dserban

USB printers for the most part work really well under Linux, but there are some that don't work at all - they might as well be paperweights, so before you buy a printer, definitely go to: http://www.openprinting.org/printers ... they have a very good, up-to-date database of printers. It's a dropdown box by manufacturer, so you click your way through it and the verdict is like a thumbs up or thumbs down for Linux compatibility - some printers work great, and some don't work at all ... and some grey areas inbetween. Additional information can be found at: http://forums.xisto.com/no_longer_exists/ or just: https://wiki.linuxfoundation.org/en/OpenPrinting (It used to be so simple, the only thing you had to do is go to linuxprinting.org, but they keep changing things on me every so often.) As a general observation, HP printers tend to work very well under Linux - not every single one, but a good number of them. (I'm not too sure about Xisto forum rules on mentioning vendor names directly. If I get in trouble, I'll replace the words "HP printers" with "printers made by one of the leading vendors".) So let's say that you have a printer and you want to network it. There are obviously lots of considerations here since everybody's network is going to be different. It's not uncommon for people to want to share printers - in a homogeneous environment / between two Linux computers, or in a heterogeneous environment / between Linux and Windows computers. The main program that you are going to need to have installed regardless of which way you're going is CUPS - Common UNIX Printing System. It is a server type piece of software - I am going to refer to it as the print server or CUPS daemon. What it does is three things: - it allows your Linux computer to print to your standalone printer, so it's a way to configure your directly attached printer without any network. - it allows your Linux computer to share that printer and process incoming print jobs. - it allows your Linux computer to print to a shared printer on a remote machine. You often see when you're reading around, about using Samba to share printers. It's actually interesting that you don't necessarily have to use Samba - you can use CUPS just by itself (there may be other packages out there, but I'm focusing on CUPS). In fact, you do need to use CUPS, no matter what. The Samba part is kind of optional. I'll explain this part in more detail further down below. A lot of times people just automatically assume they have to configure both CUPS *and* Samba in order to share printers and I guess my main point - at least initially - is just to say that's not necessarily the case. When you're sharing printers between Linux machines for example, CUPS is all you need - you don't need Samba. Obviously if you need to share files, that's a different story altogether, and sometimes people will set up Samba to share files between Linux computers anyway, so they might still be using Samba, but my point is you don't necessarily have to. Before I break it down into the different types of configurations, what you'll need to do of course is - like I said - to install CUPS. So go to your package manager, search for the CUPS package and go ahead and install that. Just like with any other service or daemon, you have the options to start / stop / restart. I have found this to be slightly different between different distributions. In most distros, what you need to do at the root shell prompt - or by prepending sudo - is: sudo /etc/init.d/cups (start|stop|restart|status) However, in a small number of distributions, you need to put a lowercase "d" after cups, like this: sudo /etc/init.d/cupsd (start|stop|restart|status) And to further complicate things, in Debian / Ubuntu it's different again: sudo /etc/init.d/cupsys (start|stop|restart|status) In some distributions you have graphcal tools for starting and stopping CUPS, so you can do that wherever the case may be, but I wanted to describe the three command line options you would have in order to start CUPS in most distributions. A few distributions (Slackware and Arch Linux for example) use a slightly different configuration (rc.d instead of init.d), but since these are fairly advanced Linux distributions, those of you who use them probably already know that. Once you have CUPS installed, there is one prerequisite you should take care of right away, and that is to enable other machines to access the remote CUPS server. At this point I'm referring strictly to a homogeneous landscape. If you have printers attached to a Windows machine and you want to print to them from within Linux, I will describe the steps further below, keep reading. But let's say you have a Linux box and you have a printer attached to it and you want to both access it directly from the local machine and also share it over the network. The very first step on the Linux machine in order to start setting up CUPS is to edit a file, just like you need to do with Samba. Open up a terminal and type: cd /etc/cups Look around and locate the file called cupsd.conf, which is the CUPS configuration file. The first thing I recommend you do is to back it up: cp cupsd.conf cupsd_backup.conf Then open cupsd.conf in you favorite editor, for example: sudo nano cupsd.conf Notice the opening and closing tags, the only tag you need to be concerned with is location. Scroll down until you get to a section that looks like this: <Location />Order Deny,AllowDeny from AllAllow from 127.0.0.1</Location>Insert a line that contains your subnet, but with the last number being a wildcard, like this:Allow from 192.168.1.* <Location />Order Deny,AllowDeny from AllAllow from 127.0.0.1Allow from 192.168.1.*</Location>You are thus allowing all machines on your local network access to the CUPS server - that's what that line does. And, just as is the case whenever you make a change in most configuration files in Linux, you will need to restart CUPS. We are going to do two more little optional changes. Although it's not necessary for the normal operation of CUPS, we are going to enable support for Windows PCL drivers. PCL drivers are Windows drivers that send raw data to the print queue, which then knows what to do to the raw data to format it and output it correctly to the printer, as opposed to the formatting having already been performed by the originating client end application. That means when you have raw printing, there is not much you can change on the client side, so this is not something you are necessarily going to need, but since we are already playing around with the terminal, we might as well enable this because this change tends to be forgotten a lot of times and it can fix a lot of problems. There are two files we need to edit: mime.types and mime.convs Open these files for editing and in both of them there will be a line in there somewhere with the following: application/octet-stream You will find this line commented in both files. Uncomment the line in both files and save them. (Yes, you should back up those files first, before you make this change.) Restart CUPS. At this point you need to set up your printer (we are still in the printer-attached-to-Linux-machine scenario). Fire up a web browser (CUPS has a web-based administration tool) and type in: http://localhost:631631 is the port that the CUPS daemon runs on, so before you do anything, make sure that your firewalls have port 631 open (again, that's something else that trips people up a lot). I'm not talking about allowing access to your 631 port from the Internet through your NAT router, I'm talking about software-based firewalls which are internal to your local network. Log in as root, click on the "Administration" button, click "Add printer" and this will walk you through a little wizard which is not very complicated to go through, so I'm not going to go into detail here about every little step, although a couple of brief comments on a few of these are in order. When it asks you for the spooler name, you will want to just type in a really short, easy to remember name that is going to be used on the system to identify the printer (that name should not contain any spaces or special characters because it's going to be embedded in URLs later on). Location is completely optional, and so is the description. On the next screen, where it's giving you the various options (HP JetDirect, Internet Printing Protocol, lpd/lpr, parallel port, USB printer) you need to know that JetDirect is a little hardware box that is going to act as a kind of a LAN port for your printer, with all that such a port entails (MAC address, IP address). Remember, what we are doing now is to set up the local printer on the local machine (we're not networking yet) so I'm assuming you aren't using JetDirect or IPP. Other self-explanatory fields are manufacturer, model number, etc. I will describe adding additional CUPS drivers further down below, but hopefully there will be a driver on the default CUPS screen here for your printer (remember, you can always go back and change the driver later on). Now that you have your printer all set up, you are back at the main CUPS admin page and you should see your printer now. You can go ahead and print a test page, hopefully that works for you. Now we get to the stage where we connect to this printer from the other machines. Should you want to connect to this printer from another Linux machine, you first need to install CUPS on that second box, open up the CUPS web-based administration tool, log in as root, click on the "Administration" button, click "Add printer", and on the next screen where it gives you all those options, go ahead and select the IPP (Internet Printing Protocol) option. The CUPS web-based administration tool gives you examples on how to type in the URL in the text field called "Device URI". So you'll see examples like: ipp://hostname-or-IP-address-of-CUPS-server-machine-where-the-printer-is-connected/printer/printernameand that's where that short spooler name comes in handy.Alternatively, try this: http://hostname-or-IP-address-of-CUPS-server-machine-where-the-printer-is-connected:631/ipp/queue-name-in-the-queue-syntax-formatbut usually ipp-based URIs work just fine. So that's what you can do to connect to that printer from another Linux machine. The interesting part is that you can perform almost the same sequence of steps to connect to that printer from Windows. So you still have your printer connected to the Linux machine, you have enabled the remote access to the printer under the location section in cupsd.conf, you have configured the printer in CUPS, and now you go over to your Windows machine and you want to connect to the remote printer. Same thing - IPP printing is supported out of the box in Windows 2000 and Windows XP. For Windows 9x and ME, you need to download and install some little piece of software to enable IPP printing, as far as I know. All you need to do in Windows 2000 and XP is go into the "Add printer" wizard, select "Network printer", and when it asks you for the URL, you type in: http://hostname-or-IP-address-of-CUPS-server-machine-where-the-printer-is-connected:631/printers/queue-name-in-the-queue-syntax-formatand it works very well - no Samba needed. But should you have had Samba set up already - let's say your Linux machine was already set up as a file server and serving files to your Windows machines - you can use Samba for printing as well, but keep in mind that Samba really uses CUPS as the back end, so you do need to get CUPS set up and working. But then you need to make two small changes to your Samba configuration file in order to share printers over Samba. The first change is in the global section of the smb.conf file. Here you will see 2 printing-related lines, and in both you need to put "cups", like this printcapname = cups printing = cups Then, scroll down to where you start defining the shares (home directories etc.) and you will see a section where you define shared printers: [printers]comment = All Printerspath = /var/spool/sambaprintable = Yesbrowseable = NoHere you need to add:public = Yes printername = <shorthand-name-of-printer> and that will make that printer discoverable in the Windows network using Samba, in which case you don't have to jump through the hoops of adding a network printer etc. So that's CUPS as a print server. The other situation is when you have a printer attached to your Windows machine already and you want to print to it from Linux. In this case you do need to be using Samba because there is no way that I know of for a Windows machine to serve printers using IPP. In this case the sequence of steps is pretty easy. Obviously in Windows you have to set up the shared printer. On the Linux side, you need to use CUPS and Samba. So go ahead and install CUPS and Samba on the Linux box, open up the CUPS web-based administration tool, log in as root, click on the "Administration" button, click "Add printer", and on the next screen where it gives you all those options, go ahead and select the "Windows printer via Samba" option, which should be there if you have installed Samba, configured it correctly and it's ready to go. When you get to the page where you need to type in the URI for the printer, you will need to use a slightly different syntax (and again, you'll see examples of this in CUPS). But basically the syntax is: smb://username:password@workgroup/server/printernameOf course username:password is a valid combination recognized by your Windows machine.If the above syntax doesn't work, there is another one you can try: smb://server/printernameOnce a printer is available in CUPS, it is automatically available to all Linux applications. There is a great CUPS tool in the KDE Control Center, and there is also of course a printer tool in Gnome, but the CUPS web tool is cross-platform and cross-desktop-environment, so once you get used to it, you'll find it very convenient to use on a regular basis in any combination of distribution and desktop environment. I promised I would talk about installing different CUPS drivers in Linux. There are a couple of them, for example the GIMP print driver (aka Gutenprint): http://gimp-print.sourceforge.net/ Basically, what you need to do is download and install that package, then you go back into CUPS to change the printer, once you get past the sections where it asks you for the manufacturer and model number of the printer, you'll see "Additional drivers" there, and you'll see GIMP print drivers for your printer (assuming they have any - they may not), which give you some additional features and settings. Another example is HP drivers (HPLIP): http://hplipopensource.com/ which give you additional options like gray scales, high quality or medium quality, etc. There are a bunch of good pages out there if you need more information. There is a really good one in the Gentoo docs, there are a couple in the Gentoo Wiki. (By the way, the Gentoo docs are always good places to start, even if you don't use Gentoo, their docs are sooo good.) And of course the documentation on the CUPS website is pretty good as well.

EMP is going to come in handy if and when RFID chips are going to start being implanted in stuff you need to carry around with you (like ID cards) to make it easy for governments to know who you are from 30 feet away. But for the moment that's the stuff of science fiction movies.

Ummmmmmmm ... no. Actually you only need something like this: $send = "GET / HTTP/1.1\r\n"; $send .= "Host: http://gameserver/housekeeping.php\r\n"; $send .= "Connection: Close\r\n\r\n";But after taking a better look at the code I gave you, I realize now that you also need to replace fsockopen( "http://yahoo.com/; ... ) with fsockopen( "<IP-address-of-your-game-server>" ... )

There is a good hands-on example on the fsockopen official documentation page at: http://php.net/function.fsockopen Look a 10th of the way down at the example code posted by: jbr at ya-right dot com There is a sequence of code that looks like this if ( ( $io = fsockopen( "yahoo.com;, 80, $errno, $errstr, 5 ) ) !== false ){ $send = "GET / HTTP/1.1\r\n"; $send .= "Host: yahoo.com;; $send .= "User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20021204\r\n"; $send .= "Referer: http://forums.xisto.com/no_longer_exists/;; $send .= "Accept: text/xml,application/xml,application/xhtml+xml,"; $send .= "text/html;q=0.9,text/plain;q=0.8,video/x-mng,image/png,"; $send .= "image/jpeg,image/gif;q=0.2,text/css,*/*;q=0.1\r\n"; $send .= "Accept-Language: en-us, en;q=0.50\r\n"; $send .= "Accept-Encoding: gzip, deflate, compress;q=0.9\r\n"; $send .= "Connection: Close\r\n\r\n"; fputs ( $io, $send );... this is the request part (fputs). But then he goes on to use fgets, which is the response part. That you don't need to do ... in fact it's a bad idea to attempt it.Now instead of $send .= "Host: yahoo.com;; you would obviously have $send .= "Host: link-to-PHP-invisible-housekeeping-script\r\n"; Your PHP invisible housekeeping script either reacts to the actions of the online gamers which are reflected in the database as table entries, or otherwise generates random events based on rules you preset - like a Pacman that keeps popping out of random places at random intervals trying to eat you alive if you're not careful. That you need to program yourself, and you need to define how those two things interact by defining the data model correctly.

If you can't find a hosting platform for your game, I would recommend - at least in a first stage - running the game on your own PC at home and giving your gaming buddies access to it. The only thing you need to do is open a port or two for direct access from the Internet.The ports you probably need to open are 80 for Apache and 3306 in extreme cases. Even with those ports open, you still have the option of controlling which IP addresses (your gaming buddies' PCs) can connect to your online game, even by using the brain-dead built-in firewall that comes bundled with Windows XP SP2.

Remember what I said ... ... emphasis on "invisible" ...The PHP script I was talking about has no way of communicating with any browser, so the idea that it would print something does not make any logical sense. It is a PHP script that does not take any input and as well does not produce any output. It's a script that is supposed to run in the background and update database tables in real time based on what online players happen to be doing as they interact with your system. The browser that is connected to your game system might actually run a combination of JavaScript and AJAX to provide interactivity between the online actions of the player and the ever-changing situation in the database tables, which is being continuously worked on by the background script.

From what I know about the architecture of Apache / PHP and from what I understand from your description, it sounds like you need to use an asynchronous call to a PHP script that does only database work (or otherwise invisible work) in an infinite loop, with a "sleep 500 miliseconds" statement or whatever frequency you want for your housekeeping activities.Apache spawns a new process in the operating system for every PHP script you invoke via browser.The whole thing works on the basis of the request-response paradigm, but for your purposes, you need to separate the request part from the response part.So you have this infinite loop housekeeping script handy and you invoke it with fsockopen once (and only once!!!).fsockopen does the request part, after which you simply omit capturing the response and close your browser window - voila, you have your housekeeping process running in the background.If you invoke it several times, you might have several instances of this script running at the same time, killing your server and stepping on each other's toes, crossing wires etc.

I think PHP function fsockopen() is the answer to your question, check out this article: http://forums.xisto.com/no_longer_exists/ and I'll be back with an architectural explanation of what I think your solution will be.

Here is an article that talks about the growing number of applications Microsoft is releasing with open source access. http://www.linuxjournal.com/node/1000260 I remember listening to a recording of a conference call not very long ago, where a Microsoft employee talked about how Microsoft does not hate the concept of open source per se, they just have a business problem with the GPL. One of the readers who posted a comment to the above linked article calls Microsoft "hypocritical, also ridiculous" for coming up with a way of circumventing GPL called "Shared Source".

The way I solved this problem is by creating a folder on the desktop called iconsfiles where I make a point of moving all the stuff that ends up for some reason or other on my desktop. But I want access to My Computer and Recycle Bin etc... so I don't hide them.

My PC is very loaded so I’m not going to list all the software that’s installed on it, but just the stuff I can’t live without.The PrimoPDF driverAdobe Acrobat ReaderAlcohol 120EZ Macros from American SystemsAntiVir PersonalEdition ClassicantiWord for WindowsApache Web ServerAstyle Java Code BeautifierBitTorrentCacheViewCamStudioCcleanerCMS RapidShare Link GrabberCode InspectorCost Centers PlusCreateCDCygWinDCPlusPlusWink from DebugModeDivX PlayerDU MeterEasy Video JoinereMuleSUPER from eRightSoftFile SplitterFileZilla FTP ServerFlashGetFreecorderAdvanced Administrative ToolsGoogle TalkGoogle Web Acceleratorgsar for Windows (low-level hex search and replace power tool)K-Meleon (browser)Ad-Aware SE Personal from LavasoftFlash MX from MacromediaFixed Assets PlusMagicISOMega Manager from MegauploadMicrosoft OfficeMiro (open source media player)MySQL ServerFireFoxLotus ScreenCam RecorderMplayerncFTPNo-IP Direct Update ClientoDCOracle Express Edition Database ServerPDF2TEXTPerl from ActiveStatePowerISOPrivoxyProcess ExplorerProcess ScannerQuickSliceToad for Oracle from Quest SoftwareRealPlayerRegScrubXPRiva FLV EncoderSiteAdvisorSkypeSlingPlayer from Sling MediaSpyBot Search and DestroySquid for Windows (proxy server software)TCP OptimizertightVNC ServerTotal CommanderTuneUp Utilities 2006TweakMASTERUltra Flash Video FLV ConverterUltra Video JoinerUltraEditUndiskerSpeedUpMyPC from UniblueVLC Player from VideoLAN (I use it very often as a video streaming server)Viewpoint Media PlayerVMware Playerwget for WindowsWinampWinMerge (the original version, the newer ones suck)Yahoo MessengerAnd a piece of software written by a former coworker and good friend of mine, that aims to recreate the effects available in GIMP, but does a better job. The effects it can do are:accented edges, angled strokes, bas relief, chalk and charcoal, charcoal, chrome, colored pencil, conte crayon, craquelure, crosshatch, cutout, dark strokes, diffuse glow, dry brush, film grain, fresco, glass, glowing edges, grain, graphic pen, halftone pattern, ink outlines, mosaic tiles, neon glow, note paper, ocean ripple, paint daubs, palette knife, patchwork, photocopy, plaster, plastic wrap, poster edges, reticulation, rough pastels, smudge stick, spatter, sponge, sprayed strokes, stained glass, stamp, sumi-e, texturizer, torn edges, underpainting, water paper, watercolor, and many others.His program does not yet have a name, but if someone is interested in the pre-alpha version, let me know, he is actually interested in having it thoroughly tested by other people.

Works as advertised, thanks faulty.lee

Not being an HTML designer, but having a requirement to make a web page look pretty, I now have the following problem (I am absolutely convinced the solution for this must be very, very simple, I just need to find the right tutorial, which I couldn't find even after googling with the keywords outsource style section external CSS file):I want to take an inline <style> ... stuff ... </style> section from a web page, create a CSS file out of it, then replace the inline style section with a reference to the CSS file.Take as an example the page you are reading now. How would I be able to accomplish that?I know it's a basic question and I'll keep researching the Internet as we speak, but if anyone has a quick solution to this, it's your chance to earn some hosting credits.

Bill Gates asleep at the wheel.Link: http://www.pcmag.com/article2/0,2817,2168700,00.asp Also, check out this article about piracy: http://forums.xisto.com/no_longer_exists/ It taks about how there is something conspicuously missing from most debate over which operating systems are better - the distortions created by the existence of pirated software. Windows, and Mac OS, are effectively free in the minds of most consumers. Most debate centered around features or stability is eclipsed by the lack of money considerations in assessing user's needs. There are some really good points in this article. I totally agree with the idea that users don't care what OS they are running- all they want is the ability to access something (ex: youtube). Youtube is the reason they have the computer- they don't care at all about what computer it is or how it works. I would even venture to guess that most users don't even know that youtube is a website. Don't laugh... explaining to my mom the difference between reading her email on the web and reading it in a program installed on her computer is an exercise in futility. Architecture (the difference between where a program runs, what runs it, and how efficient that system is) is completely irrelevant in the framework of "I just want to check my email". HOW it works is just something some Geeksquad tech is supposed to know. Just that it works is what most people care about. The vast majority of people don't consider Windows to be free because they pirate it. Most poeple have never installed an OS on their computers in their entire lives. Most pople consider it free because it comes with the computer. So they never pay for it specifically. In some countries, there is no itemized bill when you get a computer that includes the price for Windows. However, the article points out a key factor that has never been part of my paradigm when arguing for Linux. On more than one occasion when stating that "Linux is free" I have been retorted with "Windows is free too -wink wink-" which is hard argue with. But in the end it all comes down to support from developers. And you can only get support from them if you have a user base. And you get a user base by having good software... Linux has something though. A community. Nerds with a Vision. Linux will be made mainstream by the pure will of thousands of nerds, hobbyists, hackers, and geeks. I respect Windows for the time and effort put into it, although it was in vain. Microsoft doesn't care about the community. They care about their bank accounts. But There are many payment options for Linux, if free as in "free beer" is perceived by consumers as a drawback. Tech support, certifications, commercial distros, they all cost money. Of course computer geeks don't need to pay for tech support, their families maybe don't need either, but everyone else will happily pay if that means they won't have to sit clicking GUIs randomly until their problem is fixed (see "Windows"). And anything older than windows ME is freeware on the Microsoft side. The US government ruled that it is okay to crack and pirate (abandonware), or any software that has ceased being supported by its creator. Many people are in the "I'd switch to Linux if it ran my games" crowd still. But as long as Windows maintains dominance, developers, stuck with tight schedules and probably little to no Linux experience, they will generally stick to Windows and DirectX. It's just too annoying for people to jump through the WINE / Cedega hoops to get a subpar experience with the games they own. Considering that advanced Windows users don't see a blue screen of death more than once every few months, the main arguments in that article don't apply all that well to gamers.

Check out this so-called UFO (yeah right): http://forums.xisto.com/no_longer_exists/ Quick edit after reading FirefoxRocks' reply: I found this video on at least two video sharing sites (YouTube, LiveLeak, and others) and I ROFLed when I saw comments on various conspiracy theory sites about how this is a real UFO and so on. One of these conspiracy theory sites is rense.com but I had originally found the link on digg. The video shows some obviously fabricated footage of two flying objects with lights underneath passing overhead in a landscape full of palm trees.

In Linux, BIND is the most widely used DNS server package. However, most people who are just starting out tend find it very difficult to deal with BIND's many options. There is another package that few people have heard of, called dnsmasq - it's a nifty super-lightweight name server, which is very easy to configure. But first of all, let me explain what a name server is: it's a machine that holds a database against which it's performing look-ups in order to translate host names to IP addresses. In DNS jargon, this process of translation is called "resolving". For instance, when you type https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl in the address bar of your browser, your request is mapped against a list of known IP addresses for servers, and the name server returns the actual number for https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl - that, in a nutshell, is what a name server does. In order to better understand the concept, please do the following experiment with me - open a command prompt and type: https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl _linenums:0'>nslookup https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=sslThe result should look like this: https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl Server _linenums:0'># nslookup https://www.google.comserver/: dns-cache-2.<name-of-your-ISP>Address: <IP-address-of-some-DNS-server-at-your-ISP>Non-authoritative answer:Name: http://forums.xisto.com/no_longer_exists/: 209.85.129.147, 209.85.129.99, 209.85.129.104Aliases: https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl#nslookup is this cool little network troubleshooting utility which is embedded in most operating systems nowadays, regardless whether Linux or Windows. nslookup also works the other way around - grab one of the IP addresses from the above output and type this at the command prompt: nslookup 209.85.129.104The result should look like this: # nslookup 209.85.129.104Server: dns-cache-2.<name-of-your-ISP>Address: <IP-address-of-some-DNS-server-at-your-ISP>Name: fk-in-f104.google.comAddress: 209.85.129.104#One of the reasons why it's interesting to run a DNS server at home is that these host-name-IP-address pairs can be cached in order to enable faster lookups - the first time you go out to https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl, it has to go out there and resolve that against the IP address, but then your domestic DNS server can "save that result", and so the next computer on your home network that asks for https://www.google.de/?gfe_rd=cr&ei=BwkjVKfAD8uH8QfckIGgCQ&gws_rd=ssl will find a copy of the IP address right there, handy, and so it's much quicker, so your general web experience is smoother and faster. You can also couple this with a DHCP server, which is a server piece of software that hands out IP addresses to clients as they connect. This function may or may not be built into the firmware of your home router - most modern routers provide it. I am not going to explain how to download and install dnsmasq, just google it - the process should be pretty straightforward. More information about the configuration can be found here: http://forums.xisto.com/no_longer_exists/ The main configuration file is /etc/dnsmasq.conf and the script to start / stop the server is /etc/init.d/dnsmasq. The configuration file is very well commented, it really explains exactly what every single line does, so I would highly recommend going through it and really reading it. In this context, let me first describe a very important configuration file that exists in every Linux system: /etc/resolv.conf This is normally where the IP addresses of your name servers are held, and one of these IP addresses will be picked by nslookup when you invoke it like I described above, it is <IP-address-of-some-DNS-server-at-your-ISP> in this case. On the Linux server where you install dnsmasq, this file needs to contain the IP addresses that your ISP has provided to you - the real ones, so to speak. But on all other Linux machines on your network, which act as DNS clients, this file needs to contain the IP address of the newly installed dnsmasq server on your own local network. As you can see in the article that I linked to above, you need to prepend the string "nameserver " to every IP address which you maintain in any /etc/resolv.conf file. As a side note, you can actually see in the system logs that, when a client requests a DNS lookup, the dnsmasq server is picking up that request and passing it along on behalf of the client. There are a couple of other things which you can do with dnsmasq, for instance you can redirect particular "offending" domain names of advertising companies to a local IP address (127.0.0.1 localhost) in order to defeat ads. I remember from a while ago seeing that this was the topic of another discussion on Xisto, but the solution was to change the hosts file on every computer on your network as opposed to doing it once for everybody. One other thing you can do is foil SiteFinder type abusive practices. Read the article I linked above for more information on Sitefinder.

Check this other one out. "Virtualize Windows on Linux? Microsoft Says No Way!" Link: http://www.pcgamer.com/hardware/ Apparently the stakes are high. Microsoft has a history of making their competitors' products intentionally slower - in this case the roles are reversed: the competitor is the host, and Microsoft is the guest.

Panda required me to download and install an ActiveX plug-in. It's getting updates as we speak. I'll let you know of anything else.

Google now has a new program called "Business Referral Representative". They will pay you $10 per store for walking in, taking pictures, determining store hours, payment methods, etc... More on this at: https://www.google.com/services/local-businals/repfaq.html

Just found this on digg: http://forums.xisto.com/no_longer_exists/ At the bottom of the article there is a link to a video of him presenting his vision about the future of virtualization. He's describing a business environment where all safe computing (business and productivity applications) will happen on the desktop, and for all activities which are considered unsafe (browsing the web) employees will simply create a disposable installation of an OS on some server, connect to it (presumably using something like the VNC viewer, although VNC isn't specifically mentioned in this speech) do the required browsing, and if they happen to pick up any malware in the process, it will only affect this particular instance of a virtual OS. He also shows an example of a SLED (SuSe Linux Enterprise Desktop) laptop simultaneously running a virtual copy of Windows XP along with a virtual copy of Vista. What I don't get is how you can run Vista in a virtual box since Vista is supposed to be built on the principles of trusted computing, which means severe checks on hardware components.

Not all databases were created equal.There are what I call "pocket databases" (MS Access, MySQL, PostgreSQL) and then there are enterprise-strength databases (Oracle, DB2 Universal Database, SQL Server, Informix, Sybase).One of the first companies that got involved in developing database software was IBM, and the main driver for doing this was to sell more mainframes, not to provide reliable, well-performing information systems to their customers. That, and the fact that they had fallen in love with their own database product (DB2/360 for mainframe) led to IBM falling behind the growth curve of companies' (their own customers) data volumes.Instead of taking a hard look at the emerging performance problems their customers were beginning to experience and going back to the drawing board and designing from the ground up a truly scalable database product based on a radically new architecture, they added gizmos and gimmicks to the already existing code base.Oracle came along and capitalized on this opportunity by inventing the concept of rollback segments. Not many people realize this, but having these rollback segments was and continues to be the reason for Oracle's unparalleled success.Rollback segments are basically a bunch of files that store the before and after images of data when you issue so-called DML statements (update, insert, delete).Rollback segments are the hidden reason why the same application will scale to terabytes when running on Oracle, and will screech to a halt as it reaches the 500 GB mark when running on pretty much any other database software.I know this from first hand experience, being a technical SAP professional. I have personally seen many SAP systems, both R/3 (for OLTP / online transaction processing) and BW (Business Information Warehouse, for OLAP / online analytical processing) and I have noticed a pattern of the worst performance problems always happening on SAP systems which were sitting on non-Oracle databases.For the sake of disclosure, I do not work for Oracle, never have, never will, and I do not hold shares in their stock.But if it sounds like I'm bashing IBM and their DB2 UDB product, I'm not going to apologize. If a company is considering purchasing DB2 UDB, they might as well get the same level of performance and scalability by going with MySQL and saving a bunch of money in the process.The worst performance problem I have ever seen was when issuing a 3-way table join in a pretty big DB2 UDB database, where the 3 tables involved were around 2 GB in size each. Not only would this SELECT statement take forever, but we managed to bring the server to its knees just by running this 3-way join (the server being a powerful RS6000 machine running AIX). The way we ended up solving this problem was to break down the 3-way join into 2 separate SELECTs and passing a result set drom one to the other, but that ugly and against all database design best practices. However, the results were spectacular. The end users came to us and asked "What did you guys do? Did you turbo-charge the database?" and our answer was "No, but next time you have a nearly unsolvable database performance problem, get Redwood Shores on the phone." OK, enough of my soapbox, I'm calm now.

Just found this on digg: http://www.slyck.com/story1566_BitTorrent_d_Source_Issues My comments: If this doesn't smell like a backroom deal with the RIAAs and the MPAAs of the world, I don't know what does. Thoughts?

I need to find two pre-existing PHP applications and install them on my site:1. A quiz-type knowledge testing application.2. A simple poll with a world-readable IP-based voting log.The quiz script should provide the option to create questions of the following types:- single-choice (5 or more radio buttons)- multiple choice (5 or more checkboxes)- true / false- answer by entering free text (case insensitive)and create a test for someone to take, by selecting individual questions or categories of questions.Has anyone successfully used these kind of applications before?If so, can you point me to where I can get them?

What would be a good way to tell which fonts my PC never used (or only rarely used) ?By looking inside C:\WINDOWS\Fonts I can tell when they were last changed, not when they were last used by an application.Also, has anyone noticed a significant performance improvement from deinstalling fonts?

I looked at the movies and I have to say I'm not one for the eye candy.In fact, for some strange reason, it makes me sick to my stomach.I might need counseling.If I'm ever forced into a situation where I have to use Vista or Vienna, I will probably have to spend 1 month getting rid of the glitzy-glamorous stuff and removing obstacles to performance.I hope Windows XP will continue to exist for a long time to come and that they don't find a slick way to shove Vista or Vienna down people's throats.