Search Results

Search found 26978 results on 1080 pages for 'load testing'.

Page 332/1080 | < Previous Page | 328 329 330 331 332 333 334 335 336 337 338 339  | Next Page >

• How to troubleshoot a wireless networking regression?

  - by fluteflute

  I've been experiencing this perhaps slightly odd bug. It works flawlessly in Lucid, but not in Maverick and Natty. I find it seems to work when I'm booting a partition everyday (as I do for my main 10.10 partition) but for my 11.04 testing partition it's a real pain - usually refusing to connect. So given that I have both a working (10.04) and not-working (10.10 and 11.04) installs, how can I troubleshoot my problem?

  Read the article

• How to resolve concurrent ramp collisions in 2d platformer?

  - by Shaun Inman

  A bit about the physics engine: Bodies are all rectangles. Bodies are sorted at the beginning of every update loop based on the body-in-motion's horizontal and vertical velocity (to avoid sticky walls/floors). Solid bodies are resolved by testing the body-in-motion's new X with the old Y and adjusting if necessary before testing the new X with the new Y, again adjusting if necessary. Works great. Ramps (rectangles with a flag set indicating bottom-left, bottom-right, etc) are resolved by calculating the ratio of penetration along the x-axis and setting a new Y accordingly (with some checks to make sure the body-in-motion isn't attacking from the tall or flat side, in which case the ramp is treated as a normal rectangle). This also works great. Side-by-side ramps, eg. \/ and /\, work fine but things get jittery and unpredictable when a top-down ramp is directly above a bottom-up ramp, eg. < or > or when a bottom-up ramp runs right up to the ceiling/top-down ramp runs right down to the floor. I've been able to lock it down somewhat by detecting whether the body-in-motion hadFloor when also colliding with a top-down ramp or hadCeiling when also colliding with a bottom-up ramp then resolving by calculating the ratio of penetration along the y-axis and setting the new X accordingly (the opposite of the normal behavior). But as soon as the body-in-motion jumps the hasFloor flag becomes false, the first ramp resolution pushes the body into collision with the second ramp and collision resolution becomes jittery again for a few frames. I'm sure I'm making this more complicated than it needs to be. Can anyone recommend a good resource that outlines the best way to address this problem? (Please don't recommend I use something like Box2d or Chipmunk. Also, "redesign your levels" isn't an answer; the body-in-motion may at times be riding another body-in-motion, eg. a platform, that pushes it into a ramp so I'd like to be able to resolve this properly.) Thanks!

  Read the article

• How to switch your small or home office to Linux

  <b>Tech Radar:</b> "With Linux and free software making a name for itself in the world of big business, many people are testing the feasibility of switching small and home office software to their open source equivalents."

  Read the article

• organization of DLL linked functions

  - by m25

  So this is a code organization question. I got my basic code working but when I expand it will be terrible. I have a DLL that I don't have a .lib for. Therefore I have to use the whole loadLibrary()/getprocaddress() combo. it works great. But this DLL that i'm referencing at 100+ functions. my current process is (1) typedef a type for the function. or typedef short(_stdcall *type1)(void); then (2) assign a function name that I want to use such as type1 function_1, then (3) I do the whole LoadLibrary, then do something like function_1 = (type1)GetProcAddress(hinstLib, "_mangled_funcName@5"); normally I would like to do all of my function definitions in a header file but because I have to do use the load library function, its not that easy. the code will be a mess. Right now i'm doing (1) and (2) in a header file and was considering making a function in another .cpp file to do the load library and dump all of the (3)'s in there. I considered using a namespace for the functions so I can use them in the main function and not have to pass over to the other function. Any other tips on how to organize this code to where it is readable and organized? My goals are to be able to use function_1 as a regular function in the main code. if I have to a ref::function_1 that would be okay but I would prefer to avoid it. this code for all practical purposes is just plane C at the moment. thanks in advance for any advice!

  Read the article

• Google I/O 2010 - A JIT Compiler for Android's Dalvik VM

  Google I/O 2010 - A JIT Compiler for Android's Dalvik VM Google I/O 2010 - A JIT Compiler for Android's Dalvik VM Android 301 Ben Cheng, Bill Buzbee In this session we will outline the design of a JIT Compiler suitable for embedded Android devices. Topics will include an architectural overview, the rationale for design decisions and the special support for JIT verification, testing and tuning. For all I/O 2010 sessions, please go to code.google.com From: GoogleDevelopers Views: 3 0 ratings Time: 01:00:14 More in Science & Technology

  Read the article

• Opera12 browser support for WebGL

  - by Sneha

  I have enabled - enable webGL by entering Enable WebGL in opera:config by setting the value to 1, and Enable Hardware Acceleration similarly, then restart the browser - Opera 12 on Windows & tried to open - http://aleksandarrodic.com/p/jellyfish/. This gives me an alert saying - Your browser failed to initialize WebGL. Please guide me on - What's going wrong? opera:gpu says : Hardware acceleration Disabled Direct3D 10 Backend not supported Could not load library OpenGL Backend not supported Could not load library But,same thing (http://aleksandarrodic.com/p/jellyfish/) works fine on Chrome! Thanks Sneha

  Read the article

• Website Ethics / legal issues, image copyrights

  - by RailsN00b

  Ignoring the technical implementation of a website for a second, assume a website that is similar to twitter but with pictures. A user say something and puts a picture of whatever they said. As the nature of the internet, the images will most likely not be his/hers image. There are 2 options that I see for dealing with this: 1. The user will post a URL of the picture and the website will pull the picture from that URL everytime someone enters that page 2. The website will save the image in its own database of images and display the image to the visitors 'locally' The problem with option #1, while it saved storage, I see an issue with 'stealing' other websites bandwidth and if my website has many many visitors it could cost the image-hosting websites a lot and possible even crash it if the server can't handle the load. The problem with opion #2, while it saves the load to other websites, it practically takes pictures that could have copyright on them. Which option is better to implement in terms of legal issues and ethics? When do I need to contact another website to request permission to use the images from that site? Does anyone really care about that anymore. Where can I read about this?

  Read the article

• Temp profile used when User logs in

  - by TJ

  One of my users logged into his computer, Windows XP, last night only to be meet with an error message that it could not load his profile and he will be logged in using a temporary profile. Typically when this happens I shut the machine down and restart and the correct profile will load when they log in again. Not this time. In the user profile options under computer-properties-Advanced-user profiles it show that there are three profiles with his name. Two are the exact same size with the same modified date (5/5/10) and the other is what I would expect size wise for a new profile with a modified date of today. What are my options to restore his profile?

  Read the article

• What would be better in my case - apache, nginx or lighttpd ?

  - by The Devil

  Hey everybody, I'm writing a php site that's expected to get about 200-300 concurrent users browsing it. When initializing the application will load about 30 php classes, some 10 maybe 15 images and a couple of css files. So my question is what else can I do (except optimizing my code and using apc/eaccelerator for php) to get as close as possible to those numbers of concurrent users ? Currently we haven't chosen a server for the site to be hosted on but most probably it'll be a VPS Dual core + 2 or maybe 4gb ram. Is it possible for such a server to handle that load ? Also how could I test it myself and be sure that it'll be able to handle it ? Thanks in advance, Me

  Read the article

• Chrome developer tools - network panel gaps

  - by Chris Nicholson

  In the Chrome developer tools, under the network tab, I'm curious to know what is happening during the gaps. If you look at my image below, I have highlighted in orange the areas where these gaps exist. Where I'm able to load a lot of my page from cache it's a shame these large gaps occur as they make up most of my page load time. What exactly is happening in this time? EDIT Okay I found this answer which essentially sums up my question, so a different question: does anyone know a good method to reduce the length of these gaps? Presumably (albeit rather extreme) if I loaded all my CSS on the page there wouldn't be a delay after loading the CSS file before the images were loaded.

  Read the article

• With AMD style modules in JavaScript is there any benefit to namespaces?

  - by gman

  Coming from C++ originally and seeing lots of Java programmers doing the same we brought namespaces to JavaScript. See Google's closure library as an example where they have a main namespace, goog and under that many more namespaces like goog.async, goog.graphics But now, having learned the AMD style of requiring modules it seems like namespaces are kind of pointless in JavaScript. Not only pointless but even arguably an anti-pattern. What is AMD? It's a way of defining and including modules that removes all direct dependencies. Effectively you do this // some/module.js define([ 'name/of/needed/module', 'name/of/someother/needed/module', ], function( RefToNeededModule, RefToSomeOtherNeededModule) { ...code... return object or function }); This format lets the AMD support code know that this module needs name/of/needed/module.js and name/of/someother/needed/module.js loaded. The AMD code can load all the modules and then, assuming no circular dependencies, call the define function on each module in the correct order, record the object/function returned by the module as it calls them, and then call any other modules' define function with references to those modules. This seems to remove any need for namespaces. In your own code you can call the reference to any other module anything you want. For example if you had 2 string libraries, even if they define similar functions, as long as they follow the AMD pattern you can easily use both in the same module. No need for namespaces to solve that. It also means there's no hard coded dependencies. For example in Google's closure any module could directly reference another module with something like var value = goog.math.someMathFunc(otherValue) and if you're unlucky it will magically work where as with AMD style you'd have to explicitly include the math library otherwise the module wouldn't have a reference to it since there are no globals with AMD. On top of that dependency injection for testing becomes easy. None of the code in the AMD module references things by namespace so there is no hardcoded namespace paths, you can easily mock classes at testing time. Is there any other point to namespaces or is that something that C++ / Java programmers are bringing to JavaScript that arguably doesn't really belong?

  Read the article

• Windows 8 BIOS - Boot Ubuntu from External HDD

  - by F3AR3DLEGEND

  My laptop came pre-loaded with Windows 8 64-bit (only storage device is a 128 GB SSD). Since it is my school laptop/I've heard creating a Linux partition alongside Windows 8 is not very wise I installed Ubuntu onto my external hard drive. I have a 500GB external HDD with the following partitions: Main Partition - NFTS - ~400 GB Extension Partition / - ext2 - ~25gb /home - ext2 - ~30gb swap - ext2 - 10gb /boot - ? - 10gb ? = not sure of partition Using the PenDriveLinux installer, I created a LiveUSB version of Ubuntu 12.04 (LTS) on a 4GB USB drive. Using that, I installed Ubuntu onto the external hard-drive, without any errors (or at least none that I was notified of). Using the BIOS settings, I changed the OS-loading order so that it is in this order: My External USB HDD Windows Boot Loader Some other things Therefore, Ubuntu should load from my hard drive first, but it doesn't. Also, my hard drive is in working condition, and it turns on when BIOS starts (there is a light indicator). When I start my laptop, it goes directly to Windows 8 (I have the fast startup setting disabled as well). So, is there any way for me to set it up so that when my HDD is connected, it will automatically load Ubuntu? Thanks in advance!

  Read the article

• Getting the PC speaker to beep

  - by broiyan

  There has been much written on getting the beep sound from Ubuntu releases over the years. Example: fixing the beep My needs are slightly different in that I do not want to ensure sound card beeps are functioning. Instead, I want PC speaker beeps, the kind produced by the original built-in speaker because I believe they will produce less CPU load. I have confirmed that my computer has the PC speaker by unplugging the external speakers and shutting down Ubuntu. At some point in the shutdown and restart process a beep is heard even though the external speakers have no power. I have tried the following: In /etc/modprobe.d/blacklist.conf, turn these lines into comments: #blacklist snd_pcsp #blacklist pcspkr In .bashrc /usr/bin/xset b on /usr/bin/xset b 100 Enable in the gnome terminal: Edit Profile Prefs General Terminal Bell Ensure no "mute" selections in: System Prefs Sound various tabs (uncheck them all). Select "Enable window and button sounds" in: System Prefs Sound Sound Effects In gconf-editor desktop gnome sound, select the three sound check boxes. In gconf-editor apps metacity general select the audible bell check box. Still I get no PC speaker beeps when I send code 7 to the console via my Java program or use echo -e '\a' on the bash command line. What else should I try? Update Since my goal is to minimize load on the CPU, here is a comparison of elapsed times. Each test is for 100,000 iterations. Each variant was performed three times so three results are presented for each. printwriter.format("%c", 7); // 1.3 seconds, 1.5 seconds, 1.5 seconds Toolkit.getDefaultToolkit().beep(); // 0.8 seconds, 0.3 seconds, 0.5 seconds try { Runtime.getRuntime().exec("beep"); } catch (IOException e) { } // 10.3 seconds, 16.3 seconds, 11.4 seconds These runs were done inside Eclipse so multiply by some value less than 1 for standalone execution. Unfortunately, Toolkit's beep is silent on my computer and so is code 7. The beep utility works but has the most cost.

  Read the article

• Byldan

  - by csharp-source.net

  Byldan is a framework for managing the build life-cycle of .NET applications. Its goal is to support multiple-platforms (Linux/Windows) and multiple compiler vendors (Novell/Microsoft). This minor release of Byldan adds support for unit testing with NUnit and for signing of assemblies.

  Read the article

• Random users randomly being unable to connect to my static content domain

  - by jls33fsls

  I store all of my images, js, and css files on a separate domain to try and speed up page load times (it isn't a CDN, just a separate domain on the same server). This works fine for 99% of the users, 99% of the time. However, there are users that randomly are unable to connect to the static content domain for periods of 1-5 hours. They can go to the main site, but no images will load and everything is just white because no css is being loaded. If they go to the static content domain itself, the page just idles for a while and then times out with a blank white page, no error messages. I have no idea what could be causing this, and it hasn't happened to me, any ideas? I am running Apache on CentOS 5.5.

  Read the article

• Snow Leopard, PHP, and MySQL

  - by Peter

  I have just installed Snow Leopard and now my PHP/MySQL program ends in a "Segmentation fault". I have been searching the web for a solution, I realize that there are some issues with SL/PHP/MySQL, but I have not found anything that works yet. I downloaded the binary MySQL package mysql-5.1.42-osx10.6-x86_64. I have updated the php.ini file as suggested at various posts. When I run PHP and connect to the MySQL server the behavior is a bit random. In many cases it works fine to connect and read data. In my specific case the PHP program constructs a LOAD DATA LOCAL INFILE ... statement to load data from a text file. It should do several such queries after each other in a loop. It works one time but the halts in a "Segmentation fault". The program worked fine in Leopard, but not now. My versions are: OS 10.6.2, PHP 5.3.0, MySQL 5.1.42

  Read the article

• What guidelines are best suited for leveraging automatic deployments?

  - by Scott

  We are hoping to leverage a static code analysis tool (Sonar) as part of our continuous integration server, and are hoping to determine some useful guidelines to serve as a base for allowing the deployment to continue. What conditions should we make mandatory before allowing a build to proceed to the next set of testing? The obvious answers include that it compiles and the unit tests are successful. But what are some other things we should require before allowing a build to not be rolled back?

  Read the article

• vpn/Openvpn as a cloud service

  - by 8pipe

  I am working on creating a small cloud (any number of EC2 instances that can be deployed based on load) implementing a VPN as a service for the company I'm working for. This is basically a project gathering together various vpn resources under one aegis as a cloud based service. As a user of openvpn, I'm somewhat familiar with being able to connect, but I'm looking for resources to start this project. Essentially I need to be able to: run a certificate authority and manage keys to distribute to coworkers build an ami that handles openvpn as a service balance the load if necessary among machines instances as needed Any suggestions for tutorials, things to avoid, roadblocks I might not be seeing from a novice perspective, etc. or just help in visualizing this is appreciated.

  Read the article

• Debian Linux server hangs after a week or so

  - by Alex Flo

  I have 2 Debian Linux 6.0.4 servers that have a strange behaviour: after 5-7-10 days they hang. By this I mean the servers need to be restarted and before that ping won't answer. I've been struggling with this problem for a couple of months now and here's some thoughts/what I tried without being able to solve the problem. I changed the RAM on a server. Being 2 different servers I doubt that it could be something related to hardware as a 3rd identical server won't have this problem. I logged the server load and when it crashes the load is fine (quite low) I cannot find anything in the server logs, logs are fine till the server freezes. I don't have access to console unfortunately. While I have years of admin experience I have never encountered such an issue and right now I have no idea where else to investigate. If you have an idea of what I could try in order to fix the problem please share it with me:-)

  Read the article

• Ubuntu 10.04 hung on unattended apt-get upgrade?

  - by hafichuk

  I'm looking into why our ubuntu web server hung this morning and see that there was some package upgrades a few hours prior to the problem. I was able to ssh into the system and get a snapshot from top: top - 08:13:54 up 210 days, 8:25, 2 users, load average: 433.30, 422.40, 375.70 Tasks: 1192 total, 381 running, 810 sleeping, 0 stopped, 1 zombie Cpu(s): 0.5%us, 6.1%sy, 0.0%ni, 93.4%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 49549772k total, 48518392k used, 1031380k free, 960152k buffers Swap: 11595768k total, 279368k used, 11316400k free, 39355664k cached This is a 16 processor system, so I would typically expect a load in the low teens. I tried to restart apache, which didn't work, and subsequently had to do a hard reboot to get it working again (which it is). One thing I found was that the server did an unattended package update. Is it possible that upgrading php or curl (which our web sites use) might have caused apache to stop responding? Here is the snip from the unattended-upgrades.log from this morning: 2012-09-18 06:48:30,076 INFO Initial blacklisted packages: 2012-09-18 06:48:30,076 INFO Starting unattended upgrades script 2012-09-18 06:48:30,077 INFO Allowed origins are: ["['Ubuntu', 'lucid-security']"] 2012-09-18 06:49:37,017 INFO Packages that are upgraded: gnupg-curl php5-dev linux-server dhcp3-common linux-libc-dev php5-curl gpgv gnupg linux-headers-server linux-image-server php5 php5-mysql php-pear php5-cli php5-common libapache2-mod-php5 dhcp3-client 2012-09-18 06:49:37,018 INFO Writing dpkg log to '/var/log/unattended-upgrades/unattended-upgrades-dpkg_2012-09-18_06:49:37.017909.log'

  Read the article

• Incorporating libs into module pattern

  - by webnesto

  I have recently started using require.js (along with Backbone.js, jQuery, and a handful of other JavaScript libs) and I love the module pattern (here's a nice synopsis if you're unfamiliar: http://www.adequatelygood.com/2010/3/JavaScript-Module-Pattern-In-Depth). Something I'm running up against is best practices on incorporating libs that don't (out of the box) support the module pattern. For example, jQuery without modification is going to load into a global jQuery variable and that's that. Require.js recognizes this and provides an example project for download with a (slightly) modified version of jQuery to incorporate with a require.js project. This goes against everything I've ever learned about using external libs - never modify the source. I can list a ton of reasons. Regardless, this is not an approach I'm comfortable with. I have been using a mixed approach - wherein I build/load the "traditional" JS libraries in a "traditional" way (available in the global namespace) and then using the module pattern for all of my application code. This seems okay to me, but it bugs me because one of the real beauties of the module pattern (no globals) is getting perverted. Anyone else got a better solution to this problem?

  Read the article

• File store: CouchDB vs SQL Server + file system

  - by Andrey

  I'm exploring different ways of storing user-uploaded files (all are MS Office documents or alikes) on our high load web site. It's currently designed to store documents as files and have a SQL database store all metadata for those files. I'm concerned about growing out of the storage server and SQL server performance when number of documents reaches hundreds of millions. I was reading a lot of good information about CouchDB including its built-in scalability and performance, but I'm not sure how storing files as attachments in CouchDB would compare to storing files on a file system in terms of performance. Anybody used CouchDB clusters for storing LARGE amounts of documents and in high load environment?

  Read the article

• PHP hits 100% CPU and eats RAM at the same time Monday to Friday

  - by Daniel Samuels

  We run a learning platform for primary schools here in the UK and it's all been running extremely well. However at around 4PM Monday to Friday we see the same issue arise -- 1-2 PHP threads will spike to 100% CPU and gradually start eating up RAM until the server(s) fall over. 98%+ of our requests are HTTPS, these come into our Layer 7 load balancer which then decrypts the SSL data, adds the X-HTTP-Forwarded-For header and forwards the data onto an application server (we have 2 of those at the moment) on port 80. Our application servers have Varnish on port 80 which takes in the request from the load balancer and passes the request through to Nginx on port 81. Nginx then works out which 'vhost' it needs to use and passes any PHP processing through to PHP-CGI which is listening on a socket (managed through spawn-fcgi). There's an instance of Memcached running too, MySQL runs on a separate server / slave setup. Throughout the day the load will typically go no higher than 0.8 on either of the application servers, however at around 4PM our problem arises. I've managed to run strace on a few of the actual threads when they cause the problem and I always see the same thing: stat("/usr/share/zoneinfo/Europe/London", {st_mode=S_IFREG|0644,st_size=3661, ...}) = 0 stat("/usr/share/zoneinfo/Europe/London", {st_mode=S_IFREG|0644,st_size=3661, ...}) = 0 This is repeated infinitely and never stops until you SEGKILL the process or oomkiller kills it. There are no cron jobs scheduled to run at that time and I don't have any way of seeing exactly what Nginx request is associated with the PHP process which is running. We are running PHP 5.3.14 which we upgraded to from 5.3.8 last week to rule out the older version being the problem. This issue has been going on a few months now and we have no idea what is causing it. We deploy our software very frequently, so it's difficult to track down a specific release which may have started the problem - especially as we do not know the date of the first occurrence of this issue. Varnish is version 3.0.1, Nginx is 1.0.6 (which I understand is about a year old now), our servers are running CentOS release 5.7 (Final) they have Intel i3 540s at 3.07Ghz and 8GB of RAM. There's a discussion on the Debian mailing list about something very similar, you can find that here. Has anyone seen anything like this in the past, does anyone have any ideas or suggestions? Are there a way of linking an Nginx request directly to a PHP thread? Is there a better way of seeing what the PHP process is doing? (I've seen GDB mentioned, though I'll have to recompile PHP) Thanks!

  Read the article

• Coldfusion server VERY slow page loads

  - by Kevin

  I inherited a windows server 2003 coldfusion 7 server a few weeks ago. Today a network cable was unplugged by accident from the server. On plugging it back in, pages were NOT loading at all. Rather, we were receiving a generic coldfusion error page. After restarting IIS several times and coldfusion even more than that, we finally got pages to start loading. However, the loading is extremely slow (30+ seconds) on pages that used to load instantly. Loading through the local network (IE localhost/cfide/administrator) does nothing to help the load speed. I am not familiar with IIS or Coldfusion (We're in the process of migrating this to Linux/PHP), so this is all new territory to me. I'm hoping someone may have experienced this issue in the past and can help me solve it. I'm happy to provide any additional information that might be necessary....I'm just not sure what information you might need in order to help. Thanks for your time.

  Read the article

• How do I maintain a really poorly written code base?

  - by onlineapplab.com

  Recently I got hired to work on existing web application because of NDA I'm not at liberty to disclose any details but this application is working online in sort of a beta testing stage before official launch. We have a few hundred users right now but this number is supposed to significantly increase after official launch. The application is written in PHP (but it is irrelevant to my question) and is running on a dual xeon processor standalone server with severe performance problems. I have seen a lot of bad PHP code but this really sets new standards, especially knowing how much time and money was invested in developing it. it is as badly coded as possible there is PHP, HTML, SQL mixed together and code is repeated whenever it is necessary (especially SQL queries). there are not any functions used, not mentioning any OOP there are four versions of the app (desktop, iPhone, Android + other mobile) each version has pretty much the same functionality but was created by copying the whole code base, so now there are some differences between each version and it is really hard to maintain the database is really badly designed, which is causing severe performance problems also for fixing some errors in PHP code there is a lot of database triggers used which are updating data on SELECT and on INSERT so any testing is a nightmare Basically, any sin of a bad programming you can imagine is there for example it is not only possible to use SQL injections in literally every place but you can log into app if you use a login which doesn't exist and an empty password. The team which created this app is not working on it any more and there is an outsourced team which suggested that there are some problems but was never willing to deal with the elephant in the room partially because they've got a very comfortable contract and partially due to lack of skills (just my opinion). My job was supposed to be fixing some performance problems and extending existing functionality but first thing I was asked to do was a review of the existing code base. I've made my review and it was quite a shock for the management but my conclusions were after some time finally confirmed by other programmers. Management made it clear that it is not possible to start rewriting this app from scratch (which in my opinion should be done). We have to maintain its operable state and at the same time fix performance errors and extend the functionality. My question is, as I don't want just to patch the existing code, how to transform this into properly written app while keeping the existing code working at the same time? My plan is: Unify four existing versions into common code base (fixing only most obvious errors). Redesign db and use triggers to populate it with data (so data will be maintained in two formats at the same time) All new functionality will be written as separate project. Step by step transfer existing functionality into the new project After some time everything will be in the new project Some explanation about #2, right now it is practically impossible to make any updates in existing db any change requires reviewing whole code and making changes in many places. Is such plan feasible at all? Another solution is to walk away and leave the headache to someone else.

  Read the article

< Previous Page | 328 329 330 331 332 333 334 335 336 337 338 339  | Next Page >