Search Results

Search found 23098 results on 924 pages for 'multiple processes'.

Page 441/924 | < Previous Page | 437 438 439 440 441 442 443 444 445 446 447 448  | Next Page >

• Taking an image backup of an entire server?

  - by WarDoGG

  I am currently using a dedicated server for my hosting needs. However, the costs are too high and I would like to suspend everything until I work out my business strategy again. Is there a way I can take a complete backup of the filesystem and run it in VMWare ? I cannot just copy the entire filesystem because there are lots of tools installed and tight changes to the server configuration files I myself dont know about (by the developers), but I need a snapshot of the entire disk image along with processes installed and everything is as is because for development needs, I need to work on this copy in VMWare or VirtualBox etc. Is it possible for me to take a full image copy ? How do I do it ?

  Read the article

• What are the elegant ways to do MixIns in Python?

  - by Slava Vishnyakov

  I need to find an elegant way to do 2 kinds of MixIns. First: class A(object): def method1(self): do_something() Now, a MixInClass should make method1 do this: do_other() - A.method1() - do_smth_else() - i.e. basically "wrap" the older function. I'm pretty sure there must exist a good solution to this. Second: class B(object): def method1(self): do_something() do_more() In this case, I want MixInClass2 to be able to inject itself between do_something() and do_more(), i.e.: do_something() - MixIn.method1 - do_more(). I understand that probably this would require modifying class B - that's ok, just looking for simplest ways to achieve this. These are pretty trivial problems and I actually solved them, but my solution is tainted. Fisrt one by using self._old_method1 = self.method1(); self.method1() = self._new_method1(); and writing _new_method1() that calls to _old_method1(). Problem: multiple MixIns will all rename to _old_method1 and it is inelegant. Second MixIn one was solved by creating a dummy method call_mixin(self): pass and injecting it between calls and defining self.call_mixin(). Again inelegant and will break on multiple MixIns.. Any ideas?

  Read the article

• Am I understanding premature optimization correctly?

  - by Ed Mazur

  I've been struggling with an application I'm writing and I think I'm beginning to see that my problem is premature optimization. The perfectionist side of me wants to make everything optimal and perfect the first time through, but I'm finding this is complicating the design quite a bit. Instead of writing small, testable functions that do one simple thing well, I'm leaning towards cramming in as much functionality as possible in order to be more efficient. For example, I'm avoiding multiple trips to the database for the same piece of information at the cost of my code becoming more complex. One part of me wants to just not worry about redundant database calls. It would make it easier to write correct code and the amount of data being fetched is small anyway. The other part of me feels very dirty and unclean doing this. :-) I'm leaning towards just going to the database multiple times, which I think is the right move here. It's more important that I finish the project and I feel like I'm getting hung up because of optimizations like this. My question is: is this the right strategy to be using when avoiding premature optimization?

  Read the article

• select all values from a dimension for which there are facts in all other dimensions

  - by ideasculptor

  I've tried to simplify for the purposes of asking this question. Hopefully, this will be comprehensible. Basically, I have a fact table with a time dimension, another dimension, and a hierarchical dimension. For the purposes of the question, let's assume the hierarchical dimension is zip code and state. The other dimension is just descriptive. Let's call it 'customer' Let's assume there are 50 customers. I need to find the set of states for which there is at least one zip code in which EVERY customer has at least one fact row for each day in the time dimension. If a zip code has only 49 customers, I don't care about it. If even one of the 50 customers doesn't have a value for even 1 day in a zip code, I don't care about it. Finally, I also need to know which zip codes qualified the state for selection. Note, there is no requirement that every zip code have a full data set - only that at least one zip code does. I don't mind making multiple queries and doing some processing on the client side. This is a dataset that only needs to be generated once per day and can be cached. I don't even see a particularly clean way to do it with multiple queries short of simply brute-force iteration, and there are a heck of a lot of 'zip codes' in the data set (not actually zip codes, but the there are approximately 100,000 entries in the lower level of the hierarchy and several hundred in the top level, so zipcode-state is a reasonable analogy)

  Read the article

• Escape whitespace in paths using nautilus script

  - by Tommy Brunn

  I didn't think this would be as tricky as it turned out to be, but here I am. I'm trying to write a Nautilus script in Python to upload one or more images to Imgur just by selecting and right clicking them. It works well enough with both single images and multiple images - as long as they don't contain any whitespace. In fact, you can upload a single image containing whitespace, just not multiple ones. The problem is that NAUTILUS_SCRIPT_SELECTED_FILE_PATHS returns all the selected files and directories as a space separated string. So for example, it could look like this: print os.environment['NAUTILUS_SCRIPT_SELECTED_FILE_PATHS'] /home/nevon/Desktop/test image.png /home/nevon/Desktop/test.jpg What I need is a way to -either in bash or Python- escape the spaces in the path - but not the spaces that delimit different items. Either that, or a way to put quotation marks around each item. The ultimate solution would be if I could do that in bash and then send the items as separate arguments to my python script. Something like: python uploader.py /home/nevon/Desktop/test\ image.png /home/nevon/Desktop/test.jpg I've tried RTFM'ing, but there doesn't seem to be a lot of good solutions for this. At least not that I've found. Any ideas?

  Read the article

• Multidimensional array (parent and childs)

  - by Juan

  I have a category system in a MySQL database with parents and childs. The database only stores the id of it''s immediate parent (or 0 if on root). Since the system allows multiple subcategories there are cases of multiple childs. For example [98] Storage [1] External [3] Pendrives [4] Portable hhdds [2] Internal [5] Sata hhdd [6] IDE hhdd [...] [99] Clothing The database would be id parent_id name 1 98 External 2 98 Internal 3 1 Pendrives 4 1 Portable 5 2 Sata 6 2 IDE 98 0 Storage 99 0 Clothing I also have a products table with a category id and I need to get a list of all the products in the first level of categories. For example: Product Category A 3 B 4 C 5 D 6 E 74 Should return 98: A, B, C, D 99: X, Y, Z... I'm stuck and I can't think of the logic to retrieve it in that way. I started by getting the IDs of all the categories that aren't in the first level by: while ($row = mysql_fetch_assoc($result)) { if ($row['parent_id'] != 0) { $level1[$i]['name'] = utf8_encode($row['categories_name']); $level1[$i]['id'] = $row['categories_id']; } $i++; } but I'm having a burnout and can't think of a way that would nest them. I thought some kind of while but it's infinite :P Any ideas please?

  Read the article

• Safari Web Inspector is not updating when I update an element with ajax

  - by Ashley

  I have a checkout page http://www.oipolloi.com/oipolloi/shop/viewbasket.php with multiple ajax calls after certain items update (EG look up postage cost when country is changed, then update discount boxes etc). I've asked for help in the past about the best method of making sure ALL calls have returned before allowing the form to be submitted for payment processing: http://stackoverflow.com/questions/2290372/how-do-i-prevent-form-submission-until-multiple-ajax-calls-have-finished-jquery I was fairly happy that the logic in the finished solution was correct, but I have still been receiving reports that people using Safari are able to submit the form without the ajax calls returning properly. I have tried using the Safari Web Inspector to debug but it seems that when you Inspect Element, then update an element with an ajax call, the Inspector doesn't seem to update. I am updating hidden fields, so it's hard to be able to know whether the problem lies with the DOM not being updated properly, or the Inspector itself. I'm using Safari 4.0.5 on PC and you can reproduce the problem above by looking for a div id="countryFieldsBilling" with Web Inspector. It should contain three hidden fields that are initially empty. You can try to make it update (or not) by choosing a country from the select menu at the bottom of 'Shipping Address' box, and then clicking the 'click to use Shipping Address' link at the top of the 'Billing Address' just below. The behaviour I am seeing is that the country chosen in the shipping select gets copied correctly to the country in the billing select, but the hidden inputs in the Web Inspector do not get updated. When these hidden inputs do not get updated, this causes the problem that Mac Safari users report. If you can let me know either how to get Web Inspector to work properly, or something else I may have missed in the behaviour of Mac Safari that may cause these problems, that would be great. Thanks in advance

  Read the article

• How to stop nginx on Mac OS X

  - by Alex Kaushovik

  I've installed nginx server on my Mac from MacPorts: sudo port install nginx. Then I followed the recommendation from the port installation console and created the launchd startup item for nginx, then started the server. It works fine (after I renamed nginx.conf.example to nginx.conf and renamed mime.types.example to mime.types), but I couldn't stop it... I tried sudo nginx -s stop - this doesn't stop the server, I can still see "Welcome to nginx!" page in my browser on http://localhost/, also I still see master and worker processes of nginx with ps -e | grep nginx. What is the best way to start/stop nginx on Mac? BTW, I've added "daemon off;" into nginx.conf - as recommended by various resources. Thank you.

  Read the article

• Prevent member of administrator group loging in via Remote Desktop

  - by Chris J

  In order to support some build processes on our Server 2003 development servers, we require a common user account that has administrative privs. Unfortuantly, this also means that anyone that knows the password can also gain admin privs on a server. Assume that trying to keep the password secret is a failed exercise. Developers that need admin privs already have admin privs so should be able to log in as themselves. So the question is a simple one: is there anything I can configure to prevent people (ab)using the account to gain administrator on servers they shouldn't have administrator on? I'm aware that devs could disable anything that is put in place, but that's then down to process and auditing to track and manage. I don't mind where or how: it can be via the local security policy, group policy, a batch file executed in the user's profile, or something else.

  Read the article

• Django sub-applications & module structure

  - by Rob Golding

  I am developing a Django application, which is a large system that requires multiple sub-applications to keep things neat. Therefore, I have a top level directory that is a Django app (as it has an empty models.py file), and multiple subdirectories, which are also applications in themselves. The reason I have laid my application out in this way is because the sub-applications are separated, but they would never be used on their own, outside the parent application. It therefore makes no sense to distribute them separately. When installing my application, the settings file has to include something like this: INSTALLED_APPS = ( ... 'myapp', 'myapp.subapp1', 'myapp.subapp2', ... ) ...which is obviously suboptimal. This also has the slightly nasty result of requiring that all the sub-applications are referred to by their "inner" name (i.e. subapp1, subapp2 etc.). For example, if I want to reset the database tables for subapp1, I have to type: python manage.py reset subapp1 This is annoying, especially because I have a sub-app called core - which is likely to conflict with another application's name when my application is installed in a user's project. Am I doing this completely wrongly, or is there away to force these "inner" apps to be referred to by their full name?

  Read the article

• How to start with entity framework and service oriented architecture?

  - by citronas

  At work I need to create a new web application, that will connection to an MySql Database. (So far I only have expercience with Linq-To-Sql classes and MSSQL Servers.) My superior tells me to use the entity framework (he probably refers to Linq-To-Entity) and provide everything as a service based architecture. Unfortunatly nobody at work has experience with that framework and with a real nice server oriented architecture. (till now no customer wanted to pay for architecture that he can't see. This speficic project I'm leading will be long-term, meaning multiple years, so it would be best to design it the way, that multiple targetting plattforms like asp.net, c# wpf, ... could use it) For now, the main target plattform is ASP.net So I do have the following questions: 1) Where can I read best what's really behind service oriented architecture (but for now beginner tutorials work fine as well) and how to do it in best practise? 2) So far I can't seem a real difference between Linq-To-Sql classes and the information I've google so far on the 'entity framework'. So, whats the difference? Where do I find nice tutorials for it. 3) Is there any difference in the entity framework regarding the database server (MSSQL or MySQL). If not, does that mean that code snipperts I will stumble across will word database independent? 4) I do you Visual Studio 2010. Do I have to regard something specific?

  Read the article

• How to use named_scope to incrementally construct a complex query?

  - by wbharding

  I want to use Rails named_scope to incrementally build complex queries that are returned from external methods. I believe I can get the behavior I want with anonymous scopes like so: def filter_turkeys Farm.scoped(:conditions => {:animal => 'turkey'}) end def filter_tasty Farm.scoped(:conditions => {:tasty => true}) end turkeys = filter_turkeys is_tasty = filter_tasty tasty_turkeys = filter_turkeys.filter_tasty But say that I already have a named_scope that does what these anonymous scopes do, can I just use that, rather than having to declare anonymous scopes? Obviously one solution would be to pass my growing query to each of the filter methods, a la def filter_turkey(existing_query) existing_query.turkey # turkey is a named_scoped that filters for turkey end But that feels like an un-DRY way to solve the problem. Oh, and if you're wondering why I would want to return named_scope pieces from methods, rather than just build all the named scopes I need and concatenate them together, it's because my queries are too complex to be nicely handled with named_scopes themselves, and the queries get used in multiple places, so I don't want to manually build the same big hairy concatenation multiple times.

  Read the article

• URL development and mod_rewrite

  - by iRector

  My site is made-up of the main page, and multiple sub-directories, all under the same domain. My URLS are currently like .................| Ideal clean version: mysite.com mysite.com/?content=content1 ......................| mysite.com/content1/ mysite.com/?content=content2&page=4 ........| mysite.com/content2/4/ mysite.com/?content=content3 ......................| mysite.com/content3/ mysite.com/?content=content4 ......................| mysite.com/content4/ mysite.com/?content=article&id=34 ............| mysite.com/article/34/ Then the sub-directories are essentially the same: mysite.com/subdir, mysite.com/subdir2, mysite.com/subdir3, etc mysite.com/subdir/?content=content1 ...................| mysite.com/subdir/content1/ mysite.com/subdir/?content=content2&page=4 .....| mysite.com/subdir/content2/4/ mysite.com/subdir/?content=content3 ...................| mysite.com/subdir/content3/ mysite.com/subdir/?content=content4 ...................| mysite.com/subdir/content4/ mysite.com/subdir/?content=article&id=34 .........| mysite.com/subdir/article/34/ I've used mod_rewrite briefly, but I'm not sure how to approach these multiple variables. Also, how would I differentiate between the actually subfolders, and the content variable. As so to prevent 'subdir' or 'subdir2' from being plugged in as the content variable for the root site. I've played around with plenty of code snippets, but I've wiped my .htaccess slate clean, and approach you all in an attempt to help me repopulate it. Your input would thoroughly be appreciated. Note: The only time the page query string will be needed is when 'content' == 'content2' ?content=content2&page=4 **Same rule is shared by the article/id relationship, all other 'content' values are expected to be dynamic.

  Read the article

• How do I lower idle cpu usage in ubuntu linux? Gnome or KDE Variants

  - by Jasen

  My question comes from a kde desktop currently, but it also happens with the gnome instance. When just sitting there, with only the cpu monitor widget running. no open windows, no background processes other than the desktop, my cpu is at ~20%. I wanna know how to fix this, and possibly get better performance out of it. When running my windows side, the cpu will sit at zero, and i generally load new programs about 400ms faster. With windows 7 being as slow as it is, this is not acceptable. and the widget is only set to check every 500ms, so im almost completely sure its not the widget. My system is a Gateway nv 53 amd 2.0 ghz turion with 4 gb of installed ram, and 500 gb hd. both linux and windows are 64 bit. average ram use on either system is about 1.4 gb for just the os

  Read the article

• Dependency injection in constructors

  - by andre

  Hello everyone. I'm starting a new project and setting up the base to work on. A few questions have risen and I'll probably be asking quite a few in here, hopefully I'll find some answers. First step is to handle dependencies for objects. I've decided to go with the dependency injection design pattern, to which I'm somewhat new, to handle all of this for the application. When actually coding it I came across a problem. If a class has multiple dependencies and you want to pass on multiple dependencies via the constructor (so that they cannot be changed after you instantiate the object). How do you do it without passing an array of dependencies, using call_user_func_array(), eval() or Reflection? This is what i'm looking for: <?php class DI { public function getClass($classname) { if(!$this->pool[$classname]) { # Load dependencies $deps = $this->loadDependencies($classname); # Here is where the magic should happen $instance = new $classname($dep1, $dep2, $dep3); # Add to pool $this->pool[$classname] = $instance; return $instance; } else { return $this->pool[$classname]; } } } Again, I would like to avoid the most costly methods to call the class. Any other suggestions?

  Read the article

• Apache CPU usage stays at 100% even when there are no requests

  - by Leirith

  Hi, I've been running the Apache HTTP server benchmarking tool (ab) on my new Apache server to test performance. I noticed that with a command like the following: ab -n 100000 -c 1000 http://www.mysite.com/ The CPU is used 100% by the apache2 processes during the testing. When the test concludes, usually with the following error just before the last requests are made: apr_poll: The timeout specified has expired (70007) Total of 99960 requests completed the CPU usage remains at 100%, and it's all being consumed by apache. I am using the worker MPM with and running PHP with mod_fcgid. Any advice as to why this is or what can be done to stop it would be appreciated.

  Read the article

• Sharing a COM port over TCP

  - by guinness

  What would be a simple design pattern for sharing a COM port over TCP to multiple clients? For example, a local GPS device that could transmit co-ordinates to remote hosts in realtime. So I need a program that would open the serial port and accept multiple TCP connections like: class Program { public static void Main(string[] args) { SerialPort sp = new SerialPort("COM4", 19200, Parity.None, 8, StopBits.One); Socket srv = new Socket(AddressFamily.InterNetwork, SocketType.Stream, ProtocolType.Tcp); srv.Bind(new IPEndPoint(IPAddress.Any, 8000)); srv.Listen(20); while (true) { Socket soc = srv.Accept(); new Connection(soc); } } } I would then need a class to handle the communication between connected clients, allowing them all to see the data and keeping it synchronized so client commands are received in sequence: class Connection { static object lck = new object(); static List<Connection> cons = new List<Connection>(); public Socket socket; public StreamReader reader; public StreamWriter writer; public Connection(Socket soc) { this.socket = soc; this.reader = new StreamReader(new NetworkStream(soc, false)); this.writer = new StreamWriter(new NetworkStream(soc, true)); new Thread(ClientLoop).Start(); } void ClientLoop() { lock (lck) { connections.Add(this); } while (true) { lock (lck) { string line = reader.ReadLine(); if (String.IsNullOrEmpty(line)) break; foreach (Connection con in cons) con.writer.WriteLine(line); } } lock (lck) { cons.Remove(this); socket.Close(); } } } The problem I'm struggling to resolve is how to facilitate communication between the SerialPort instance and the threads. I'm not certain that the above code is the best way forward, so does anybody have another solution (the simpler the better)?

  Read the article

• Drupal 6 - I am using drupal_execute to insert a CCK node into my site. Everything is working excep

  - by rdurbin

  $form_state['values']['field_prx_mp3_labels'][0][value] = $mp3_labels; $form_state['values']['taxonomy'][0][value] = array('tags'=>array('1'=>'Music')); $errs = drupal_execute('prx_content_node_form', $form_state, (object) $nodeTmp); This is a Drupal 6 site. I am using drupal_execute to create a node programatically. The first line is working for field_prx_mp3_labels. The second (for taxonomy) is not. Here is what my select on the node add for my cck looks like: <select name="taxonomy[2][]" multiple="multiple" class="form-select" id="edit-taxonomy-2" size="9"><option value="">- None -</option><option value="5">Music</option><option value="6">-Rock/Pop</option><option value="7">-Jazz/Blues</option><option value="8">-Classical</option><option value="9">-Music Documentaries</option><option value="10">-Festivals/Concerts</option><option value="11">Arts</option><option value="19">-Literature</option><option value="12">Nature</option><option value="13">History</option><option value="15">-Music</option><option value="14">Culture</option><option value="17">-American Indian</option><option value="18">-Latino</option><option value="16">-Youth Perspective</option></select> I have tried many many variations for line 2 (relating to the taxonomy). This comment seemed close but it hasn't worked for me: http://drupal.org/node/178506#comment-1155576 Thanks!

  Read the article

• Entity Framework and WCf

  - by Nihilist

  Hi I am little confused on designing WCf services with EF. When using WCf and EF, where do we draw this line on what properties to return and what not to with the entity. Here is my scenario I have User. Here are the relations. User [1 to many] Address, User [ 1 to many] Email, User [ 1 to many] Phone So now on the webform, on page1 I can edit user information. say I can edit few properties on the user entity and can also edit address, phone, email entities[ like add / delete and update any] On page2, i can only update user properties and nothing related to navigation properties [ address, email, phone]. So when I return the User Entity [ OR DTO] should i be returning the navigation properties too? Or should the client make multiple calls to get navigation properites. Also, how does it go with Save? Like should the client make multiple calls to save user and related entites or just one call to save the graph? Lets say, if I just have a Save(User user) [ where user has all the related entities too] both page1 and page2 will call save and pass me the user. but one page1 i will need a lot more information. but on page2 i just need the user primitive properties. So my question is, where do we draw this line, how do we design theses services ? Is the WCF operation designed on the page and the fields it has ? I am hoping i explained my problem well enough.

  Read the article

• A way to auto cycle (close) through all screen sessions

  - by JBWhitmore

  I frequently use screen when I log into the interactive nodes to a supercomputer that I have access to -- and I often run things and move on. There are about 20 separate nodes that I can log into; and if I check any one of them I'll have something like 4 detached sessions. Each of those sessions will have maybe 5 screen sessions within that. Is there a quick way to cycle through all of these and close them down if they are not running any processes? My current process is to screen -ls and then screen -r #### then type exit until I'm back to the base screen.

  Read the article

• How do I count how many arrays have the same name within a multidimensional array with php?

  - by zeckdude

  I have a multidimensional array, and I would have multiple arrays within it. Some of those arrays contain multiple arrays within them as well, and I would like to count how many arrays are within the second array(the date). This is an example of the structure of the multidimensional array: $_SESSION['final_shipping'][04/03/2010][book] $_SESSION['final_shipping'][04/12/2010][magazine] $_SESSION['final_shipping'][04/12/2010][cd] This is the foreach statement I am currently using to count how many of the second array(the one with the dates) exists. foreach($_SESSION['final_shipping'] as $date_key => $date_value) { foreach ($date_value as $product_key => $product_value) { echo 'There are ' . count($date_key) . ' of the ' . $date_key . ' selection.<br/>'; } } It is currently outputting this: There are 1 of the 04/03/2010 selection. There are 1 of the 04/12/2010 selection. There are 1 of the 04/12/2010 selection. I would like it to output this: There are 1 of the 04/03/2010 selection. There are 2 of the 04/12/2010 selection.

  Read the article

• No grammar constraints (DTD or XML schema) detected for the document.

  - by fastcodejava

  I have this dtd : http://fast-code.sourceforge.net/template.dtd But when I include in an xml I get the warning : No grammar constraints (DTD or XML schema) detected for the document. The xml is : <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE templates PUBLIC "//UNKNOWN/" "http://fast-code.sourceforge.net/template.dtd"> <templates> <template type="type"> <description>Some</description> <variation></variation> <variation-field></variation-field> <allow-multiple-variation></allow-multiple-variation> <class-pattern></class-pattern> <getter-setter>setter</getter-setter> <allowed-file-extensions>java</allowed-file-extensions> <number-required-classes>1</number-required-classes> <template-body> <![CDATA[ Some Data ]]> </template-body> </template> </templates> Any clue?

  Read the article

• Why is process not being displayed by TOP

  - by drN

  I am running a Mathematica script (this question probably doesn't fit in Mathematica.SE however) and I know that it generally takes up a lot of RAM and loads up my cores. However, althought pgrep MathKernel is showing a pid, I find that top doesn't show this in the top processes, although I notice that it is taking up about 2.25GB of the 8GB available to me. pmap -x my_process_id total kB 2243132 1907404 1892108 AND ps aux | grep MathKernel dnaneet 20837 12.6 23.3 2234944 1907404 pts/1 Sl 09:23 8:01 /share/apps/mathematica/8.0.4/SystemFiles/Kernel/Binaries/Linux-x86-64/MathKernel -runfirst $TopDirectory="/share/apps/mathematica/8.0.4" -script ./dcm_10micrometer_2x -- ./dcm_10micrometer_2x ps aux shows that the process is taking about 12% (In asterisks) dnaneet 20601 0.0 0.0 68264 1660 pts/1 Ss 09:15 0:00 -bash **dnaneet 20837 12.2 23.3 2234944 1907404 pts/1 Sl 09:23 8:01 /share/apps/mat** dnaneet 21922 0.0 0.0 65604 948 pts/1 R+ 10:29 0:00 ps -aux Did this process fail and is the MathKernel just lingering?

  Read the article

• Logging off does not kill process in Windows Server 2003

  - by user25951

  I have a Windows Server 2003(Enterprise, SP2). My understanding was that any process created by a user will be terminated when the user loggs off the account. But its not happening. I login via Administrator account. Start a simple java process and logoff. But the process is not killed. Is there any configuration for this or something? I am mostly a software programmer and not much in to servers and so I am stuck. I found out that while logging off, 1) Win32 is supposed to send a CTRL_LOGOFF_EVENT to all processes started by that user. 2) JVM is supposed to handle this event and terminate the VM. But I can't understand why my java process is not killed when i logoff. Any idea!!!

  Read the article

• How do I handle the messages for a simple web-based live chat, on the server side?

  - by Carson Myers

  I'm building a simple live chat into a web application running on Django, but one thing I'm confused about is how I should store the messages between users. The chat will support multiple users, and a chat "session" is composed of users connected to one user that is the "host." The application is a sort of online document collaboration thing, so user X has a document, and users Y and Z would connect to user X to talk about the document, and that would be one chat session. If user Y disconnected for five minutes, and then signed back in and reconnected to user X, he should not get any of the messages shared between users X and Z while he was away. if users X, Y, and Z can have a chat session about user X's document, then users X and Y can connect to a simultaneous, but separate discussion about user Z's document. How should I handle this? Should I keep each message in the database? Each message would have an owner user and a target user (the host), and a separate table would be used to connect users with messages (which messages are visible to what users). Or should I store each session as an HTML file on the server, which messages get appended to? The problem is, I can't just send messages directly between clients. They have to be sent to the server in a POST request, and then each client has to periodically check for the messages in a GET request. Except I can't just have each message cleared after a client fetches it, because there could be multiple clients. How should I set this up? Any suggestions?

  Read the article

< Previous Page | 437 438 439 440 441 442 443 444 445 446 447 448  | Next Page >