Do a search on Yahoo and Google and you'll see the same results.
No, the results are quite often different. Yes, I know that Yahoo licenses Google's technology, but Google.com still returns different (and generally better, IMHO) results.
It is ironic that the biggest Usenet archive is now owned by a company that indexes the Web.
Ironic in an Alanis Morissette sense, or an Oxford English Dictionary sense?
AWT isn't really slow. Swing is somewhat slow, but plain AWT is actually quite quick. The problem is that most people who code directly to AWT don't know what they're doing, and they overide paint() but not update(). Check out the applets at Java on the Brain to see some good AWT-intensive Java applets.
In fact, all the good OSes make testing malloc() for 0 effectively redundant. You simply have all the memory you want or something is horribly wrong and the machine needs to be reset.
That would only be true if you had unlimited swap space.
From my experience, Sawfish is a lot faster than E, even when using the same theme and options. I had used E for several months, and one day I decided to try out Sawfish (back when it was called Sawmill). I stopped using E, because with sawfish I had exactly the same theme, but windows rendered much faster, and it has a real configuration language.
Okay, so E isn't slower than Sawfish as long as you use a crappy theme. Perhaps. Why should I have to put up with a crappy theme though? With Sawfish, I can use a theme that looks exactly like the one I used to use in E, and it's much faster.
Plus, if I need to do some really complicated modifications to my setup or my theme, I can use a real language (rep, a Lisp dialect), rather than the hacked up C preprocessor kludge that E uses.
Most of the people who complain about GNOME being slugggish are actually complaining about E. GNOME with Sawfish is actually fairly quick. I've enev got an old Pentium 133 with GNOME+Sawfish on it, and it runs fine. I tried running E on the same machine, with the E version of the same theme, and it was unbearably slow. (and yes, I turned off animations and the snapshot pager)
Remember: you have to fight copyrights at all times, lest you lose them; you can let patents and trademarks go undefended and fight them at the very end without problems from the legal side
No, patents and copyrights don't need to be defended. Trademarks do.
I am sick and tired of everyone saying, "We should do X, so newbies won't have to learn Y". I am happy if people want to use GNU/Linux, but if it means having to dumb down everything, then what is the point anymore?
So why do you call it "GNU/Linux". Wasn't RMS's excuse that he wants newbies to know about GNU? So isn't calling is "GNU/Linux" really "dumbing down" the name so that newbies won't have to spend some time using Linux so they can find out for themselves that some of the command line utils and libc are GNU?
I think Windows has less of this problem simply because most of the software used on Windows boxes is mostly self-contained commercial software. DLL hell on Windows boxes usually results from a small number of DLLs... usually VB or VC++ DLLs, actually. Those are the only ones that "other programs" generally depend upon. And sometimes it really can lead to a nasty mess. I've seen programs that would only work with one version of the MSVCRT.DLL. They'd become unstable with a later version, so if you install some other program that upgrades that DLL, you're screwed. (and of course, most Windows installers silently upgrade DLLs, and there's often no easy way to examine an installer to see what it'll install, because the files are generally hidden away in some sort of weird proprietary archive format)
The first monolith in 2001 was transparent, and later became black. It was only black in the movie because it wasn't feasible to haul around a 20' tall slab of glass...
Incidently, something that reflects no light is black, not nearly invisible... unless it's on a perfectly black background. When the monoliths were black, they were supposed to reflect no light, but again, that wasn't feasible for the movies. (and it probably would've been confusing for viewers too... there would be fewer depth cues due to the lack of different shading)
as if we just found it, on a remote island somewhere? before that everyone was just drifting off into space...
[I agree with your point about computers, but I wouldn't call "gravity" a discovery either. The way in which gravity works and the mathematical relationship between mass, distance and gravity, that was discovered. Gravity itself wasn't discovered any more than the ground was.]
That's like saying people shouldn't be allowed to look at the windows on your house because you don't have curtains, and your SO won't let you install any.
No, it's a bug in MKS Toolkit. Yes, you're right that the FAT filesystem is limited in that it's case insensitive (but case preserving). That isn't an excuse for the behavior of MKS ksh though. It should find the largest common prefix in a case-insensitive way. It doesn't. Even the cmd.exe file completion at least got that part of it right. (it does, unfortunately, get a whole bunch of other things wrong, like the fact that it replaces what you typed with a full absolute pathname).
While the MKS Toolkit is certainly much better than cygwin, it does have bugs. The most irritating bug I know of:
touch foobar1
touch FooBar2
Then type cat foo and try to use filename completion. It should give you "foobar" or perhaps even "FooBar". Instead, it completely wipes out the beginning of your filename! I think what's happening is:
- it notices that foobar1 and FooBar2 both start with "foo" (case-insensitive)
- it tries to find the longest common prefix, but it does this in a case-sensitive way. So the result is the empty string.
- it replaces what you typed with the (empty) common prefix
This has been there since 5.x, and I know it's still there in 6.1. I don't know if they finally fixed this in 7.0.
(incidently, the #1 reason why MKS Toolkit is better than cygwin is that it uses a pathname convention that's compatible with other Win32 tools. Cygwin doesn't play nice with non-cygwin executable, like VIM, for instance.)
I think you're thinking of Puerto Rico, or the US Virgin Islands. I'm pretty sure Hawaii (like Alaska) is a "real state". FWIW, hawaii.gov Calls Hawaii the "Aloha state".
Right now I'm seeing a banner ad on Slashdot that has the read-ends of a bunch of cartoon animals, with the punchline "What does your backend look like?" Given the place where they're advertising, they really should've had the goatse.cx "back-end" in the ad too...
In most civilized countries, you can't patent something one it's been published, even if you're the one who published it. I'm not sure if this is true for the US now (it wasn't back when RSA was patented, but that was a while ago...).
Anyway, releasing GPLed code is "publishing" something, for all intents and purposes. So if the US does have a provision against patenting things that have already been published, then the patent should be invalid for that reason.
Slashdot Mirror
Do a search on Yahoo and Google and you'll see the same results.
No, the results are quite often different. Yes, I know that Yahoo licenses Google's technology, but Google.com still returns different (and generally better, IMHO) results.
It is ironic that the biggest Usenet archive is now owned by a company that indexes the Web.
Ironic in an Alanis Morissette sense, or an Oxford English Dictionary sense?
Then that version of Windows should also refuse to install itself, using Windows' history of security holes as an excuse.
AWT isn't really slow. Swing is somewhat slow, but plain AWT is actually quite quick. The problem is that most people who code directly to AWT don't know what they're doing, and they overide paint() but not update(). Check out the applets at Java on the Brain to see some good AWT-intensive Java applets.
In fact, all the good OSes make testing malloc() for 0 effectively redundant. You simply have all the memory you want or something is horribly wrong and the machine needs to be reset.
That would only be true if you had unlimited swap space.
Out of curiosity, what is it you hate about sawfish?
From my experience, Sawfish is a lot faster than E, even when using the same theme and options. I had used E for several months, and one day I decided to try out Sawfish (back when it was called Sawmill). I stopped using E, because with sawfish I had exactly the same theme, but windows rendered much faster, and it has a real configuration language.
Okay, so E isn't slower than Sawfish as long as you use a crappy theme. Perhaps. Why should I have to put up with a crappy theme though? With Sawfish, I can use a theme that looks exactly like the one I used to use in E, and it's much faster.
Plus, if I need to do some really complicated modifications to my setup or my theme, I can use a real language (rep, a Lisp dialect), rather than the hacked up C preprocessor kludge that E uses.
Most of the people who complain about GNOME being slugggish are actually complaining about E. GNOME with Sawfish is actually fairly quick. I've enev got an old Pentium 133 with GNOME+Sawfish on it, and it runs fine. I tried running E on the same machine, with the E version of the same theme, and it was unbearably slow. (and yes, I turned off animations and the snapshot pager)
Remember: you have to fight copyrights at all times, lest you lose them; you can let patents and trademarks go undefended and fight them at the very end without problems from the legal side
No, patents and copyrights don't need to be defended. Trademarks do.
I am sick and tired of everyone saying, "We should do X, so newbies won't have to learn Y". I am happy if people want to use GNU/Linux, but if it means having to dumb down everything, then what is the point anymore?
So why do you call it "GNU/Linux". Wasn't RMS's excuse that he wants newbies to know about GNU? So isn't calling is "GNU/Linux" really "dumbing down" the name so that newbies won't have to spend some time using Linux so they can find out for themselves that some of the command line utils and libc are GNU?
I think Windows has less of this problem simply because most of the software used on Windows boxes is mostly self-contained commercial software. DLL hell on Windows boxes usually results from a small number of DLLs... usually VB or VC++ DLLs, actually. Those are the only ones that "other programs" generally depend upon. And sometimes it really can lead to a nasty mess. I've seen programs that would only work with one version of the MSVCRT.DLL. They'd become unstable with a later version, so if you install some other program that upgrades that DLL, you're screwed. (and of course, most Windows installers silently upgrade DLLs, and there's often no easy way to examine an installer to see what it'll install, because the files are generally hidden away in some sort of weird proprietary archive format)
You mentioned that you'll only be able to download packages that are signed?
I think he said "be able to only", not "only be able to". Also, he said "install", not "download"...
The first monolith in 2001 was transparent, and later became black. It was only black in the movie because it wasn't feasible to haul around a 20' tall slab of glass...
Incidently, something that reflects no light is black, not nearly invisible... unless it's on a perfectly black background. When the monoliths were black, they were supposed to reflect no light, but again, that wasn't feasible for the movies. (and it probably would've been confusing for viewers too... there would be fewer depth cues due to the lack of different shading)
discovery of gravity?
as if we just found it, on a remote island somewhere? before that everyone was just drifting off into space...
[I agree with your point about computers, but I wouldn't call "gravity" a discovery either. The way in which gravity works and the mathematical relationship between mass, distance and gravity, that was discovered. Gravity itself wasn't discovered any more than the ground was.]
Biotech = Controlling genes and DNA (ala Galactica).
I think you mean Gattaca. Note that the name only contains the letters G, A, T and C.
That's like saying people shouldn't be allowed to look at the windows on your house because you don't have curtains, and your SO won't let you install any.
I don't think you've been trying very hard. Did you try doing a search for "algorithms" on a bookstore's web site? Introduction to Algorithms is pretty popular. As for how to structure code, Code Complete is pretty good, as is Design Patterns. There are also numerous other domain-specific language independant books like Computer Graphics: Principles and Practice (while it does have code snippets, you can effectively treat them as pseudo code), Artificial Intelligence: A Modern Approach, and Database System Implementation.
This post is making me wish I had an Amazon affiliate ID...
[If any of the above links are busted, it's probably Slash's "long line breaking" bug. Remove the spaces in the URL, and try again...]
No, it's a bug in MKS Toolkit. Yes, you're right that the FAT filesystem is limited in that it's case insensitive (but case preserving). That isn't an excuse for the behavior of MKS ksh though. It should find the largest common prefix in a case-insensitive way. It doesn't. Even the cmd.exe file completion at least got that part of it right. (it does, unfortunately, get a whole bunch of other things wrong, like the fact that it replaces what you typed with a full absolute pathname).
While the MKS Toolkit is certainly much better than cygwin, it does have bugs. The most irritating bug I know of:
touch foobar1
touch FooBar2
Then type cat foo and try to use filename completion. It should give you "foobar" or perhaps even "FooBar". Instead, it completely wipes out the beginning of your filename! I think what's happening is:
- it notices that foobar1 and FooBar2 both start with "foo" (case-insensitive)
- it tries to find the longest common prefix, but it does this in a case-sensitive way. So the result is the empty string.
- it replaces what you typed with the (empty) common prefix
This has been there since 5.x, and I know it's still there in 6.1. I don't know if they finally fixed this in 7.0.
(incidently, the #1 reason why MKS Toolkit is better than cygwin is that it uses a pathname convention that's compatible with other Win32 tools. Cygwin doesn't play nice with non-cygwin executable, like VIM, for instance.)
I think you're thinking of Puerto Rico, or the US Virgin Islands. I'm pretty sure Hawaii (like Alaska) is a "real state". FWIW, hawaii.gov Calls Hawaii the "Aloha state".
Google says it's Kramer.
Right now I'm seeing a banner ad on Slashdot that has the read-ends of a bunch of cartoon animals, with the punchline "What does your backend look like?" Given the place where they're advertising, they really should've had the goatse.cx "back-end" in the ad too...
Tell that to the Hawaiians.
Why do we need more phone numbers for ISPs? I don't use a modem for my net connection.
I don't believe I've ever heard Dennis Hopper say anything that sounded even remotely like a backspace.
In most civilized countries, you can't patent something one it's been published, even if you're the one who published it. I'm not sure if this is true for the US now (it wasn't back when RSA was patented, but that was a while ago...).
Anyway, releasing GPLed code is "publishing" something, for all intents and purposes. So if the US does have a provision against patenting things that have already been published, then the patent should be invalid for that reason.
ObIANAL: IANAL