You install the OS, then lock the sucker up tight in a server room. Then you log in remotely.
Think about this: no OS can possibly guarantee security without physical security. If you don't have physical security, what will prevent someone from opening up the box, refreshing the CMOS so it can boot to the floppy (you did put a password on the BIOS setup, didn't you?), booting from a floppy, and doing whatever he wants to (reinstalling the OS, copying files off of the disk, creating an account for himself)? Encrypted file systems can help a little bit...
Re:You must not do anything interesting on them
on
Microsoft's DNS Down
·
· Score: 1
So you are implying that the same things don't bring down a *NIX system? Registry=/etc, and I've occasionally made a dumb change to/etc that brought down my system and made me reboot to single user mode to fix and get back up again. I've also had to debug daemon code from coworkers that forgot to deallocate resources and ate up all file handles. Yes, the OS didn't crash, but I couldn't log in to run kill either.
I currently run NT4 SP6a. At first, it would only run for two weeks or so before the graphics driver crashed, so I backed down to the previous version of the driver. I went 2 months without rebooting last year, and then the power went out. I haven't duplicated that since, because I do heavy development on all kinds of different things, necessitating installing and uninstalling things all the time (and yes, there is something to be said about having to reboot when you install something).
As has been said before, just use the right tool for the job. I love my FreeBSD machine. It has a place of honor under my bed (with no monitor or keyboard) and uptime indicates 39 days (days since the above mentioned power failure). It works wonders without ever complaining, and I keep it going because there are a lot of times when it is much better for certain tasks than NT. But when NT is better, I don't let my feelings for MS get in the way of my using their OS.
To some extent, there is something to be said about boycotting an evil industry. However, I've got some doubts about the utter evilness of MS. They work hard to make sure it is in our best interest to purchase their products. Sometines the methods they use tick me off, but as often as they do stupid things, they do smart things like produce a useful product.
So just use the right tool.
Re:Microsoft == bad partner, no multimedia savvy
on
Live Streaming Video?
·
· Score: 1
I've got a theory about why Macs get those kinds of tools, and it has nothing to do with Microsoft sucking. Which is simpler to get the hang of, set up, get running: Windows or Mac? Ok, so assume you are a graphics/publishing/video edition/etc. person. You are an expert in liberal arts. Do you choose a PC or a Mac?
On the other hand, assume you are a business person. You need a lot more flexibility with your configurations, and a lot more power under the hood for less money. You need two buttons on your mouse. You need a platform that you can configure properly for enterprise. You choose... Linux, but then you realize that the apps you need for your business aren't available yet, so you choose Windows instead.
So, logically, the business apps will continue to lag on the Mac, and the graphics apps will lag on the PC. Part of this also is due to simple history, as the Mac had a GUI before the PC, splitting the business people from the liberal arts people early on.
Streaming is a very nice thing for some purposes, and real has something that I don't think anybody else has. There is a checkbox on the encoder that allows you to mark the file as "saveable." That means that even if it was streamed, you can save-as afterwards. I don't think QT or WMP have this. It makes the no-save problem 100% the fault of the content provider.
Re:Quicktime Streaming Server is slashdot friendly
on
Live Streaming Video?
·
· Score: 1
I used to work a lot with streaming media, and I keep in touch somewhat with the guys who still do who used to be working with me on it. From what I know, Quicktime has a LOT of nice ideas that have been halfway implemented. It isn't as efficient (it requires a faster machine to view the same video as compared to Real and Windows Media), and both the server and the client were so screwed up that they decided quicktime wasn't an option for their uses. All the hype aside about Sorenson codecs, they determined that for modem-bandwidths, sorenson is worse than MPEG4 and Real.
I used to do a lot of really intricate client-side JavaScript programming. I started out my web programming career pretty solid on the side of Netscape, and last I checked I was pretty solid on the side of IE. I'm very glad Netscape exists: I need something good to run when I'm not running MS operating systems. However, they've done a lot of things wrong, MS did a lot of things right with IE, and I'm not seeing much change in the situation in the near future. I would love for Netscape to become as stable, capable, and easy to program for as IE, and once it happens, I will renew my efforts to support it. But it hasn't happened yet, and I can't sacrifice my productivity to the cause of supporting something that doesn't deserve my support.
Can I just say that IE is much easier to program for (in JavaScript)? The document object model is more consistent and powerful. A few things are easier to do in Netscape, but most are better in IE. In addition, CSS works much closer to the way it is supposed to in IE. Finally, while trying to create some normal (non-malicious) programs, I found some innocent looking JavaScript code that would crash any Netscape browser and, if running Windows 95/98, would put the OS into an unstable state (NT just had to log out and back in). That is really unacceptable.
Back to the main topic: standards on the web. No, I don't think things will get much worse than they are now. I've personally been seeing about the same percentage of browser-specific sites over the past 3 years or so, without much change. If anything, things will get better. Standards committees have been getting more attention recently than they did a few years ago. Even though there will always be proprietary extensions making some sites browser-specific, as long as the standards are adequate, it is easy enough to write to them and avoid the incompatible stuff.
Try to see the other point of view. Programmers for some sites simply get tired of pounding their heads against the wall to get their code to work on all browsers, and just spend their time working on the one that accounts for 96% of their web traffic, inviting the rest to try a different browser. If I were in that situation, yes I would try to at least make it visible to other browsers, but I would probably not want to spend a whole lot of time on it. Especially on a site where they discuss technology that is inherently MS specific, it makes a lot of sense to require IE: all of the customers are running Windows anyway (or they wouldn't be using this MS specific product), so why not save thousands of dollars in web programmer time and just support IE?
Anyway, that's my $0.25. YMMV. And I'm still waiting, for the Second Coming (of Netscape, that is!)
It had a few bugs, but nothing out of the ordinary for a x.0 release. It installed two different versions of the Java Runtime (1.3.0 and 1.3.0_xxxx). And it has already crashed once. But that isn't so unexpected. Much better than 4.0 was! I'm sure it will get a lot better with time.
My 2 bits: Why can't Netscape allow drag-and-drop favorites like Explorer? They've stolen most of the other good ideas, but this is one of the big reasons why I like Explorer. It is much easier to just drag a shortcut up to the menu bar than it is to add a shortcut, then edit shortcuts to put it where you want it.
Then it would be nice (although I can see good reasons, such as Unix compatibility) to handle the middle mouse button like most other Windows apps (scrolling).
As is mentioned elsewhere, the 45-55 days thing is Windows 98, I believe. Windows NT has a fairly good kernel (at least compared to 95/98). My desktop machine currently has uptime of 17 days. I was getting blue screens from time to time, sometimes right after boot, sometimes after 10-12 days of uptime, so I downgraded to a previous version of my video driver and haven't had a blue screen yet.
As far as uptime on NT, the best way I know of is with event viewer. Look at the system log and find the most recent entries from "EventLog." It will probably be an event indicating when the eventlog was started.
I think I know what the previous poster was trying to get at: you would have to use some symmetric encryption, and the symmetric key would have to be on a machine that could be cracked.
Reminder about the structure of Public/Private key encryption: In most use, you only use Asymmetric encryption to encrypt a very small amount of data (usually a session key for a symmetric algorithm) because asymmetric encryption is pretty darned slow. Trying to encrypt all logs using Public/Private encryption would be very CPU intensive.
If you do want to encrypt logs, you could maybe generate a new session key every hour and encrypt that with asymmetric encryption when it is no longer being used...
Re:Anti-aliasing on conventional monitors.
on
Cleartype In Depth
·
· Score: 1
It isn't so much an issue with CRT screens. The underlying idea is that colors on CRT screens are blended over the whole pixel, while colors on LCD's are separated into specific R, G, and B stripes. If you want to triple the number of dots per inch you can fit on the screen, just address each stripe independantly.
Although I'm not an expert, I got the MS newsletter a few months ago about this. If I remember correctly, LCD screens form the colors with exactly three points of light, arranged in a definite pattern. Let's assume it is point-R at the left, point-G in the middle, and point-B at the right (it may be different, I don't remember exactly).
If you were to light up a pixel as blue, then change it to red, the pixel would actually move over 2/3 pixel. Do this with a completely blue image on black background, and the whole image would move 2/3 pixel over. If instead of changing the pixel colors, you had moved the entire image one pixel to the left while changing the color to blue, the image would only move over 1/3 pixel.
On a CRT, this doesn't happen - each pixel is made of several points of light of each color, so the blend is over the entire pixel area.
You would think that the shift of 1/3 or 2/3 pixel isn't that big. Well, for a picture, it isn't. But for a character on the screen that is only a few pixels high, this is a big deal. The borders of the letter are chunky, especially on LCD's. Using simple blending techniques for CRT's is about as good as it gets, since each color is blended over the entire pixel area, so it tends to blend nicely and make things smooth.
On LCD's, where the colors are separated, this tends to make things blurry. With resolution at a premium, blurry is a bad idea. Instead of using the tried and true CRT method (which doesn't work for LCD's), just treat each of the three colors as a pixel. Now you have triple the resolution across the screen! Although you might get funny colors at the edges, it will probably average out. And the result is supposedly much crisper text, since it is effectively composed of more pixels.
Don't mean to be flamebait or anything, but the main reason I use IE is that it crashed far less often than Netscape, and many people that I've talked to agree. YMMV.
He just might be right. However, the problem with this is that it seems to take some of the wind out of the sails of the Justice Department. They went to all kinds of trouble to show how IE was inextricably linked to the OS, and then made the case that this was damaging to industry. Now we have someone important saying, "it wouldn't be such a big deal to have the OS linked with the browser..."
No loss of functionality seems likely with JVM, client applications, compilers (assuming they release complete specs on the OS), or server software separated from the main OS group. The only disadvantage is that things don't get bundled anymore. This is a good and a bad thing.
For example, buying NT Server means you get a free streaming video server (unlimited # of streams). Pro: free, fairly good technology. Con: requires Windows clients (insert cynical laugh here about their pretended attempts to port ASF clients to Mac and Unix). This is going to be a painful but necessary transition: many streaming video producers have been happy with telling their clients to view from a windows machine. Now they will be forced to buy the server software anyway, so they will be much more likely to pick one that is alternate-OS friendly. Splitting off server apps would probably be a good thing. (As far as file sharing goes, you can consider that an add-on, just as advanced server capabilities are added on to the NT platform to get NT Server, even though the underlying OS is the same)
On the other hand, unbundling client apps makes me uncomfortable. Buying Windows means you get standard ftp, telnet, WordPad, browser with JVM, defrag, scandisk, file manager, email client, news reader on EVERY MACHINE. I always replace the telnet app and get a real word processor, but it would really suck to have to buy all of these things separately, or to try to find freeware copies of everything. Pro: You can do basic things on your computer after installing just the OS from one CD, and everybody with Windows has the basic software. (Don't discount how useful this is: imagine trying to fix somebody's computer over the phone, and having to ask them, "which text editor do you have installed? What? No text editor? No way to edit your config files? Uh, do you have an ftp client? Which one? Oh, the new one I've never seen before. Great...") Con: They get to steal the wind from the sails of anybody who makes anything useful.
I dunno. I like having ftp, telnet, defrag, scandisk, a basic web browser, a wordpad, a notepad, and a file manager come with Windows. I don't know where to draw the line, but I think they should be able to bundle basic utilities.
Interesting... I extracted the new client (under Windows 98), and it took over a minute to extract, with 100% CPU usage, jerky mouse, and no HDD activity. I finally figured out that it was the F-PROT (F-STOPW) that was the culprit. It reported no error, didn't detect anything, and didn't even report it as suspicious, just took over a minute to check it. Weird.
So anyone running a virus checker, don't be surprised if copying/extracting the file takes forever, and you may want to set your virus checker to ignore it once you have checked it in the first place. Especially if you use F-STOPW.
This may be an unpopular opinion, but after having some experience as a web developer, I'm about fed up to here with browser variety. I don't mind competition, and I understand that two full-powered browsers should exist so the competition can continue. On the other hand, I have been frustrated to no end by having to remember all of the little (or big) differences between browsers, having to remember a bug list for each browser, etc. Most JavaScript programmers prefer IE because the DOM makes sense, but you have to program for Netscape anyway.
Bottom line: I am all for competition, but it better be standards based competition or forget it!
Too bad he's an important guy, so they actually pay attention to things like his age.
The first time I went to Comdex, I was 17 as well. I got a lot out of it. A non-technical entrepreneur relative took me down there as his "technical advisor." I didn't even know there was an age limit (I guess he put me down as age 18). But I didn't see anything that would have made it important to keep the audience 18 or over. I went again last year, and again, there wasn't any reason to keep people out based on age.
I can see why they would want to keep some young people out: I know a lot of people that would be immature enough to cause a problem at Comdex. At least once you turn 18, you are legally responsible for your own stupidity. Even though I know a bunch of older people who probably shouldn't be admitted based on maturity level, and many 15 year olds who should, the legal repercussions make 18 a logical cutoff point. Too bad rules have to be solid (you give an inch, they take a mile).
I put my solution on a web page. (I want to see if I can get SlashDotted ). However, I recommend that you verify the answer. It seems to have worked out, and all of the answers are grammatically correct, but I may have messed up.
IIS is much more a part of the operating system than Explorer, which MS argued for weeks was a part of the server. And it is much more a part of the OS than Apache.
You buy Windows NT *SERVER*. You can make it a file server, a domain server, a DHCP server, a WINS server, or an Internet server. If you want it to be an Internet server, you install IIS. IIS is supplied as part of the OS by Microsoft to all owners of NT Server ON THE INSTALL CD's (Apache happens to come with some distros, but it comes as part of the applications library, not as a part of the kernel or base install). It was created by the creators of NT Server (Apache != Linux). It integrates itself into the OS as a system service (Apache doesn't run in kernel space, and doesn't need Admin priviledges).
Now, if you said Netscape Server, things would be different.
I completely agree. Nonetheless, I am working on this topic quite a bit. Ideally, it wouldn't be necessary. However, not everyone has the opportunity to go to a regular university. For those in (insert name of remote desert isle here), those who can't afford regular tuition, or those who cannot quit their jobs, virtual universities offer a great opportunity. It is definitely not as good as a real university, but it beats nothing hands down.
I did a lot of work on this last year, and it looks like I will be continuing with this over the next few months. We determined that it is NOT FEASIBLE to transmit a video representation of slides or overhead material at this point. The resolution of a video camera is too coarse, for one thing, and assuming you COULD get legible data on the video input, it is an incredible waste of bandwidth.
What we found to work best was to use a 20kbps or 35kbps (depending on whether we were aiming for 28k or 56k modems) video/audio stream. You can get a 128x96 video image at a respectable frame rate (3-6 fps) and a 6k audio stream (adequate for mono voice, terrible for music, though) into a 20k stream, and a better frame rate at 35k. Then we redid the slides/overheads in either Flash or Powerpoint (exporting to GIF) and transmitted them separately from the video/audio stream. Admittedly, the slides updated slowly over a 28k modem, but it worked. In fact, I can even say it worked well. (Of course, it worked better over the LAN...).
You don't need a full screen of a talking head. What is the purpose, anyway? The important thing is the slides, the text, and the audio. Video is really secondary. Therefore, we even got things set up to use a transcribed copy of the lecture text (searchable, and synchronized with the audio/video stream so you can goto a position in the text and the stream jumps to the correct position and vice versa).
Looking through the release notes, I would say that there really isn't enough changes in 3.3 to justify upgrading unless you have a device that was not supported by 3.2 but is supported by 3.3. That said, I also agree that CVSUP is a valueable thing to learn. So if you decide to upgrade, do it for the learning experience, not for the upgrade itself.
In addition, if you are going to try CVSUP, I would recommend that you wait a few weeks before doing so. Those people who have good reason to install right away (not me!) will appreciate you not taking up the limited bandwidth.
I feel for you. I just installed 3.2, and then one of the first email messages I got after the install was a bugfix announcement that indicated that the fixes would be available shortly in 3.3. I screamed. Then I realized that there really wasn't much in 3.3 that I cared about anyway. Then I was happy again.;)
I would guess that a lot of those people who don't follow Phrack closely do know what Phrack is. They just don't realize that Phrack 55 means "Phrack, issue #55." The first time I saw it mentioned in that format, I thought something along the lines of "I thought Phrack was an e-magazine... Is Phrack XX a new version of some utility for script kiddies or something?" It is kindof confuzzing when words like "issue" get left out. Not that it is bad to name the issues that way, but give the people a break when they don't "get it" right away.
I agree. We're massively parallel. All they proved is that there is a limit to our parallelism. Duh! I can only focus on so many details at once. I knew that without any brain-wave processing.
I already knew that when I read a book, I don't stare at the whole page until the text of the book sinks in. I read word after word. When I'm looking for a detail in a picture, I don't just stare at the picture until I find what I'm looking for. I scan small areas that look interesting until I can focus in on the area in question.
Slashdot Mirror
You install the OS, then lock the sucker up tight in a server room. Then you log in remotely.
Think about this: no OS can possibly guarantee security without physical security. If you don't have physical security, what will prevent someone from opening up the box, refreshing the CMOS so it can boot to the floppy (you did put a password on the BIOS setup, didn't you?), booting from a floppy, and doing whatever he wants to (reinstalling the OS, copying files off of the disk, creating an account for himself)? Encrypted file systems can help a little bit...
So you are implying that the same things don't bring down a *NIX system? Registry=/etc, and I've occasionally made a dumb change to /etc that brought down my system and made me reboot to single user mode to fix and get back up again. I've also had to debug daemon code from coworkers that forgot to deallocate resources and ate up all file handles. Yes, the OS didn't crash, but I couldn't log in to run kill either.
I currently run NT4 SP6a. At first, it would only run for two weeks or so before the graphics driver crashed, so I backed down to the previous version of the driver. I went 2 months without rebooting last year, and then the power went out. I haven't duplicated that since, because I do heavy development on all kinds of different things, necessitating installing and uninstalling things all the time (and yes, there is something to be said about having to reboot when you install something).
As has been said before, just use the right tool for the job. I love my FreeBSD machine. It has a place of honor under my bed (with no monitor or keyboard) and uptime indicates 39 days (days since the above mentioned power failure). It works wonders without ever complaining, and I keep it going because there are a lot of times when it is much better for certain tasks than NT. But when NT is better, I don't let my feelings for MS get in the way of my using their OS.
To some extent, there is something to be said about boycotting an evil industry. However, I've got some doubts about the utter evilness of MS. They work hard to make sure it is in our best interest to purchase their products. Sometines the methods they use tick me off, but as often as they do stupid things, they do smart things like produce a useful product.
So just use the right tool.
I've got a theory about why Macs get those kinds of tools, and it has nothing to do with Microsoft sucking. Which is simpler to get the hang of, set up, get running: Windows or Mac? Ok, so assume you are a graphics/publishing/video edition/etc. person. You are an expert in liberal arts. Do you choose a PC or a Mac?
On the other hand, assume you are a business person. You need a lot more flexibility with your configurations, and a lot more power under the hood for less money. You need two buttons on your mouse. You need a platform that you can configure properly for enterprise. You choose... Linux, but then you realize that the apps you need for your business aren't available yet, so you choose Windows instead.
So, logically, the business apps will continue to lag on the Mac, and the graphics apps will lag on the PC. Part of this also is due to simple history, as the Mac had a GUI before the PC, splitting the business people from the liberal arts people early on.
Streaming is a very nice thing for some purposes, and real has something that I don't think anybody else has. There is a checkbox on the encoder that allows you to mark the file as "saveable." That means that even if it was streamed, you can save-as afterwards. I don't think QT or WMP have this. It makes the no-save problem 100% the fault of the content provider.
I used to work a lot with streaming media, and I keep in touch somewhat with the guys who still do who used to be working with me on it. From what I know, Quicktime has a LOT of nice ideas that have been halfway implemented. It isn't as efficient (it requires a faster machine to view the same video as compared to Real and Windows Media), and both the server and the client were so screwed up that they decided quicktime wasn't an option for their uses. All the hype aside about Sorenson codecs, they determined that for modem-bandwidths, sorenson is worse than MPEG4 and Real.
I used to do a lot of really intricate client-side JavaScript programming. I started out my web programming career pretty solid on the side of Netscape, and last I checked I was pretty solid on the side of IE. I'm very glad Netscape exists: I need something good to run when I'm not running MS operating systems. However, they've done a lot of things wrong, MS did a lot of things right with IE, and I'm not seeing much change in the situation in the near future. I would love for Netscape to become as stable, capable, and easy to program for as IE, and once it happens, I will renew my efforts to support it. But it hasn't happened yet, and I can't sacrifice my productivity to the cause of supporting something that doesn't deserve my support.
Can I just say that IE is much easier to program for (in JavaScript)? The document object model is more consistent and powerful. A few things are easier to do in Netscape, but most are better in IE. In addition, CSS works much closer to the way it is supposed to in IE. Finally, while trying to create some normal (non-malicious) programs, I found some innocent looking JavaScript code that would crash any Netscape browser and, if running Windows 95/98, would put the OS into an unstable state (NT just had to log out and back in). That is really unacceptable.
Back to the main topic: standards on the web. No, I don't think things will get much worse than they are now. I've personally been seeing about the same percentage of browser-specific sites over the past 3 years or so, without much change. If anything, things will get better. Standards committees have been getting more attention recently than they did a few years ago. Even though there will always be proprietary extensions making some sites browser-specific, as long as the standards are adequate, it is easy enough to write to them and avoid the incompatible stuff.
Try to see the other point of view. Programmers for some sites simply get tired of pounding their heads against the wall to get their code to work on all browsers, and just spend their time working on the one that accounts for 96% of their web traffic, inviting the rest to try a different browser. If I were in that situation, yes I would try to at least make it visible to other browsers, but I would probably not want to spend a whole lot of time on it. Especially on a site where they discuss technology that is inherently MS specific, it makes a lot of sense to require IE: all of the customers are running Windows anyway (or they wouldn't be using this MS specific product), so why not save thousands of dollars in web programmer time and just support IE?
Anyway, that's my $0.25. YMMV. And I'm still waiting, for the Second Coming (of Netscape, that is!)
It had a few bugs, but nothing out of the ordinary for a x.0 release. It installed two different versions of the Java Runtime (1.3.0 and 1.3.0_xxxx). And it has already crashed once. But that isn't so unexpected. Much better than 4.0 was! I'm sure it will get a lot better with time.
My 2 bits: Why can't Netscape allow drag-and-drop favorites like Explorer? They've stolen most of the other good ideas, but this is one of the big reasons why I like Explorer. It is much easier to just drag a shortcut up to the menu bar than it is to add a shortcut, then edit shortcuts to put it where you want it.
Then it would be nice (although I can see good reasons, such as Unix compatibility) to handle the middle mouse button like most other Windows apps (scrolling).
As is mentioned elsewhere, the 45-55 days thing is Windows 98, I believe. Windows NT has a fairly good kernel (at least compared to 95/98). My desktop machine currently has uptime of 17 days. I was getting blue screens from time to time, sometimes right after boot, sometimes after 10-12 days of uptime, so I downgraded to a previous version of my video driver and haven't had a blue screen yet.
As far as uptime on NT, the best way I know of is with event viewer. Look at the system log and find the most recent entries from "EventLog." It will probably be an event indicating when the eventlog was started.
I think I know what the previous poster was trying to get at: you would have to use some symmetric encryption, and the symmetric key would have to be on a machine that could be cracked.
Reminder about the structure of Public/Private key encryption: In most use, you only use Asymmetric encryption to encrypt a very small amount of data (usually a session key for a symmetric algorithm) because asymmetric encryption is pretty darned slow. Trying to encrypt all logs using Public/Private encryption would be very CPU intensive.
If you do want to encrypt logs, you could maybe generate a new session key every hour and encrypt that with asymmetric encryption when it is no longer being used...
It isn't so much an issue with CRT screens. The underlying idea is that colors on CRT screens are blended over the whole pixel, while colors on LCD's are separated into specific R, G, and B stripes. If you want to triple the number of dots per inch you can fit on the screen, just address each stripe independantly.
Although I'm not an expert, I got the MS newsletter a few months ago about this. If I remember correctly, LCD screens form the colors with exactly three points of light, arranged in a definite pattern. Let's assume it is point-R at the left, point-G in the middle, and point-B at the right (it may be different, I don't remember exactly).
If you were to light up a pixel as blue, then change it to red, the pixel would actually move over 2/3 pixel. Do this with a completely blue image on black background, and the whole image would move 2/3 pixel over. If instead of changing the pixel colors, you had moved the entire image one pixel to the left while changing the color to blue, the image would only move over 1/3 pixel.
On a CRT, this doesn't happen - each pixel is made of several points of light of each color, so the blend is over the entire pixel area.
You would think that the shift of 1/3 or 2/3 pixel isn't that big. Well, for a picture, it isn't. But for a character on the screen that is only a few pixels high, this is a big deal. The borders of the letter are chunky, especially on LCD's. Using simple blending techniques for CRT's is about as good as it gets, since each color is blended over the entire pixel area, so it tends to blend nicely and make things smooth.
On LCD's, where the colors are separated, this tends to make things blurry. With resolution at a premium, blurry is a bad idea. Instead of using the tried and true CRT method (which doesn't work for LCD's), just treat each of the three colors as a pixel. Now you have triple the resolution across the screen! Although you might get funny colors at the edges, it will probably average out. And the result is supposedly much crisper text, since it is effectively composed of more pixels.
Don't mean to be flamebait or anything, but the main reason I use IE is that it crashed far less often than Netscape, and many people that I've talked to agree. YMMV.
He just might be right. However, the problem with this is that it seems to take some of the wind out of the sails of the Justice Department. They went to all kinds of trouble to show how IE was inextricably linked to the OS, and then made the case that this was damaging to industry. Now we have someone important saying, "it wouldn't be such a big deal to have the OS linked with the browser..."
No loss of functionality seems likely with JVM, client applications, compilers (assuming they release complete specs on the OS), or server software separated from the main OS group. The only disadvantage is that things don't get bundled anymore. This is a good and a bad thing.
For example, buying NT Server means you get a free streaming video server (unlimited # of streams). Pro: free, fairly good technology. Con: requires Windows clients (insert cynical laugh here about their pretended attempts to port ASF clients to Mac and Unix). This is going to be a painful but necessary transition: many streaming video producers have been happy with telling their clients to view from a windows machine. Now they will be forced to buy the server software anyway, so they will be much more likely to pick one that is alternate-OS friendly. Splitting off server apps would probably be a good thing. (As far as file sharing goes, you can consider that an add-on, just as advanced server capabilities are added on to the NT platform to get NT Server, even though the underlying OS is the same)
On the other hand, unbundling client apps makes me uncomfortable. Buying Windows means you get standard ftp, telnet, WordPad, browser with JVM, defrag, scandisk, file manager, email client, news reader on EVERY MACHINE. I always replace the telnet app and get a real word processor, but it would really suck to have to buy all of these things separately, or to try to find freeware copies of everything. Pro: You can do basic things on your computer after installing just the OS from one CD, and everybody with Windows has the basic software. (Don't discount how useful this is: imagine trying to fix somebody's computer over the phone, and having to ask them, "which text editor do you have installed? What? No text editor? No way to edit your config files? Uh, do you have an ftp client? Which one? Oh, the new one I've never seen before. Great...") Con: They get to steal the wind from the sails of anybody who makes anything useful.
I dunno. I like having ftp, telnet, defrag, scandisk, a basic web browser, a wordpad, a notepad, and a file manager come with Windows. I don't know where to draw the line, but I think they should be able to bundle basic utilities.
(Hope this didn't post twice...)
Interesting... I extracted the new client (under Windows 98), and it took over a minute to extract, with 100% CPU usage, jerky mouse, and no HDD activity. I finally figured out that it was the F-PROT (F-STOPW) that was the culprit. It reported no error, didn't detect anything, and didn't even report it as suspicious, just took over a minute to check it. Weird.
So anyone running a virus checker, don't be surprised if copying/extracting the file takes forever, and you may want to set your virus checker to ignore it once you have checked it in the first place. Especially if you use F-STOPW.
This may be an unpopular opinion, but after having some experience as a web developer, I'm about fed up to here with browser variety. I don't mind competition, and I understand that two full-powered browsers should exist so the competition can continue. On the other hand, I have been frustrated to no end by having to remember all of the little (or big) differences between browsers, having to remember a bug list for each browser, etc. Most JavaScript programmers prefer IE because the DOM makes sense, but you have to program for Netscape anyway.
Bottom line: I am all for competition, but it better be standards based competition or forget it!
Too bad he's an important guy, so they actually pay attention to things like his age.
The first time I went to Comdex, I was 17 as well. I got a lot out of it. A non-technical entrepreneur relative took me down there as his "technical advisor." I didn't even know there was an age limit (I guess he put me down as age 18). But I didn't see anything that would have made it important to keep the audience 18 or over. I went again last year, and again, there wasn't any reason to keep people out based on age.
I can see why they would want to keep some young people out: I know a lot of people that would be immature enough to cause a problem at Comdex. At least once you turn 18, you are legally responsible for your own stupidity. Even though I know a bunch of older people who probably shouldn't be admitted based on maturity level, and many 15 year olds who should, the legal repercussions make 18 a logical cutoff point. Too bad rules have to be solid (you give an inch, they take a mile).
I put my solution on a web page. (I want to see if I can get SlashDotted ). However, I recommend that you verify the answer. It seems to have worked out, and all of the answers are grammatically correct, but I may have messed up.
The spoiler
Have fun! And let me know if I made any mistakes.
IIS is much more a part of the operating system than Explorer, which MS argued for weeks was a part of the server. And it is much more a part of the OS than Apache.
You buy Windows NT *SERVER*. You can make it a file server, a domain server, a DHCP server, a WINS server, or an Internet server. If you want it to be an Internet server, you install IIS. IIS is supplied as part of the OS by Microsoft to all owners of NT Server ON THE INSTALL CD's (Apache happens to come with some distros, but it comes as part of the applications library, not as a part of the kernel or base install). It was created by the creators of NT Server (Apache != Linux). It integrates itself into the OS as a system service (Apache doesn't run in kernel space, and doesn't need Admin priviledges).
Now, if you said Netscape Server, things would be different.
I completely agree. Nonetheless, I am working on this topic quite a bit. Ideally, it wouldn't be necessary. However, not everyone has the opportunity to go to a regular university. For those in (insert name of remote desert isle here), those who can't afford regular tuition, or those who cannot quit their jobs, virtual universities offer a great opportunity. It is definitely not as good as a real university, but it beats nothing hands down.
I did a lot of work on this last year, and it looks like I will be continuing with this over the next few months. We determined that it is NOT FEASIBLE to transmit a video representation of slides or overhead material at this point. The resolution of a video camera is too coarse, for one thing, and assuming you COULD get legible data on the video input, it is an incredible waste of bandwidth.
What we found to work best was to use a 20kbps or 35kbps (depending on whether we were aiming for 28k or 56k modems) video/audio stream. You can get a 128x96 video image at a respectable frame rate (3-6 fps) and a 6k audio stream (adequate for mono voice, terrible for music, though) into a 20k stream, and a better frame rate at 35k. Then we redid the slides/overheads in either Flash or Powerpoint (exporting to GIF) and transmitted them separately from the video/audio stream. Admittedly, the slides updated slowly over a 28k modem, but it worked. In fact, I can even say it worked well. (Of course, it worked better over the LAN...).
You don't need a full screen of a talking head. What is the purpose, anyway? The important thing is the slides, the text, and the audio. Video is really secondary. Therefore, we even got things set up to use a transcribed copy of the lecture text (searchable, and synchronized with the audio/video stream so you can goto a position in the text and the stream jumps to the correct position and vice versa).
Looking through the release notes, I would say that there really isn't enough changes in 3.3 to justify upgrading unless you have a device that was not supported by 3.2 but is supported by 3.3. That said, I also agree that CVSUP is a valueable thing to learn. So if you decide to upgrade, do it for the learning experience, not for the upgrade itself.
In addition, if you are going to try CVSUP, I would recommend that you wait a few weeks before doing so. Those people who have good reason to install right away (not me!) will appreciate you not taking up the limited bandwidth.
I feel for you. I just installed 3.2, and then one of the first email messages I got after the install was a bugfix announcement that indicated that the fixes would be available shortly in 3.3. I screamed. Then I realized that there really wasn't much in 3.3 that I cared about anyway. Then I was happy again. ;)
I got my announcement...
I would guess that a lot of those people who don't follow Phrack closely do know what Phrack is. They just don't realize that Phrack 55 means "Phrack, issue #55." The first time I saw it mentioned in that format, I thought something along the lines of "I thought Phrack was an e-magazine... Is Phrack XX a new version of some utility for script kiddies or something?" It is kindof confuzzing when words like "issue" get left out. Not that it is bad to name the issues that way, but give the people a break when they don't "get it" right away.
I agree. We're massively parallel. All they proved is that there is a limit to our parallelism. Duh! I can only focus on so many details at once. I knew that without any brain-wave processing.
I already knew that when I read a book, I don't stare at the whole page until the text of the book sinks in. I read word after word. When I'm looking for a detail in a picture, I don't just stare at the picture until I find what I'm looking for. I scan small areas that look interesting until I can focus in on the area in question.