Mozilla is multi threaded, some linux system monitoring tools don't grok multi threaded so it appears that you get number of active threads x process memory allocated. That said the ratio is probably right, althought the actual memory usage is probably a fifth or a sixth of the value quoted. Then again, maybe you have some huge plugins.
NT can exist on the same physical device as 2000. To do it install 2000 near the end of your disk. blank the partition table, install NT at the start of your disk, upgrade to sp6 and then reinstate the partition table. Et voila... (Guess who had to do this recently...)
Previous years have seen valid (although crap) INTERCAL entrants. Though as they say the amazment of the elephant tap dancing is not that he does it well, rather that he does it at all.
Previous years, however cat(1) has been a valid entrant, which is probably how the INTERCAL got done in time.
See that word after X, it is server. Server has (at least) two meanings.
The big box in the computer room that doesn't need graphics (second use of server in your sentence)
Something that a client talks to. (first use of server in your sentence)
Oh my, a very amusing play on words, how funny am I (not very)? Look, you even used the word twice in your sentence! Its like those bits in films when you laugh at a joke, and then they spell out the punchline again. Get some caffeine down you.
I think you'll have a harder time breaking a ring made out of diamond than one made out of titanium. Probably a lot harder to make one in the first place too
England's dot-uk is the fourth largest Internet domain with more than 3.5 million registrations.
I'm pretty sure they do for Scotland, Wales, Northern Ireland and some smaller Islands as well. In fact uk probably stands for United Kingdom, altough like most things United it isn't (I'm sure thats someone's Law)
The -X flag for ssh sets up secure X tunnelling. In the case in point, the cygwin window will be the X server for the host being connected to. Whether or not encryption is really required in DenialS' environment may be more pertinent, but if there is sufficient processor power I guess its useful to get into the habit of using secure networking, even when not strictly necessary.
I guess the thorny problem is that before you can get any interesting feedback based on testers typing they would need to train to use the given keyboard layout effectively.
Maybe some kind of two phase operation could be developed where you have one algorithm similar to the authors, and a second that attempts to modify the fitness function based on human feedback on the layouts generated by the first alogorithm. Perhaps that would reduce the amount of input required by humans.
I think I read that quote slightly differently to you. Subsitute the assertion in the first part of the quote into the second part and you get...
Threads are for people who can't program computers.
What it is not saying is "Threads enable people who cannot program computers to program computers"
Basically I think Alan shares your point of view. That the way I was thinking when I remembered the quote. I certainly agree with you, and try to aviod them wherever possible.
The only place I regularly use them (thus proving that I can't program, damn) is for Java user interfaces. Then its only if there is a moderately long running action, e.g. five seconds for a button. To stop the window becoming frozen with respect to the windowing system during this time a second thread appears to be necessary. I'd certainly be interested to hear of another way of doing it. In *nix style you would separate out the long running part into a separate process and get the second process to signal the first when it is done (effectively the completion becomes part of the same event loop that is processing the window events). I don't think Java gives the necessary infrastructure to do this, though I stand to be corrected.
Slashdot Mirror
Mozilla is multi threaded, some linux system monitoring tools don't grok multi threaded so it appears that you get number of active threads x process memory allocated. That said the ratio is probably right, althought the actual memory usage is probably a fifth or a sixth of the value quoted. Then again, maybe you have some huge plugins.
I wonder if you ever read the Bottom Inspectors cartoon in Viz before. It was a few years ago, so its probably been replaced by now.
1/2 tracks, laser burned discs (I'm waiting for this on CD-ROM!), and 1541 misalignments
. They all seem pretty pointless (asuming C64 is a Commodore 64) when the reset button doesn't clear memory...Or did they get round that somehow?
NT can exist on the same physical device as 2000. To do it install 2000 near the end of your disk. blank the partition table, install NT at the start of your disk, upgrade to sp6 and then reinstate the partition table. Et voila... (Guess who had to do this recently...)
Reminds me of a time we had a guy from Poland working for us. The only time I've ever seen soldered vamps...
One cpu second on their machine (which is the guideline) is a lot more than one second if you are working on a different machine.
There is no mentioin of publically accessable servers either, however...
Previous years have seen valid (although crap) INTERCAL entrants. Though as they say the amazment of the elephant tap dancing is not that he does it well, rather that he does it at all. Previous years, however cat(1) has been a valid entrant, which is probably how the INTERCAL got done in time.
That's _SO_ 1990s
That's so twentieth century. In fifteen years time it will be cutely retro
I payed about GBP100 for a 32kB upgrade for my ZX Spectrum.
16kB ram packs for the ZX81 were pricier still. I'm sure people have even more extreme examples.
See that word after X, it is server. Server has (at least) two meanings.
Oh my, a very amusing play on words, how funny am I (not very)? Look, you even used the word twice in your sentence! Its like those bits in films when you laugh at a joke, and then they spell out the punchline again. Get some caffeine down you.
Except an X server, that is.
Yeah, but that's the at least the second girl he has asked out in this story alone, so he is probably out of luck.
I think you'll have a harder time breaking a ring made out of diamond than one made out of titanium. Probably a lot harder to make one in the first place too
This is equivalent to storing the contents of 7,800 DVDs in one square inch of material.
It may be one inch square in cross section. Unfortuately its twelve light years high. Oh, well.
England's dot-uk is the fourth largest Internet domain with more than 3.5 million registrations.
I'm pretty sure they do for Scotland, Wales, Northern Ireland and some smaller Islands as well. In fact uk probably stands for United Kingdom, altough like most things United it isn't (I'm sure thats someone's Law)
Grumble, grumble, whinge, whinte.
Heck, how many Americans go to another country full stop? I heard one completely unreliable source say it was only 10%
I think there is some scaremongeroing going on here, sure, but I also think that our next 100 years or so is going to see a lot of problems.
A simple extrapolation from the last seven thousand years would indicate that, without worrying about the mechanism too much.
no space on device /dev/hda4
The technology was patently bollocks, as it were.
The -X flag for ssh sets up secure X tunnelling. In the case in point, the cygwin window will be the X server for the host being connected to. Whether or not encryption is really required in DenialS' environment may be more pertinent, but if there is sufficient processor power I guess its useful to get into the habit of using secure networking, even when not strictly necessary.
How about the best keyboard layout for brainf*ck?
<curse> <run>
> < + - . , [ ]
<space> <enter>
Space and enter are probably optional. One finger for each key that you need while programming, cool.
I guess the thorny problem is that before you can get any interesting feedback based on testers typing they would need to train to use the given keyboard layout effectively.
Maybe some kind of two phase operation could be developed where you have one algorithm similar to the authors, and a second that attempts to modify the fitness function based on human feedback on the layouts generated by the first alogorithm. Perhaps that would reduce the amount of input required by humans.
Funny the first line in my .bashrc for cygwin has always been
alias more=less
Its a wonder I get any useful work done.
I think I read that quote slightly differently to you. Subsitute the assertion in the first part of the quote into the second part and you get...
Threads are for people who can't program computers.
What it is not saying is "Threads enable people who cannot program computers to program computers"
Basically I think Alan shares your point of view. That the way I was thinking when I remembered the quote. I certainly agree with you, and try to aviod them wherever possible.
The only place I regularly use them (thus proving that I can't program, damn) is for Java user interfaces. Then its only if there is a moderately long running action, e.g. five seconds for a button. To stop the window becoming frozen with respect to the windowing system during this time a second thread appears to be necessary. I'd certainly be interested to hear of another way of doing it. In *nix style you would separate out the long running part into a separate process and get the second process to signal the first when it is done (effectively the completion becomes part of the same event loop that is processing the window events). I don't think Java gives the necessary infrastructure to do this, though I stand to be corrected.
I'd think you appreciate the quote on threads attributed to Alan Cox on Larry McVoy's page.