Search Results

Search found 33445 results on 1338 pages for 'single instance storage'.

Page 420/1338 | < Previous Page | 416 417 418 419 420 421 422 423 424 425 426 427  | Next Page >

• How can i handle a form submit using REAL OOP in PHP

  - by Lorence

  Im used to java and creating UML.. and i was wondering how can PHP be OOP, the objects live only until you make a request.. then they destroy, so if im using a database is useless to create a class and add the members (variables) to the class, they will be useless.. i cant pass the main system object from one page to another, or similar so how can PHP be compare to jave? you never do OOP .. i mean REAL OOP.. not creating classes , in fact your index will be a procedural file with some object instance and then ? how about if i make a html form and i want to submit the data.. i have to call a file which is not a class is a php procedural file were i grab the submited data with POST, from that file you will instance a class and do some logic there.. but for me thats not pure OOP.. can somebody point me to the right way of OOP using a form submit example ? Thanks!

  Read the article

• Why is Selenium RC so slow?

  - by Pete

  Hi. For some time I have been investigating Selenium RC in order to do functional testing of my web application. I have now found a test strategy that is so effective, that I do not want to move away from Selenium RC (after spending weeks trying to figure out a good way to validate ASP.NET validation controls). But now that my Selenium RC adventure is moving from a POC to be something that I actually use, I'm running into a problem. It is insanely slow. Executing a single test that loads a page, fills in some fields, and clicks a button takes in the magnitude of seconds to execute. When it is executing, I can easily see each individual field being filled out one at a time. Using Selenium IDE in Firefox is not that slow. I found this page, that clearly specifies that Selenium RC is slow http://selenium-grid.seleniumhq.org/how_it_works.html But why is that? Is it because the browser is polling the selenium server? If so, can this polling interval not be modified? Or is there another reason. I am not accustomed to a remote call taking a humanly noticable amount of time to execute. It is horrible that executing a few tests should take so long. I can execute my entire presentation (MVP), business, and database layer test suite (500+ tests) way quicker than it takes to run 10 tests for a single web page.

  Read the article

• Recover a lost IBM DS4300 SAN password?

  - by Daniel

  I have a pair of IBM DS4300 SAN units that I need to perform a firmware upgrade on. Unfortunately the admin passwords on these units have been lost and now need to be reset. I had hoped the default password of 'infiniti' would work but it seems that it must have been changed. I know the method here is to contact IBM but the cost of that phone call is ludicrous if someone out here knows what they're going to tell me to do. These units are out of warranty and the cost of support is beyond what we have the budget for. Is there an interrupt I can enter during the boot process (similar to a cisco password recovery) or is there a hardware/software tool I require? Please, if anyone has ever gone through the process of an IBM password recovery I’m asking for a little help. EDIT: Just to clarify, the password I need to reset is the one used for serial cable access and not for storage manager. Sorry if I have caused more confusion.

  Read the article

• Swig typecast to derived class?

  - by Zack

  I notice that Swig provides a whole host of functions to allow for typecasting objects to their parent classes. However, in C++ one can produce a function like the following: A * getAnObject() { if(someBoolean) return (A *) new B; else return (A *) new C; } Where "A" is the parent of classes "B" and "C". One can then typecast the pointer returned into being a "B" type or "C" type at one's convenience like: B * some_var = (B *) getAnObject(); Is there some way I can typecast an object I've received from a generic-pointer-producing function at run-time in the scripting language using the wrappers? (In my case, Lua?) I have a function that could produce one of about a hundred possible classes, and I'd like to avoid writing an enormous switch structure that I'd have to maintain in C++. At the point where I receive the generic pointer, I also have a string representation of the data type I'd like to cast it to. Any thoughts? Thanks! -- EDIT -- I notice that SWIG offers to generate copy constructors for all of my classes. If I had it generate those, could I do something like the following?: var = myModule.getAnObject(); -- Function that returns an object type-cast down to a pointer of the parent class, as in the function getAnObject() above. var = myModule.ClassThatExtendsBaseClass(var); -- A copy constructor that SWIG theoretically creates for me and have var then be an instance of the inheriting class that knows it's an instance of the inheriting class?

  Read the article

• serializing type definitions?

  - by Dave

  I'm not positive I'm going about this the right way. I've got a suite of applications that have varying types of output (custom defined types). For example, I might have a type called Widget: Class Widget Public name as String End Class Throughout the course of operation, when a user experiences a certain condition, the application will take that output instance of widget that user received, serialize it, and log it to the database noting the name of the type. Now, I have other applications that do something similar, but instead of dealing with Widget, it could be some totally random other type with different attributes, but again I serialize the instance, log it to the db, and note the name of the type. I have maybe a half dozen different types and don't anticipate too many additional ones in the future. After all this is said and done, I have an admin interface that looks through these logs, and has the ability for the user to view the contents of this data thats been logged. The Admin app has a reference to all the types involved, and with some basic switch case logic hinged upon the name of the type, will cast it into their original types, and pass it on to some handlers that have basic display logic to spit the data back out in a readable format (one display handler for each type) NOW... all this is well and good... Until one day, my model changed. The Widget class now has deprecated the name attribute and added on a bunch of other attributes. I will of course get type mismatches in the admin side when I try to reconstitute this data. I was wondering if there was some way, at runtime, i could perhaps reflect through my code and get a snapshot of the type definition at that precise moment, serialize it, and store it along with the data so that I could somehow use this to reconstitute it in the future?

  Read the article

• ExtJS: dynamically ajust layout after adding/removing some form fields

  - by Sergei Stolyarov

  I have form layout with some TextField elements and one HtmlEditor element. Some TextField elements are "hideable", i.e. could be hid or showed. After hiding/showing elements HtmlEditor instance break layout — there appear empty space or element doesn't end at the window border. Is it possible to tell to HtmlEditor instance use all remaining available space? Even in case when some elements are hidden/showed. I've tried to use anchor property, but it works well until some element removed from the layout. Updated Here is a sample code: var htmlEditor = new Ext.form.HtmlEditor({ anchor: '100% -54', hideLabel: true }); var fp = new Ext.form.FormPanel({ items: [{xtype: 'textfield', fieldLabel: 'zzz', mode: 'local'}, {xtype: 'textfield', fieldLabel: 'nnn', id: 'id-one', mode: 'local'}, htmlEditor] }); var w = new Ext.Window({layout: 'fit', height: 400, width: 600, tbar: [{text: 'click', handler: function() { // hide element Ext.getCmp('id-one').getEl().up('.x-form-item').setDisplayed(false); w.doLayout(); } }], items: fp }); w.show(); Execute, click toolbar button "click", one field should disappear, then look at empty space below htmleditor.

  Read the article

• How to manage Javascript modules in django templates?

  - by John Mee

  Lets say we want a library of javascript-based pieces of functionality (I'm thinking jquery): For example: an ajax dialog a date picker a form validator a sliding menu bar an accordian thingy There are four pieces of code for each: some Python, CSS, JS, & HTML. What is the best way to arrange all these pieces so that: each javascript 'module' can be neatly reused by different views the four bits of code that make up the completed function stay together the css/js/html parts appear in their correct places in the response common dependencies between modules are not repeated (eg: a javascript file in common) x-------------- It would be nice if, or is there some way to ensure that, when called from a templatetag, the templates respected the {% block %} directives. Thus one could create a single template with a block each for CSS, HTML, and JS, in a single file. Invoke that via a templatetag which is called from the template of whichever view wants it. That make any sense. Can that be done some way already? My templatetag templates seem to ignore the {% block %} directives. x-------------- There's some very relevant gasbagging about putting such media in forms here http://docs.djangoproject.com/en/dev/topics/forms/media/ which probably apply to the form validator and date picker examples.

  Read the article

• File Translator to Export Animated 3D Character from Autodesk Maya as Quake MD2

  - by Andy R

  I'm wondering if anyone knows of a way to export geometry/textures for a rigged, animated character as Quake MD2? I’m developing an app for mobile devices, and I’ve found that MD2 works great for lightweight OpenGL rendering. I have several animated characters, and I’d like to export them as MD2 from Maya. Here are some of the things I have tried: Exporting FBX to 3D Studio Max and using the QTip plugin -- this works, but only if I convert my animation to point cache (pc2) and bring that into Max. When I do that, QTip plugin doesn’t honor the point cache node, and all I get is the model, no animation Exporting OBJ Sequence to Blender, exporting MD2 from there -- I can’t seem to get blender to compile the imported obj sequence into a single animated object Exporting FBX to Milkshape -- the maya rig gets corrupted in the process Bake point cache for animation, export to FBX, import to Milkshape -- again, Milkshape doesn’t honor the geometry cache, so all that appears is a static character. I’m currently writing a plugin for Maya using the blender MD2 export script as reference (but re-writing using the C++ Maya API), but I’m just wondering if anyone has done this before or has any suggestions of how to do this. Also, if anyone has other suggestions for the best format to render a single animated character on an OpenGL ES (Android or iPhone) device, I would be very grateful! I’m open to trying anything to get these animations rendering on device Thanks!

  Read the article

• Metaprogramming - self explanatory code - tutorials, articles, books

  - by elena

  Hello everybody, I am looking into improving my programming skils (actually I try to do my best to suck less each year, as our Jeff Atwood put it), so I was thinking into reading stuff about metaprogramming and self explanatory code. I am looking for something like an idiot's guide to this (free books for download, online resources). Also I want more than your average wiki page and also something language agnostic or preferably with Java examples. Do you know of such resources that will allow to efficiently put all of it into practice (I know experience has a lot to say in all of this but i kind of want to build experience avoiding the flow bad decisions - experience - good decisions)? EDIT: Something of the likes of this example from the Pragmatic Programmer: ...implement a mini-language to control a simple drawing package... The language consists of single-letter commands. Some commands are followed by a single number. For example, the following input would draw a rectangle: P 2 # select pen 2 D # pen down W 2 # draw west 2cm N 1 # then north 1 E 2 # then east 2 S 1 # then back south U # pen up Thank you!

  Read the article

• Database Change Management - Setup for Initial Create Scripts, Subsequent Migration Scripts

  - by Martin Aatmaa

  I've got a database change management workflow in place. It's based on SQL scripts (so, it's not a managed code-based solution). The basic setup looks like this: Initial/ Generate Initial Schema.sql Generate Initial Required Data.sql Generate Initial Test Data.sql Migration 0001_MigrationScriptForChangeOne.sql 0002_MigrationScriptForChangeTwo.sql ... The process to spin up a database is to then run all the Initlal scripts, and then run the sequential Migration scripts. A tool takes case of the versioning requirements, etc. My question is, in this kind of setup, is it useful to also maintain this: Current/ Stored Procedures/ dbo.MyStoredProcedureCreateScript.sql ... Tables/ dbo.MyTableCreateScript.sql ... ... By "this" I mean a directory of scripts (separated by object type) that represents the create scripts for spinning up the current/latest version of the database. For some reason, I really like the idea, but I can't concretely justify it's need. Am I missing something? The advantages would be: For dev and source control, we would have the same object-per-file setup that we're used to For deployment, we can spin up a new DB instance to the latest version either by running the Initial+Migrate, or by running the scripts from Current/ For dev, we do not need a DB instance running in order to do development. We can do "offline" development on the Current/ folder. The disadvantages would be: For each change, we need to update the scripts in the Current/ folder, as well as create a Migration script (in the Migration/ folder) Thanks in advance for any input!

  Read the article

• Sync Vs. Async Sockets Performance in C#

  - by Michael Covelli

  Everything that I read about sockets in .NET says that the asynchronous pattern gives better performance (especially with the new SocketAsyncEventArgs which saves on the allocation). I think this makes sense if we're talking about a server with many client connections where its not possible to allocate one thread per connection. Then I can see the advantage of using the ThreadPool threads and getting async callbacks on them. But in my app, I'm the client and I just need to listen to one server sending market tick data over one tcp connection. Right now, I create a single thread, set the priority to Highest, and call Socket.Receive() with it. My thread blocks on this call and wakes up once new data arrives. If I were to switch this to an async pattern so that I get a callback when there's new data, I see two issues The threadpool threads will have default priority so it seems they will be strictly worse than my own thread which has Highest priority. I'll still have to send everything through a single thread at some point. Say that I get N callbacks at almost the same time on N different threadpool threads notifying me that there's new data. The N byte arrays that they deliver can't be processed on the threadpool threads because there's no guarantee that they represent N unique market data messages because TCP is stream based. I'll have to lock and put the bytes into an array anyway and signal some other thread that can process what's in the array. So I'm not sure what having N threadpool threads is buying me. Am I thinking about this wrong? Is there a reason to use the Async patter in my specific case of one client connected to one server?

  Read the article

• SD to PCI adapter

  - by vy32

  We are looking for a way to plug a SD card into a PCI slot without going through a USB interface. We want to read the SD card's serial number directly and issue SD commands. The Chumby and the Google Android G1 phones both have SD slots that you can read from Linux without going through USB, but the Chumby only has a single SD card that's used for booting, and the G1 doesn't have any other reasonable storage if you use the SD card for that purpose. I'd really like a desktop with a few SD slots that are directly accessible. Anybody know of anything??

  Read the article

• RotatingFileHandler throws an exception when delay parameter is set

  - by Eli Courtwright

  When I run the following code under Python 2.6 import logging from logging.handlers import RotatingFileHandler rfh = RotatingFileHandler("testing.log", delay=True) logging.getLogger().addHandler(rfh) logging.warning("Boo!") then the last line throws AttributeError: RotatingFileHandler instance has no attribute 'level'. So I add the line rfh.setLevel(logging.DEBUG) before the call to addHandler, and then the last line throws AttributeError: RotatingFileHandler instance has no attribute 'filters'. So if I manually set filters to be an empty list, then it complains about not having the attribute lock, etc. When I remove the delay=True to leave it as the default value of False as documented here, the problem completely goes away. Am I missing something? How do I properly use the delay parameter of the RotatingFileHandler class? EDIT: Upon further analysis (presented in my own answer below), this looks like a bug, but I can't find a bug report on this in the Python bug tracker, even trying different search terms, so I guess I'll report it. However, if someone can locate the actual bug report, then I can avoid submitting a duplicate reporting and wasting the time of the Python developers. I'll hold off on reporting the bug for a few hours, and if someone posts an answer that has the current bug report, then I'll accept that answer for this question.

  Read the article

• Silverlight 4, combobox databinding problem

  - by synergetic

  In my Silverlight 4 app, CustomerView (UserControl) shows Customer object as it's DataContext. Customer object has IndustryCode string property. I created combobox called cboIndustryCode and bind it to the IndustryCode property the following way: <ComboBox x:Name="cboIndustryCode" SelectedValue="{Binding IndustryCode, Mode=TwoWay}" ... /> In code-behind I populate cboIndustryCode with List of Industry object, which has Code and Name properties: cboIndustryCode.ItemsSource = industries; //which is of List<Industry> type Now, to show everything properly, in XAML I added the following: <ComboBox x:Name="cboIndustryCode" SelectedValue="{Binding IndustryCode, Mode=TwoWay}" DisplayMemberPath="Name" SelectedValuePath="Code" ... /> So, when I get a customer class from my data layer and set the DataContext to this customer instance, the cboIndustryCode properly displays industry name. But, then I edit customer (not necessarily IndustryCode) and save the object (which resets DataContext = new Customer()) and retrieve the customer again from database, and I see that cboIndustryCode no longer working. It just displays nothing, and if I select new value from the list, it does not update underlying customer object's IndustryCode property. The problem goes away, if I put the following code in the place where I set DataContext to a instance of customer, retrieved from database: Binding binding = new Binding("IndustryCode"); binding.Mode = BindingMode.TwoWay; cboIndustryCode.SetBinding(ComboBox.SelectedValueProperty, binding); So, in short, combobox's binding is reset somehow every time I save my data. Can someone tell me the reason?

  Read the article

• Force float left with no line break no matter what

  - by Tesserex

  I'm guessing this isn't possible, but here goes. I have two tables, and I'm trying to get them to sit side-by-side so that they look like one table. The reason for this, instead of using one larger table, is that the data in the second table needs to be handled on a column basis, not row basis, for performance reasons like caching and AJAX-fetching data. So rather than have to reload the whole table for a single column, I decided to break the column out into a separate table, but have it visually seem like a single table. I can't find a way to forcibly put the second table next to the first. I can float them, but when the first table is too wide, the second one breaks to the next line. Here's the kicker: the width of the first table is dynamic. So I can't just set a huge width to their container. Well, I could set a huge width, like 1000%, but then I have a huge ugly horizontal scroll bar. So is there any way to tell the second table "Stay on that same line, no matter what! And line up right next to the previous element please!"

  Read the article

• Directory name for non-generic Proprietary stuff

  - by George Bailey

  Is there a common or standard directory name for the company-specific stuff that exists in a server? This would include any crons, scripts, webserver docroots, programs, non-database storage areas, service codebases, etc. We could of course put crons in /etc/cron.d, put docroots in /home/webservd, scripts in one of the bin directories, but that would be messy. If XYZ Technology Corp wanted to have all the non-generic stuff in one place, would they make a directory /xyz or /home/xyz or is there an alternative directory name that is not company-specific, but intended for company-specific stuff? What is most common?

  Read the article

• Connecting to a Ghost User in Flex RTMFP

  - by Dan

  I have a simple Flex RTMFP P2P video app in the same mold as the Adobe Cirrus VideoPhone Sample application. A problem I've been encountering in developing this app (the same problem occurs in the sample) is when you try to connect to a ghost Stratus instance i.e you try to call someone whose Stratus id is in the database but who is no longer on the page. So here's an example of what I mean: Let's say you go to the Adobe Stratus sample and connect as Dan. Then open up a new tab, go to the sample again and connect as Fred. If from this point, you (as Fred) call Dan everything will work fine. But, if you close the tab in which you connected as Dan, and then from the Fred tab try to connect to Dan the program will just hang. I would have thought there would be a NetStream event that would be triggered if you tried to connect to a Stratus instance that is not longer online but I can't seem to find anything besides NetStream.Connect.Rejected which doesn't seem to be called. Any help is much appreciated!

  Read the article

• Updating Workflow Task without Correlation Token

  - by Khurram Aziz

  I have inhertied a sequential sharepoint workflow which deals with multiple tasks for different people; multi step approval based on certain condition..For each approval new task is created and monitored...For some reason; we have decided to use single task for the whole workflow and the single task will get assigned to required person at different stages...this will help us reduce the cluter in the task list For refactoring it; I am trying to create "CreateOrUpdateTaskAndWaitForCompletition" activity...so that I can use this component multiple times as per given workflow. Create/Wait branch of my activity works fine; as I have the correlation token within the activity. But when I try two instances of this activity; task is created in first activity and it needs to be updated in second instance where I dont have the correlation token. In the second instance; (Update/Wait branch) I have tried updating the task through code activity but its not working and I am getting "This task is currently locked by a running workflow and cannot be edited" exception! Can I use UpdateTask activity without correlation token? Can I programmatically update the workflow task? Can I programmatically unlock the workflow task?

  Read the article

• Can I have different ESX hosts accessing the same LUN over different protocols?

  - by Kevin Kuphal

  I currently have a cluster of two ESX 3.5U2 servers connected directly via FiberChannel to a NetApp 3020 cluster. These hosts mount four VMFS LUNs for virtual machine storage. Currently these LUNs are only made available via our FiberChannel initator in the Netapp configuration If I were to add an ESXi host to the cluster for internal IT use can I: Make the same VMFS LUNs available via the iSCSI initiator on the Netapp Connect this ESXi host to those LUNs via iSCSI Do all of this while the existing two ESX hosts are connected to those LUNs via FiberChannel Does anyone have experience with this type of mixed protocol environment, specifically with Netapp?

  Read the article

• NFS and KVM. Slow Speed

  - by Javier Martinez

  I have a KVM virtualization in Debian with 2 guests (Debian and Windows 2008). I want to have a 'mount point' shared that can be accessed by the 3 system (host and 2 guests) at the same time. So the only thing that I found was a NFS/SMB network storage. I picked NFS Due to my Ethernet network (10/100), the speed average that I get between accessing/transfering files between the 3 system is always 8~10MB/s. The point is if is there any chance of get a boost system for sharing files between 3 system (at the same time) without wasting the speed of my SATA disks. I mean, without the Ethernet limitation of 10 MB/s

  Read the article

• Singletons, thread safety and structuremap

  - by Ben

  Hi, Currently I have the following class: public class PluginManager { private static bool s_initialized; private static object s_lock = new object(); public static void Initialize() { if (!s_initialized) { lock (s_lock) { if (!s_initialized) { // initialize s_initialized = true; } } } } } The important thing here is that Initialize() should only be executed once whilst the application is running. I thought that I would refactor this into a singleton class since this would be more thread safe?: public sealed class PluginService { static PluginService() { } private static PluginService _instance = new PluginService(); public static PluginService Instance { get { return _instance; } } private bool s_initialized; public void Initialize() { if (!s_initialized) { // initialize s_initialized = true; } } } Question one, is it still necessary to have the lock here (I have removed it) since we will only ever be working on the same instance? Finally, I want to use DI and structure map to initialize my servcices so I have refactored as below: public interface IPluginService { void Initialize(); } public class NewPluginService : IPluginService { private bool s_initialized; public void Initialize() { if (!s_initialized) { // initialize s_initialized = true; } } } And in my registry: ForRequestedType<IPluginService>() .TheDefaultIsConcreteType<NewPluginService>().AsSingletons(); This works as expected (singleton returning true in the following code): var instance1 = ObjectFactory.GetInstance<IPluginService>(); var instance2 = ObjectFactory.GetInstance<IPluginService>(); bool singleton = (instance1 == instance2); So my next question, is the structure map solution as thread safe as the singleton class (second example). The only downside is that this would still allow NewPluginService to be instantiated directly (if not using structure map). Many thanks, Ben

  Read the article

• KVM slow guest i/o

  - by Akarot

  Host: Debian 6.0 (squeeze) with qemu-kvm and libvirt from squeeze-backports ii qemu-kvm 1.0+dfsg-8~bpo60+1 ii libvirt-bin 0.9.8-2~bpo60+2 Has 3TB sata drives with software raid and lvm. It has a sequential write speed of ~140MB/s measured with dd bs=1M count=512 if=/dev/zero of=test conv=fdatasync Elevator set to cfq Guest Debian 6.0 (squeeze) Uses LVM as storage. Drivers are virtio and cache='none' Sequential write speed is considerably slower with only 25-50MB/s Elevator set to noop I'm kind of running out of ideas for further tweaks but I'm sure that I/O speed should be much faster because many people are reporting almost native performance with lvm.

  Read the article

• Wireless Network Issue, Disconnecting Randomly From Network

  - by Surfer513

  I'm having an odd problem with my wireless network. Here is the background information: Server (Windows Server 2008) 1 to 10 end user machines connecting to the network Layer 3 Access Point (Asus WL-330 gE) connected to ethernet of Server and all machines connect to the network via the AP The end user machines get a connection to the server with no problems initially. But then connections are randomly lost throughout the day to the server/network. The wireless NICs of the machines still see the wireless network but are unable to connect to it. Then after some time the connection is regained automatically. I initially thought there was a problem with this particular AP, but then I took the same make/model AP out of storage and still ran into the problem. Any ideas what could be causing this??? Very confusing that the wireless nics on the end user machines can still see the network but not connect, and that the connections are randomly lost/gained. Thanks in advance!

  Read the article

• Setting an instanced class property overwrites the property in all instances.

  - by Peter Moore

  I have two instances of a class. Setting a property in the first instance will set it for the second instance too. Why? It shouldn't. The property is not a "prototype property". Check the code below, what happens to Peter Griffin? Two objects are created and given the name "Peter Griffin" and "Roger Moore" but the alert boxes will say "Peter Moore" and "Roger Moore". What happened to Peter Griffin? var BaseClass = function(){ this.name = ""; this.data = {}; this.data.lastname = ""; } var ExtendedClass = function(){ this.setName = function(fname, lname){ this.name = fname; this.data.lastname = lname; } this.hello = function(){ alert("Full name: " + this.name + " " + this.data.lastname); } } ExtendedClass.prototype = new BaseClass(); pilot = new ExtendedClass(); driver = new ExtendedClass(); pilot.setName("Peter", "Griffin"); driver.setName("Roger", "Moore"); pilot.hello(); // Full name: Peter Moore driver.hello(); // Full name: Roger Moore

  Read the article

• Best way to automatically synchronize files between Linux and Windows

  - by Gregory

  My first choice was rsync but it caused some issues and is too manual. My second choice, currently under evaluation is Unison. Are there any other good options for bi-directional auto-syncing? The synching tool cannot add it's own files to the directories to be synched. Which removes CVS/SVN as a choice. Plus they are too manual. The requirements are user-level program on both sides, no root account access available. Only scanning on linux. On windows it could be a virtual drive/path. Very fast and efficient like rsync. Some other requirements include: machines are not on the same network, files cannot fall into the wrong hands, nor can they be handled by 3rd parties, this pretty much excludes all online storage sites.

  Read the article

< Previous Page | 416 417 418 419 420 421 422 423 424 425 426 427  | Next Page >