drawn? I'll tell you where - nowhere - because the lawyers who deal with these issues don't fully understand what they are regulating.
IANAL, but for those who ARE, how do they deal with fuzzy situations like this, where even the technical people can't really come up with a strict definition of where to draw the line. (This also is why the plan to split Microsoft into OS and App companies was doomed from the start - you can't even get a precise definition of what is and isn't core functionality the OS should have. Technically, in Windows the entire desktop gui is an application called EXPLORER.EXE and not really a low-level OS component. But I don't think anyone, even people like me who hate Microsoft, would think it fair to force them to pull that program out of the OS distribution. While it's not technically part of the OS in the strictest sense, neither is stuff like the UNIX "ls" command, but everyone sort of expects it to be bundled with the OS.
This same problem exists in the DMCA, where it makes no provision for how cryptic a thing must be before running a program to transform it into something human readable counts as an act of "decrpytion". Technically every plain ascii text file is "encrypted" in a code known as ASCII that replaces letters with single-byte numeric codes, and every time you pull it up into a text file viewer you are "decrypting" it into human-readable form. Granted, it's a pretty *BAD* encrpytion algorithm, with many programs out there that can decrypt it, but then again so are a lot of the things that ARE protected by the DMCA, like the DVD CSS algorithm.
A vendor who distributes non-GPL modules is _not_ protected by the module interface per se, and should feel very confident that they can show in a court of law that the code is not derived.
That last statement proves that it IS possible to have your module not be GPL-infected. Otherwise there'd be no point to having to "feel very confident that they can shoud in a court of law that the code is not derived", since such a thing would be moot if the GPL (technically LGPL in this case) was "virally" infecting said code just by it's very presense somewhere near the kernel.
Get a sense of proportion. There is a huge difference between someone who wants to take a GPL'ed app, tweak it a bit, and sell it under a non-GPL license, and someone who wants to take a GPL'ed library and include it as a small part of a non-GPL'ed program. That would be stretching the definition of "derivative work" beyond the level of common sense.
And on the day it becomes possible to publish an executable with source that cannot be simply turned into a library file by sticking a simple wrapper around it, that argument might start to make sense. Until that time, there is very little difference.
So maybe you ought to think about this issue a little more before you post this tripe again.
If you want to convince people of your position, it helps not to come off sounding like a pompous ass.
I'd much rather have the case where the lawyers *know* they don't know anything than the case where they *think* they do because one of them downloaded and installed a device driver once.
What's noteworthy is that AOL is getting companies to pay AOL to fix a problem AOL created themselves. Pretty sweet deal. Kinda like the Far Side cartoon where a guy gets a brick thrown through his window, and attached to the brick is an advertisement for a window glass repair shop.
I try, I really, really try, to like SuSe. It comes with so many packages you can install, and is just very cool ("have a lot of fun"), but they keep fscking up their kernel source that gets installed and that ruins it for me. I've tried SuSe 6, 7, and now 8. Each time their installation gets screwed up whenever I need to make a small change to the kernel config and recompile. The reason, I only just found out recently, is that the kernel source they install on your system is NOT configured the same way as the kernel source from which they made their prebuilt kernels. This means I have to just make a wild guess as to what the settings were for all those config options. With over hundreds of config options to choose from, the probability of my random guess being correct approaches zero. So inevitably whenever I want to change something small (like setting it up to read IDE CD-ROMS under SCSI emulation so I can run burner software), I end up accidentally changing a heck of a lot of settings because I had to guess what the whole list of settings originally were.
Suddenly stuff breaks all over the place on the next reboot as drivers can't be found (I configured as a module what had been originally compiled-in). And then I look in the source tree and can't even FIND the drivers existing anywhere in the source for what had obviously been working using the prebuilt binary, so now I know I don't even have the same SOURCE CODE they were using, much less the same config options for that source.
No, it's not. Competition isn't about attacking all competitors how ever they see fit. (Note the emphasis.) It's about doing so withing certain rules of good ethics and fair conduct where you let your products compete on their own merits. This is not the Microsoft business model.
Slashdot Mirror
The phrase he used was "case law regarding journalism, satire, and commentary", not "case law regarding satire alone".
Number 23 on the "101 things mozilla can do that IE cannot." list was colored source viewing, with HTML syntax markup. This is NOT a win for mozilla, seeing as how in their attempt to add color highlighting, they screwed up the primary purpose of "view source" which is to try to determine what's wrong when a page isn't displaying right. Their color highlighting algorithm, whatever it is, tends to LIE about what the source looked like omitting things that it didn't understand.
I understand that it would PROBABLY tend to be be more readable, but on what authority can you make the statement that you know this is the case. You will only ever be able to see a very unrepresentative sample of closed source code. You can only see that closed source which is put out by companies you have worked for or are working for. That's what "closed source" means. So what are you comparing with to make the judgement that open source "tends" to be more readable. If you could make the comparasin with it, it wouldn't be closed source.
Like the subject says. Automatic updates are not a feature that will make people love MS over Linux. Even people who like MS would typically still prefer to decide for THEMSELVES when it's a good time to upgrade instead of having no choice over the matter.
Not buying something something you would have otherwise bought can be a boycott. Not buying something because you can't even make use of it in the first place is not a boycott. For example, the fact that I haven't bought any motorcycle helmets isn't that I'm boycotting them - it's that I don't own a motorcycle, so there'd be no point.
I pronounce it "squirrel". I worked in a place where everyone said "sequel" and I hated that name. The strange thing is, I don't know why. I just did.
To the moderator who said "overrated". I hope you realize that by definition it's not actually possible for something that hasn't been rated yet to be "overrated". Seing "overrated" as the first and only moderation so far doesn't make any sense at all. Hopefully a metamoderator will notice this fact.
My post obviously referred to Mozilla in the past tense, as in "has had some problems". No, I was not speaking of version 1.2
Viper was awful the last time I tried using it. :wq
The real problem is that while someone went and implemented viper in emacs, vi evolved and vi users started using vim, which has a few emacs-like features like split-screen - the problem is that people used to vim will find that their vim keybindings don't work in emacs/viper. viper instead insists that you use the emacs keys for things like, for example, resizing the window. So it's not a simple drop-in-replacement for vim, although it is for old fashioned vanilla vi, which people don't use much anymore.
viper mode doesn't emulate the colon prompt well. It is not a drop-in replacment for using vi, although someone who only uses vi casually wouldn't notice the difference.
It's a Mac Thing. You wouldn't understand the fact that some people think there should be only one text editor. It goes along with the mentality that less choice is a good thing, which is also a Mac thing you won't understand.
And I'm sure the fact that "pico" is a common prefix for a lot of scientific terms has nothing to do with that high result from google.
M-x viper has some of the keybindings of vi, but misses out on all the colon-prompt commands (actually ex commands). I'll take :%s/findme/replace/g any day over emacs's "Let's bind most of the useful versions of functions to keys your keyboard doesn't have, or better yet not bind them at all and let you try to wade through info pages to figure out how to give them a binding, but the pages that show you the lisp code to write don't have links pointing you to something telling you WHERE these things are supposed to go" mentality.
Configuring a simple keybinding shouldn't require hours of documentation reading.
Who has burden of proof here? Presumably if someone wanted to accuse a company of de-GPL-ing a derived work, it's up to the prosecution to show that this is the case, isn't it? If so, the company doesn't need to be sure it can prove it's work is not derived from GPL work, as the comment claims. It only needs to be sure it cannot be proved that it DID derive the work.
Asking somoene to prove a negative is dirty pool.
link with program != link with library.
Libraries are things explicitly DESIGNED such that the finished product is a pile of functions you want other people to call.
This is one place where the British way is less logical than the American (there aren't many, I'll admit). In American usage, a group that wants [Note, "group that wants", not "group that want"] to be referred to as plural just has to pick a plural noun for their name. So for example, when watching sporting events here you'll hear phrases like "The Dallas Cowboys *are* coming onto the field.", but also phrases like, "The Milwaukee Wave is coming onto the field." - The usage depends on what sort of name the group is using for itself. The British way of sometimes using a singular noun as if it was plural when it refers to a group is a bit illogical to me because it requires knowing contextual information beyond what part of speech the word really is. To me this seems like a breackdown of the well ordered layer between grammar and semantics. Consider: is this grammar correct: "Slipknot are hard to get out."? Well, in British usage you can't tell yet, because I might be talking about it being difficult to remove a certain type of knot, in which case my grammar was wrong, or I might have been talking about the band called "Slipknot", and complaining about how they just won't seem to leave, in which case my grammar was correct.
To my logical mind, this is unneccessarily messy. Just say that plural nouns are plural, and singular ones are singular and be done with it.
Executable binaries *ARE* source code - in machine language. The difference between a *.c file and a list of machine language program code is not one of classification, but of degree. They both require specialized training to understand. They both are completely unfathomable to the casual observer who hasn't studied them. The only difference is in the degree of the obfuscation, not that such obfuscation exists.
And on that topic, There were once legit uses for C code obfuscation - companies would sometimes find themselves in a position where they HAD to release source for a technical reason, but they didn't want that source to be human readable, so they'd run an obfuscater on the code that would do things like remove all the whitespace, break lines at random
places, transform variable names into gibberish strings, and so on.
Is that really enough obfuscation to legally count as not being source code anymore?
Where is the legal definition of that line drawn? I'll tell you where - nowhere - because the lawyers who deal with these issues don't fully understand what they are regulating.
IANAL, but for those who ARE, how do they deal with fuzzy situations like this, where even the technical people can't really come up with a strict definition of where to draw the line. (This also is why the plan to split Microsoft into OS and App companies was doomed from the start - you can't even get a precise definition of what is and isn't core functionality the OS should have. Technically, in Windows the entire desktop gui is an application called EXPLORER.EXE and not really a low-level OS component. But I don't think anyone, even people like me who hate Microsoft, would think it fair to force them to pull that program out of the OS distribution. While it's not technically part of the OS in the strictest sense, neither is stuff like the UNIX "ls" command, but everyone sort of expects it to be bundled with the OS.
This same problem exists in the DMCA, where it makes no provision for how cryptic a thing must be before running a program to transform it into something human readable counts as an act of "decrpytion". Technically every plain ascii text file is "encrypted" in a code known as ASCII that replaces letters with single-byte numeric codes, and every time you pull it up into a text file viewer you are "decrypting" it into human-readable form. Granted, it's a pretty *BAD* encrpytion algorithm, with many programs out there that can decrypt it, but then again so are a lot of the things that ARE protected by the DMCA, like the DVD CSS algorithm.
I'd much rather have the case where the lawyers *know* they don't know anything than the case where they *think* they do because one of them downloaded and installed a device driver once.
What's noteworthy is that AOL is getting companies to pay AOL to fix a problem AOL created themselves. Pretty sweet deal. Kinda like the Far Side cartoon where a guy gets a brick thrown through his window, and attached to the brick is an advertisement for a window glass repair shop.
If you assume all silly illogical posts are jokes, you are giving posters too much credit.
Like yours?
I try, I really, really try, to like SuSe. It comes with so many packages you can install, and is just very cool ("have a lot of fun"), but they keep fscking up their kernel source that gets installed and that ruins it for me. I've tried SuSe 6, 7, and now 8. Each time their installation gets screwed up whenever I need to make a small change to the kernel config and recompile. The reason, I only just found out recently, is that the kernel source they install on your system is NOT configured the same way as the kernel source from which they made their prebuilt kernels. This means I have to just make a wild guess as to what the settings were for all those config options. With over hundreds of config options to choose from, the probability of my random guess being correct approaches zero. So inevitably whenever I want to change something small (like setting it up to read IDE CD-ROMS under SCSI emulation so I can run burner software), I end up accidentally changing a heck of a lot of settings because I had to guess what the whole list of settings originally were. Suddenly stuff breaks all over the place on the next reboot as drivers can't be found (I configured as a module what had been originally compiled-in). And then I look in the source tree and can't even FIND the drivers existing anywhere in the source for what had obviously been working using the prebuilt binary, so now I know I don't even have the same SOURCE CODE they were using, much less the same config options for that source.
No, it's not. Competition isn't about attacking all competitors how ever they see fit. (Note the emphasis.) It's about doing so withing certain rules of good ethics and fair conduct where you let your products compete on their own merits. This is not the Microsoft business model.