Um, system extensions were little more than shared libraries. A feature every operating system since the 70s has had. The only difference was that apple showed you an icon as it loaded them in.
Actually it has everything to do with the twist in the wire. If you twist a +/- wire pair together, the signals they put out are cancelled some. The tighter the twist the higher frequency is cancelled. I'm not sure the exact reason, but varying the twist along the length of cable could be to remove any resonance.
For most of those cases, app servers have completely replaced CGI. The problem with CGI is lack of persistence, and process startup costs. These days, any decent web app is written with some persistent layer that interacts directly with the web server, and loads code snipets of its own language, starts objects, calls callback routines or something similar to handle requests. mod_perl, J2EE/JSP, PHP, Zope, and Ruby on Rails are all examples of this.
The mindless force of nature is quite effective at differentiating what's interesting and what isn't. The "interesting" changes survive. By interesting we mean more capable of survival than what came before. If you take a colony of bacteria and let it sit around, tons of mutations happen. Each set of offspring is slightly different than its parents. Some of those changes are drastic and some aren't. The ones that aren't, don't really change much. Of the ones that are, given that they are random, most kill the reciever. Then every once in a while one is beneficial. That bacterium is more suited to survive than the rest, and so reproduces more. The resulting group is more suited to reproduce more, and so does. Soon the colony is taken over by this change. The colony has evolved.
You keep saying that randomness doesn't evolve, but you've obviously never seen the Game of Life. Go download it sometime. A bunch of random numbers are piped on screen and set to evolve. Soon patterns emerge from it. Its random stuff that's happening, but in any random set, some patterns are self propogating. Since non-propogating patterns die out as soon as they appear, the whole set if left long enough will eventually fall into self propogating patterns. Life is a self propogating pattern. In a random universe life is inevitable.
Bullshit. The only people in my life I've ever heard that come from are the people who refuse to do their time with the grunt work, and are waiting around, leeching off the people around them, waiting for their "big chance" to come through so they can do interesting work like the assume people like me have always done. I'm not going to bother telling you how much grunt work I (and everyone around me who is doing interesting things) I had to do to get where I am.
That was SOMEWHAT of the case with Win9x. What a lot of people don't realize is that NT and later actually have a display server. Its just very well hidden.
Ok, something I always wondered why no one's done. Yes for most things, unix domain sockets are fast enough. But the X-SHM extension is MUCH faster for some operations (that mostly involve sending lots of pixmaps). But an application has to be written to use it. Well, we have an abstraction layer here that could dynamically do this, it's called X-lib. Why has no one written a modified X-lib that detects if you're connecting directly and automatically converts some operations to X-SHM?
Editor utility functions are different from UI and environment plugins. For quick and dirty write-these-lines-for-me, syntax hilighting rules, common task macros, emacs can't be beat (I sure as hell can't live without it). But for managing projects, integrating whole other apps (like debuggers, web browsers, class browsers, db interfaces) emacs doesn't hold a candle to eclipse. The thing I have the most hope for is the plugin that uses emacs as eclipse's editor using gnuserv. That is the one thing tempting me from moving from pure emacs.
That's because the only people using java use it server side. JSP, Weblogic, and J2EE are pretty much the way web dev is done these days, both on and off windows.
I realize alot of people see this game initially, realize the controls for each character are the same and assume it's a simple game with no depth. 90% they've never tried to do a smash move (or decided they don't work), and haven't realized that even though the controls for each characters moves are the same, the end results are subtly (or in some cases drastically) different for each one. And the differences aren't just animations like in some fighting games. If you see a character swing his fist in a 23 degree arc over his head, that's exactly how much of a hit area they have. Then you take in the fact that each of these moves behaves completely differently with any of the MANY items in the game, and every character reacts to these items differently. Lastly you throw in the fact that unlike every other fighter, your end goal is to knock them out of the ring, not do them damage (damage just makes them fly farther with each hit), and that every move in the game has a balance between how far it throws and how much damage it does (which sometimes even varies based on timing or placement), and the vast number of recovery moves in the game, the complex, dynamic levels, and you start to realize the depth.
The game is the kind of genius that nintendo is famous for. There is no learning curve. Special moves are a button and a direction. Anyone can pick it up and play it. But you can keep learning things about this game up through the 2300 versus matches the thing's stat screen says I've played.
You could look at it as ptr not being a boolean type, hense ! not making sense. I on the other hand see ptr == NULL meaning "the thing ptr points to doesn't exist". Therefore !ptr easily goes in my mind to "ptr does not exist" which is more semantically correct for most cases. There may be some (pointer arithmetic in particular) where ptr == NULL makes more sense, but they're not the common case.
The modern GUI equivalent is component systems. To name a few:
OLE/ActiveX/COM Java Beans Bonobo DCOP CORBA
Now a lot of people are going to respond about how these aren't as simple as unix pipes. That's because the data comming from GUIs isn't as simple as the data comming from simple unix text apps. All of the unix apps you could stream into/out of expected one directional constant flow data. In fact any time you have a unix console app that's at all interactive you can't do anything with its output (think top or vi). You can redirect these utilities simply because the things they do are simpler.
So with GUIs we have bi-directional data that stops and starts erratically. Effectively we have an event system. What is the best way of representing an event source/destination? As an object. But different objects have different capabilities, so you can't arbitrarily connect objects together, unless an object could tell you what it's capable of. And that right there is the defining feature of every component system out there. And no, it doesn't allow a user on a command line to just string components together, because the data is more complex than a user on a command line could deal with. It does however let a low level semi programmer throw together an impressive looking app with not a lot of VB script, or a graphic designer make a completely dynamic website out of a little JSP.
CRTs are a better picture than plasma or lcd yes, but you're forgetting about DLP. Best picture bar none, and slightly smaller than CRT. There is NO reason to get a CRT over a DLP.
I played Far Cry, and I'd say looks about the same (environments look better, characters look worse (HL2s incredible characters are one of the best parts of the game), and while it was really good, I haven't gotten drawn into a game like this since the original Half Life.
Yes there are lots fo perfectly good other games, but frankly, they're not as good as this one. I was with everyone about the annoyances until I actually played it. This is the best FPS ever made in my opinion, and the fact that I've barely slept the last two nights is testament to this fact.
Even that isn't always a bad thing. The drawback of the video chip using main memory, is both the fact that it...well...uses a big chunk of your memory, and the fact that doing a render from main memory uses memory bandwidth the processor could be using. The benefit is that software doesn't have to send video data over the agp bus to get it onto the video card. It just moves it from one part of memory to another (or possibly even just remaps which blocks of memory the video chip is using to the blocks that software built up the model in, depending on how good the driver is).
Slashdot Mirror
Um, system extensions were little more than shared libraries. A feature every operating system since the 70s has had. The only difference was that apple showed you an icon as it loaded them in.
Hmm, on the other hand, the old slashdot used to look perfect on the Blazer browser that comes with the Palm Treo 650, but now it's unreadable.
DC has a few that I'd kill to see in movie form. Transmetropolitan being top of that list.
On the other hand Hellblazer was on that list too before that abomination with Keanu Reeves.
Actually it has everything to do with the twist in the wire. If you twist a +/- wire pair together, the signals they put out are cancelled some. The tighter the twist the higher frequency is cancelled. I'm not sure the exact reason, but varying the twist along the length of cable could be to remove any resonance.
When its done right, its the same source code base with the extra features compiled out, not disabled by a runtime option.
For most of those cases, app servers have completely replaced CGI. The problem with CGI is lack of persistence, and process startup costs. These days, any decent web app is written with some persistent layer that interacts directly with the web server, and loads code snipets of its own language, starts objects, calls callback routines or something similar to handle requests. mod_perl, J2EE/JSP, PHP, Zope, and Ruby on Rails are all examples of this.
The mindless force of nature is quite effective at differentiating what's interesting and what isn't. The "interesting" changes survive. By interesting we mean more capable of survival than what came before. If you take a colony of bacteria and let it sit around, tons of mutations happen. Each set of offspring is slightly different than its parents. Some of those changes are drastic and some aren't. The ones that aren't, don't really change much. Of the ones that are, given that they are random, most kill the reciever. Then every once in a while one is beneficial. That bacterium is more suited to survive than the rest, and so reproduces more. The resulting group is more suited to reproduce more, and so does. Soon the colony is taken over by this change. The colony has evolved.
You keep saying that randomness doesn't evolve, but you've obviously never seen the Game of Life. Go download it sometime. A bunch of random numbers are piped on screen and set to evolve. Soon patterns emerge from it. Its random stuff that's happening, but in any random set, some patterns are self propogating. Since non-propogating patterns die out as soon as they appear, the whole set if left long enough will eventually fall into self propogating patterns. Life is a self propogating pattern. In a random universe life is inevitable.
Bullshit. The only people in my life I've ever heard that come from are the people who refuse to do their time with the grunt work, and are waiting around, leeching off the people around them, waiting for their "big chance" to come through so they can do interesting work like the assume people like me have always done. I'm not going to bother telling you how much grunt work I (and everyone around me who is doing interesting things) I had to do to get where I am.
That was SOMEWHAT of the case with Win9x. What a lot of people don't realize is that NT and later actually have a display server. Its just very well hidden.
Ok, something I always wondered why no one's done. Yes for most things, unix domain sockets are fast enough. But the X-SHM extension is MUCH faster for some operations (that mostly involve sending lots of pixmaps). But an application has to be written to use it. Well, we have an abstraction layer here that could dynamically do this, it's called X-lib. Why has no one written a modified X-lib that detects if you're connecting directly and automatically converts some operations to X-SHM?
Editor utility functions are different from UI and environment plugins. For quick and dirty write-these-lines-for-me, syntax hilighting rules, common task macros, emacs can't be beat (I sure as hell can't live without it). But for managing projects, integrating whole other apps (like debuggers, web browsers, class browsers, db interfaces) emacs doesn't hold a candle to eclipse. The thing I have the most hope for is the plugin that uses emacs as eclipse's editor using gnuserv. That is the one thing tempting me from moving from pure emacs.
That's because the only people using java use it server side. JSP, Weblogic, and J2EE are pretty much the way web dev is done these days, both on and off windows.
Do a google search for xorg and sub-pixel rendering. Cleartype is not a microsoft exclusive thing.
Hey, this winter we're having a drought. At least its a drought by our standards (I look out the window and notice it's raining right now).
I realize alot of people see this game initially, realize the controls for each character are the same and assume it's a simple game with no depth. 90% they've never tried to do a smash move (or decided they don't work), and haven't realized that even though the controls for each characters moves are the same, the end results are subtly (or in some cases drastically) different for each one. And the differences aren't just animations like in some fighting games. If you see a character swing his fist in a 23 degree arc over his head, that's exactly how much of a hit area they have. Then you take in the fact that each of these moves behaves completely differently with any of the MANY items in the game, and every character reacts to these items differently. Lastly you throw in the fact that unlike every other fighter, your end goal is to knock them out of the ring, not do them damage (damage just makes them fly farther with each hit), and that every move in the game has a balance between how far it throws and how much damage it does (which sometimes even varies based on timing or placement), and the vast number of recovery moves in the game, the complex, dynamic levels, and you start to realize the depth.
The game is the kind of genius that nintendo is famous for. There is no learning curve. Special moves are a button and a direction. Anyone can pick it up and play it. But you can keep learning things about this game up through the 2300 versus matches the thing's stat screen says I've played.
Dammit, should have hit preview.
Ctrl-< beginning of page
Ctrl-> end of page
At this point, I wouldn't be able to survive on a system that didn't support:
Ctrl-a beginning of line
Ctrl-e end of line
Ctrl- end of page
Taking my fingers off the main set of keys is far too cumbersome. Non-emacs users will disagree with me however.
You could look at it as ptr not being a boolean type, hense ! not making sense. I on the other hand see ptr == NULL meaning "the thing ptr points to doesn't exist". Therefore !ptr easily goes in my mind to "ptr does not exist" which is more semantically correct for most cases. There may be some (pointer arithmetic in particular) where ptr == NULL makes more sense, but they're not the common case.
The modern GUI equivalent is component systems. To name a few:
OLE/ActiveX/COM
Java Beans
Bonobo
DCOP
CORBA
Now a lot of people are going to respond about how these aren't as simple as unix pipes. That's because the data comming from GUIs isn't as simple as the data comming from simple unix text apps. All of the unix apps you could stream into/out of expected one directional constant flow data. In fact any time you have a unix console app that's at all interactive you can't do anything with its output (think top or vi). You can redirect these utilities simply because the things they do are simpler.
So with GUIs we have bi-directional data that stops and starts erratically. Effectively we have an event system. What is the best way of representing an event source/destination? As an object. But different objects have different capabilities, so you can't arbitrarily connect objects together, unless an object could tell you what it's capable of. And that right there is the defining feature of every component system out there. And no, it doesn't allow a user on a command line to just string components together, because the data is more complex than a user on a command line could deal with. It does however let a low level semi programmer throw together an impressive looking app with not a lot of VB script, or a graphic designer make a completely dynamic website out of a little JSP.
No, GLU provides that.
CRTs are a better picture than plasma or lcd yes, but you're forgetting about DLP. Best picture bar none, and slightly smaller than CRT. There is NO reason to get a CRT over a DLP.
I played Far Cry, and I'd say looks about the same (environments look better, characters look worse (HL2s incredible characters are one of the best parts of the game), and while it was really good, I haven't gotten drawn into a game like this since the original Half Life.
Yes there are lots fo perfectly good other games, but frankly, they're not as good as this one. I was with everyone about the annoyances until I actually played it. This is the best FPS ever made in my opinion, and the fact that I've barely slept the last two nights is testament to this fact.
My mother was from Philly. Its pretty much a Jersey accent only whinier.
Even that isn't always a bad thing. The drawback of the video chip using main memory, is both the fact that it...well...uses a big chunk of your memory, and the fact that doing a render from main memory uses memory bandwidth the processor could be using. The benefit is that software doesn't have to send video data over the agp bus to get it onto the video card. It just moves it from one part of memory to another (or possibly even just remaps which blocks of memory the video chip is using to the blocks that software built up the model in, depending on how good the driver is).