If you had only 5 books or 100 books at home, sure. If you only ever read the books that you had at home, sure. But half of the time if I'm browsing through a used bookstore and see a nice book, chances are I've already read it (if I'm willing to buy it in hardback), and I often can't remember if I read a copy of mine, borrowed a friends, or checked a copy out of the library.
I have lots of series that I know I'm missing a book, but when I've seen copies in stores I don't remember if I needed book 2 or 3.
Whether I've read it already is irrelevant, because I have... whether I already own it is a seperate question. Plus, if you're married, your SO may have bought or read the book and you might not have gotten around to it yet.
My wife and I have a nice library at home, I'd guess several thousands of paperbacks, hundreds to a thousand+ hardbacks. I'm sure between us we've read them all (or at least 95%+), but that doesn't mean I've read every one of them.
Personally I'd LOVE to have a DB like that... but no way in hell I'd spend the time it would take to enter everything by hand.
>> I have seen what you're taling about [...] >> Usually it's because there is a zombie process [...]
Hrm. I could buy that if it was a single application (and I can bring up the task manager and look at the process list), but it won't launch *any* application, even one that I haven't run since a reboot (ie, word, quicken, NOTEPAD, anything). Perhaps whatever process w2k uses to do the equivalent of fork/exec is zombied, but it isn't explorer (I've killed explorer w/ task manager, then done a run->explorer and brought the shell back up, but the problem persists).
Still weird. Ah well.
I never did install BeOS 4.5, I found that I was only using Be as a "toy"... wrote a couple of programs for the hell of it, and to check out their api (very cool!) but never used it seriously. Win2k does run great (when it's running) on 128 Mb of RAM, but I haven't tried it on a lesser box. *shrug*
I've been running rc1 for the last couple of months because I'd picked up a cheap 56k modem that win98 would *NOT* recognize. Period, end of story. I tried everything I could think off, but if those two vxd files were in the windows directory, windows would hang, every time.
w2k picked up and ran flawlessly, and didn't take that long to install (although I didn't sit at the computer the whole time). I've been using that computer to run Everquest almost exclusively for the last couple of months.
I am having a problem, though, and I'm curious if others have had the same thing happen. Periodically, (meaning about once every day or two) programs refuse to execute. You can double click, you can click on a shortcut, anything, and nothing will run. You get the hourglass for a fraction of a second and then nothing. If it were unix, I'd say the fork failed w/o warning.
My solution is to reboot, which of course means that my windows box has an average uptime of probably 18 hours. Linux, of course, sits next to it happily chugging away.:)
Anyone else see this symptom? Jon
Re:Security 101... Not offered on campus.
on
Linux Lite?
·
· Score: 1
Very cool, this is exactly what I was looking for! Thanks a million...
Jon
Re:Security 101... Not offered on campus.
on
Linux Lite?
·
· Score: 1
OK, so as someone recently basking in the glow of his cable modem's little green lights, here's my question:
Any pointers to consolidated information on steps to take to secure your box? I've done some dinking around looking for info, and have certainly found some, but a single, coherent site would be marvelous.
I find it ironic that now that I'm really "on" I'm now more worried about my linux box than my windows box! The reason, of course, is that I don't do *anything* (I think Quicken and Everquest are the only two progs I've installed since I installed NT5rc1) major on Windows... whereas Linux is running telnet/ftp/http services. But I'm not willing to give up *any* of these. One of my major pleasures is [going to be] finally being able to read my home mail through the firewall at work (socksified telnet) without having to use a web based email service.
I'm not exactly a newbie (Yggdrasil anyone?), but I've never had to worry about it before since Linux has been pretty much strictly recreational, and not connected.
Allow me to harken back to the halcyon days of Graduate Computer Architecture.
VLIW stands for "Very Long Instruction Word" and basically just concatenates several smaller instructions into one humongous one. In a simple example, consider a chip that has 2 int and 1 fp ALUs (Arithematic logic unit). Today's chip would have to find independant ops and either queue them or stagger the start through the ALUs. VLIW moves the process of finding independent operations into the compile stage, allowing the compiler to determine and "schedule" operations that can execute in parallel. At least in theory, you can get quite a bit of speedup. IF (Big If) you can find enough "parallelism" to be sending mostly full instructions through. You'll be spinning a lot of wheels if you have one add and three noops in each instruction word.;)
A couple of years ago I saw a printout of what the "compiled" (probably was simulated at the time) results were from the HP VLIW compiler, and what it basically looked like was:
Essentially, all the independent operations would get moved up and shoved into the first few instructions, then there would be a (possibly long) series of instructions that had to complete in sequence. Rinse, repeat.
Theoretically, it should be a win, if they can figure out better ways to find/create/massage instructions that can be executed simultaneously. Of course, I haven't been paying attention for the last couple of years, maybe they've worked out some good solutions.
I haven't seen this point elsewhere, what about the theatre turning away someone *legally* of age (17) with an arbitrary policy of "nobody under the age of 18 can see Eyes Wide Shut."
Don't think so. Says "Reflective CMOS"... TI uses a teeter-totter type system, where they turn "pixels" on and off by biasing the "board" one way or the other, and changing the reflection vector.
If this is anything like DisplayTech's stuff, it's a LCD layer on top of a conventional chip that has a silvered or highly reflective layer just under the LCD. Then by turning each LCD pixel on/off they can change whether light is absorbed/scattered or reflected back (by the "mirror" under the LCD).
Then you can either put one of these at the other end of a colorwheel, and index your primaries, relying on the eye to merge the images, or you can take three of them, and use a special prism (Philips?) to separate out and recombine the RGB images.
Pretty cool technology. But it'll never be used on a "large" scale... like the paper thread. Silicon's too expensive! But projection now...
What I really want is my $300 HDTV projector. I know that DisplayTech has 1600x1200 (at least in the proto stage, although all I saw working was an 800x600). That's not too much lower than the 19??x???? needed for HDTV.
Slashdot Mirror
If you had only 5 books or 100 books at home, sure. If you only ever read the books that you had at home, sure. But half of the time if I'm browsing through a used bookstore and see a nice book, chances are I've already read it (if I'm willing to buy it in hardback), and I often can't remember if I read a copy of mine, borrowed a friends, or checked a copy out of the library.
I have lots of series that I know I'm missing a book, but when I've seen copies in stores I don't remember if I needed book 2 or 3.
Whether I've read it already is irrelevant, because I have... whether I already own it is a seperate question. Plus, if you're married, your SO may have bought or read the book and you might not have gotten around to it yet.
My wife and I have a nice library at home, I'd guess several thousands of paperbacks, hundreds to a thousand+ hardbacks. I'm sure between us we've read them all (or at least 95%+), but that doesn't mean I've read every one of them.
Personally I'd LOVE to have a DB like that... but no way in hell I'd spend the time it would take to enter everything by hand.
Jon
>> I have seen what you're taling about
... wrote a couple of programs for the hell of it, and to check out their api (very cool!) but never used it seriously. Win2k does run great (when it's running) on 128 Mb of RAM, but I haven't tried it on a lesser box. *shrug*
[...]
>> Usually it's because there is a zombie process
[...]
Hrm. I could buy that if it was a single application (and I can bring up the task manager and look at the process list), but it won't launch *any* application, even one that I haven't run since a reboot (ie, word, quicken, NOTEPAD, anything). Perhaps whatever process w2k uses to do the equivalent of fork/exec is zombied, but it isn't explorer (I've killed explorer w/ task manager, then done a run->explorer and brought the shell back up, but the problem persists).
Still weird. Ah well.
I never did install BeOS 4.5, I found that I was only using Be as a "toy"
Later,
Jon
I've been running rc1 for the last couple of months because I'd picked up a cheap 56k modem that win98 would *NOT* recognize. Period, end of story. I tried everything I could think off, but if those two vxd files were in the windows directory, windows would hang, every time.
:)
w2k picked up and ran flawlessly, and didn't take that long to install (although I didn't sit at the computer the whole time). I've been using that computer to run Everquest almost exclusively for the last couple of months.
I am having a problem, though, and I'm curious if others have had the same thing happen. Periodically, (meaning about once every day or two) programs refuse to execute. You can double click, you can click on a shortcut, anything, and nothing will run. You get the hourglass for a fraction of a second and then nothing. If it were unix, I'd say the fork failed w/o warning.
My solution is to reboot, which of course means that my windows box has an average uptime of probably 18 hours. Linux, of course, sits next to it happily chugging away.
Anyone else see this symptom?
Jon
Very cool, this is exactly what I was looking for! Thanks a million...
Jon
OK, so as someone recently basking in the glow of his cable modem's little green lights, here's my question:
Any pointers to consolidated information on steps to take to secure your box? I've done some dinking around looking for info, and have certainly found some, but a single, coherent site would be marvelous.
I find it ironic that now that I'm really "on" I'm now more worried about my linux box than my windows box! The reason, of course, is that I don't do *anything* (I think Quicken and Everquest are the only two progs I've installed since I installed NT5rc1) major on Windows... whereas Linux is running telnet/ftp/http services. But I'm not willing to give up *any* of these. One of my major pleasures is [going to be] finally being able to read my home mail through the firewall at work (socksified telnet) without having to use a web based email service.
I'm not exactly a newbie (Yggdrasil anyone?), but I've never had to worry about it before since Linux has been pretty much strictly recreational, and not connected.
Anyway, point is, pointers anyone?
Jon
Allow me to harken back to the halcyon days of Graduate Computer Architecture.
;)
VLIW stands for "Very Long Instruction Word" and basically just concatenates several smaller instructions into one humongous one. In a simple example, consider a chip that has 2 int and 1 fp ALUs (Arithematic logic unit). Today's chip would have to find independant ops and either queue them or stagger the start through the ALUs. VLIW moves the process of finding independent operations into the compile stage, allowing the compiler to determine and "schedule" operations that can execute in parallel. At least in theory, you can get quite a bit of speedup. IF (Big If) you can find enough "parallelism" to be sending mostly full instructions through. You'll be spinning a lot of wheels if you have one add and three noops in each instruction word.
A couple of years ago I saw a printout of what the "compiled" (probably was simulated at the time) results were from the HP VLIW compiler, and what it basically looked like was:
-----
-----
-----
---
--
-
-
-
-
-----
-----
----
...
Essentially, all the independent operations would get moved up and shoved into the first few instructions, then there would be a (possibly long) series of instructions that had to complete in sequence. Rinse, repeat.
Theoretically, it should be a win, if they can figure out better ways to find/create/massage instructions that can be executed simultaneously. Of course, I haven't been paying attention for the last couple of years, maybe they've worked out some good solutions.
Who knows...
Jon
I haven't seen this point elsewhere, what about the theatre turning away someone *legally* of age (17) with an arbitrary policy of "nobody under the age of 18 can see Eyes Wide Shut."
How can that be legal?
Jon
Aaah, tradewars.
;)
*sigh*
Those were the days.
If this is anything like DisplayTech's stuff, it's a LCD layer on top of a conventional chip that has a silvered or highly reflective layer just under the LCD. Then by turning each LCD pixel on/off they can change whether light is absorbed/scattered or reflected back (by the "mirror" under the LCD).
Then you can either put one of these at the other end of a colorwheel, and index your primaries, relying on the eye to merge the images, or you can take three of them, and use a special prism (Philips?) to separate out and recombine the RGB images.
Pretty cool technology. But it'll never be used on a "large" scale... like the paper thread. Silicon's too expensive! But projection now...
What I really want is my $300 HDTV projector. I know that DisplayTech has 1600x1200 (at least in the proto stage, although all I saw working was an 800x600). That's not too much lower than the 19??x???? needed for HDTV.
Later,
Jon
Yikes-o-rama. Any programmer willing to work for those rates deserves what (s)he gets, I guess.
Not totally sure, but I'm guessing starting salaries for CS-types around here (NW USA) are:
No degree: 30-40k
BS: 40-50k
MS: 50-60k
PhD: WTH knows.