However, I'm not aiming for HIGH, but PREDICTABLE performance.
You take a look at a loop and you'll roughly know how long it's going to take to run for C++ without GC.
Until Java or C# supports true destructors (i.e. I should be able to tell the GC that I want to delete this object whenever I want, not the GC want), a lot of people, myself included, will not use it for systems programming.
Disclaimer: I'm a happy user of Slackware for years.
Still, I'm going to say I don't expect Linux to take over the Desktop anytime soon. The problem? Choice. Choice is good, but excess of anything is bad - even choice.
If the only distribution were MEPIS, I agree Linux is a viable option - even with hundreds of desktop-oriented distributions, it might still be possible if everyone and his dog is not reinventing the wheel by writing their "minimalist" editor and all distributions choosing a different default.
Linux is a nightmare to support. On Windows, you're almost 100% sure your user will have a clean, easy-to-work-with editor, aka Notepad. On Linux, you'll have to ask - open a text editor...do you have Gvim? Nope? How about Nedit? Nope? Xemacs? Nope? How about xedit - yeah, that's it...with its archic and ugly interface. You get the idea. Too many choices WITHOUT A DEFAULT is a problem.
Also, higher assumptions are difficult to make when you develop software (apart from POSIX, you almost cannot assume anything) - which version of libc do you use? On Windows it does not matter - for example, msvcrt.dll just seems to be better at being backwards compatible than glibc - just rename the library and it'll most likely work. With glibc, it'll most likely complain - I know it is in theory a better safeguard - but I don't want to care if all I want to do is to run a browser.
Linux will be viable for the Desktop on the day all distributions agree on a set of defaults for GUI apps, which unfortunately LSB does not cover.
And for some reason I don't think that day is coming soon.
Section 202 -- Gives federal officials the authority to intercept wire, oral, and electronic communications relating to computer fraud and abuse offenses.
Seems that national security interests are already covered by Section 201. Section 202 smells like something totally irrelevant to counterterrorism at all.
Think about it - depending on how they interpret "computer abuse", Section 201 is practially a ticket to arbitrary wiretap.
One way to success, is to commoditize your complement.
If the RIAA senses that more of their future income might come from online sales than from the record stores, it only makes sense to make the online stores a commodity.
I believe they're showing that they're not working for the money but for the passion in their work.
It's like saying, "Competitors - I'm now working not for the money but for the sheer joy of it - it implies I must do a better job than you, so PHEAR!!"
I've seen a lot of posts asking for "complete documentation" of a piece of hardware.
Basically, you're just asking for the hardware maker to back away.
Hardware makers often make use of 3rd-party technologies. Making documentation open might sound easy, but is legally difficult.
There are trade secrets to protect (if you don't choose to patent a novel design, you do the best that you can to keep it secret).
(for those of you who think there shouldn't be a reason for trade secrets and only evil corporations use it - think again - trade secrets also protect small companies from the big fishes, too)
The solution I can see is a public binary ABI for drivers that is *STABLE* (i.e. not completely revised every other major kernel release).
I'm a Linux user. In my short-sighted opinion, maintaining binary compatibility is the biggest weakness of all open source operating systems.
There should be a guideline - that, say, binaries compiled within a major versino of GLIBC should all at least load - with various levels of buginess.
It's fine if it *recommends* a version. But if it *requires* a version there might be something wrong - I was furious to see a binary that refused to load on a GLIBC 2.3.2 box just because it was compiled to link with GLIBC 2.3.3.
No matter what you do, replacing a perfectly-working X is just crazy.
I went through the painful experience of installing Dropline Gnome in Slack 9, and then ending up having to uninstall it when I swaret to Slack 10 - its X is not compatible with the Dropline X.
Dropline should be modified to make use of whatever X version Slackware is supplied with.
FWIW, I've been using a D70 and a ist DS for some time now and I prefer the Pentax over the Nikon for it's lack of Moire pattern, and a generally more natural image.
Sharpness in DSLR depends more on the lens than anything else. If image is sharper because of software, it's called "in camera sharpening", something better done with the Gimp.
There was Gnome 1. Then it was Pango. I thought different Gnome-native IME would be implemented (some already have been) but to this date a lot of people still are waiting for the CJK IME...especially since other input modules don't have good integration and sometimes just don't work well at all.
Once you can load a program and run it from a memory card, what prevents people from making interfaces that you can read from the GC discs and load huge programs through the memory card port?
Since the memory card port is likely something standard in a different form factor (e.g. USB, IDE, etc.) making interfaces to large storage devices are not far-fetched at all.
And try to be the first one to distribute contents there and find ways to prevent independent artists from using it - after all, independent artists is a bigger potential threat than piracy...
Slashdot Mirror
Indeed, the reason I use C++ is performance.
However, I'm not aiming for HIGH, but PREDICTABLE performance.
You take a look at a loop and you'll roughly know how long it's going to take to run for C++ without GC.
Until Java or C# supports true destructors (i.e. I should be able to tell the GC that I want to delete this object whenever I want, not the GC want), a lot of people, myself included, will not use it for systems programming.
Has little to do with annoyance.
It has to do with the phone being close to too many base stations, loading them up with traffic.
The Powerbook duo was HUGE.
By "PDA", I mean...PDA-sized portable device, which should be at most as big as a Blackberry...
I'm still waiting for a PDA that when plugged into a docking station, will become a full-fledged PC.
Someone MUST have done it right. Come on, it's 2005.
Indeed. The Bush administration has formed one of the most ridiculous government regarding the respect to scientists.
Twisting scientific results to fit their political, Christian fundamentalist propaganda.
Example: Selectively provide funding only to researches that they like (stem cell research, abstinence, global warming), and numerous others...
I now perceive the US government as the most successful "Cargo Cult" ever existed.
Disclaimer: I'm a happy user of Slackware for years.
Still, I'm going to say I don't expect Linux to take over the Desktop anytime soon. The problem? Choice. Choice is good, but excess of anything is bad - even choice.
If the only distribution were MEPIS, I agree Linux is a viable option - even with hundreds of desktop-oriented distributions, it might still be possible if everyone and his dog is not reinventing the wheel by writing their "minimalist" editor and all distributions choosing a different default.
Linux is a nightmare to support. On Windows, you're almost 100% sure your user will have a clean, easy-to-work-with editor, aka Notepad. On Linux, you'll have to ask - open a text editor...do you have Gvim? Nope? How about Nedit? Nope? Xemacs? Nope? How about xedit - yeah, that's it...with its archic and ugly interface. You get the idea. Too many choices WITHOUT A DEFAULT is a problem.
Also, higher assumptions are difficult to make when you develop software (apart from POSIX, you almost cannot assume anything) - which version of libc do you use? On Windows it does not matter - for example, msvcrt.dll just seems to be better at being backwards compatible than glibc - just rename the library and it'll most likely work. With glibc, it'll most likely complain - I know it is in theory a better safeguard - but I don't want to care if all I want to do is to run a browser.
Linux will be viable for the Desktop on the day all distributions agree on a set of defaults for GUI apps, which unfortunately LSB does not cover.
And for some reason I don't think that day is coming soon.
but can you bring it to the washroom?
Deathblow of what? Even without backwards compatibility and FF, the PS3 can still hold its own against the XBox 360.
Section 202 -- Gives federal officials the authority to intercept wire, oral, and electronic communications relating to computer fraud and abuse offenses.
Seems that national security interests are already covered by Section 201. Section 202 smells like something totally irrelevant to counterterrorism at all.
Think about it - depending on how they interpret "computer abuse", Section 201 is practially a ticket to arbitrary wiretap.
One way to success, is to commoditize your complement.
If the RIAA senses that more of their future income might come from online sales than from the record stores, it only makes sense to make the online stores a commodity.
I believe they're showing that they're not working for the money but for the passion in their work.
It's like saying, "Competitors - I'm now working not for the money but for the sheer joy of it - it implies I must do a better job than you, so PHEAR!!"
If this is true, then the agreement sucks.
At the very least, it must include an admission of guilt and a formal apology, or some form of other punitive measures.
It seems they can *totally* get away with it now, and nobody will even know they did something wrong.
Don't let it happen.
It smells more like a half-assed effort to calm down legal threats than anything.
The moment you see their words "popular demand" you know they're STILL trying to lie and get away with something.
I run a script that uses xnest to run Gimp in a its own X. The result looks much like an "MDI-like" environment, very clean, very little distraction.
After all, it's their network, it's their rules.
Look for an alternate network, or sign up for DSL with the phone line.
If you don't have a choice because they are the exclusive provider, fight for alternate network access.
Lack of an alternative is the root of liberal use of censorship in school networks.
Have they fixed the memory leaks in 1.0 yet?
Having an idle browser leaking 300 MB of memory per day is like a self-inflicted DOS attack.
I've seen a lot of posts asking for "complete documentation" of a piece of hardware.
Basically, you're just asking for the hardware maker to back away.
Hardware makers often make use of 3rd-party technologies. Making documentation open might sound easy, but is legally difficult.
There are trade secrets to protect (if you don't choose to patent a novel design, you do the best that you can to keep it secret).
(for those of you who think there shouldn't be a reason for trade secrets and only evil corporations use it - think again - trade secrets also protect small companies from the big fishes, too)
The solution I can see is a public binary ABI for drivers that is *STABLE* (i.e. not completely revised every other major kernel release).
I'm a Linux user. In my short-sighted opinion, maintaining binary compatibility is the biggest weakness of all open source operating systems.
There should be a guideline - that, say, binaries compiled within a major versino of GLIBC should all at least load - with various levels of buginess.
It's fine if it *recommends* a version. But if it *requires* a version there might be something wrong - I was furious to see a binary that refused to load on a GLIBC 2.3.2 box just because it was compiled to link with GLIBC 2.3.3.
How about X?
No matter what you do, replacing a perfectly-working X is just crazy.
I went through the painful experience of installing Dropline Gnome in Slack 9, and then ending up having to uninstall it when I swaret to Slack 10 - its X is not compatible with the Dropline X.
Dropline should be modified to make use of whatever X version Slackware is supplied with.
FWIW, I've been using a D70 and a ist DS for some time now and I prefer the Pentax over the Nikon for it's lack of Moire pattern, and a generally more natural image.
Sharpness in DSLR depends more on the lens than anything else. If image is sharper because of software, it's called "in camera sharpening", something better done with the Gimp.
There was Gnome 1. Then it was Pango. I thought different Gnome-native IME would be implemented (some already have been) but to this date a lot of people still are waiting for the CJK IME...especially since other input modules don't have good integration and sometimes just don't work well at all.
Can anyone shine some light on it?
He said "it is the right and responsibility" to know where they're going and how fast.
Gimme a break. If you need to know these things about your kids, there's something wrong how you brought them up.
It's too late to track them.
Could make a valuable textbook example of how not to manage a company...
Once you can load a program and run it from a memory card, what prevents people from making interfaces that you can read from the GC discs and load huge programs through the memory card port?
Since the memory card port is likely something standard in a different form factor (e.g. USB, IDE, etc.) making interfaces to large storage devices are not far-fetched at all.
The treaty that US is interested in:
Banning stem cell research.
How enlightened.
And try to be the first one to distribute contents there and find ways to prevent independent artists from using it - after all, independent artists is a bigger potential threat than piracy...