Search Results

Search found 31501 results on 1261 pages for 'event log'.

Page 18/1261 | < Previous Page | 14 15 16 17 18 19 20 21 22 23 24 25  | Next Page >

  • Customizing Mail Message in SSIS Event Handler

    - by Eric Ness
    I want to add an email notification to an SSIS 2005 package event handler. I've added a Send Mail task to the event handler. I'd like to customize the email body to include things like the error description. I've tried including @[System::ErrorDescription] in the MessageSource field, but the mail message doesn't include the value of ErrorDescription only the name of the variable.

    Read the article

  • How to add a SaveOrUpdateCopy event listener in NHibernate

    - by skrishna
    How can I add a event listener for SaveOrUpdateCopy in NHibernate ? I see that the ListenerType enumeration does not have a 'SaveOrUpdateCopy' type. I tried using the 'Merge' type, but that adds it to the MergeEventListeners collection. The SaveOrUpdateCopy invokes the events from the SaveOrUpdateCopyEventListeners collection. How can I add my event class to the SaveOrUpdateCopyEventListeners collection in NHibernate? Any help is appreciated.

    Read the article

  • C# wpf: Need to add mouseclick event to certain text within a textbox

    - by Michael
    I have a textbox with a paragraph of information. There are certain words in the paragraph that i want the user to be able to click on, and when clicked, a different textbox is populated with more information. I know that you can have the event for the whole textbox, but that isn't want i want. I only want to call that event when certain words within the box are clicked.

    Read the article

  • where store event handler method in WPF - MVVM

    - by netmajor
    Hey, Where should I store event methods for button Click event ?Normally it's store in code behind of wpf page, <Button Name="myButton" Click="myButton_Click">Click Me</Button> but in MVVM it should be store in other view-model class and bind to click property of button like that?? <Button Name="myButton" Click="{Binding StaticResouces myButton_Click}">Click Me</Button>

    Read the article

  • Making a Statement: How to retrieve the T-SQL statement that caused an event

    - by extended_events
    If you’ve done any troubleshooting of T-SQL, you know that sooner or later, probably sooner, you’re going to want to take a look at the actual statements you’re dealing with. In extended events we offer an action (See the BOL topic that covers Extended Events Objects for a description of actions) named sql_text that seems like it is just the ticket. Well…not always – sounds like a good reason for a blog post. When is a statement not THE statement? The sql_text action returns the same information that is returned from DBCC INPUTBUFFER, which may or may not be what you want. For example, if you execute a stored procedure, the sql_text action will return something along the lines of “EXEC sp_notwhatiwanted” assuming that is the statement you sent from the client. Often times folks would like something more specific, like the actual statements that are being run from within the stored procedure or batch. Enter the stack Extended events offers another action, this one with the descriptive name of tsql_stack, that includes the sql_handle and offset information about the statements being run when an event occurs. With the sql_handle and offset values you can retrieve the specific statement you seek using the DMV dm_exec_sql_statement. The BOL topic for dm_exec_sql_statement provides an example for how to extract this information, so I’ll cover the gymnastics required to get the sql_handle and offset values out of the tsql_stack data collected by the action. I’m the first to admit that this isn’t pretty, but this is what we have in SQL Server 2008 and 2008 R2. We will be making it easier to get statement level information in the next major release of SQL Server. The sample code For this example I have a stored procedure that includes multiple statements and I have a need to differentiate between those two statements in my tracing. I’m going to track two events: module_end tracks the completion of the stored procedure execution and sp_statement_completed tracks the execution of each statement within a stored procedure. I’m adding the tsql_stack events (since that’s the topic of this post) and the sql_text action for comparison sake. (If you have questions about creating event sessions, check out Pedro’s post Introduction to Extended Events.) USE AdventureWorks2008GO -- Test SPCREATE PROCEDURE sp_multiple_statementsASSELECT 'This is the first statement'SELECT 'this is the second statement'GO -- Create a session to look at the spCREATE EVENT SESSION track_sprocs ON SERVERADD EVENT sqlserver.module_end (ACTION (sqlserver.tsql_stack, sqlserver.sql_text)),ADD EVENT sqlserver.sp_statement_completed (ACTION (sqlserver.tsql_stack, sqlserver.sql_text))ADD TARGET package0.ring_bufferWITH (MAX_DISPATCH_LATENCY = 1 SECONDS)GO -- Start the sessionALTER EVENT SESSION track_sprocs ON SERVERSTATE = STARTGO -- Run the test procedureEXEC sp_multiple_statementsGO -- Stop collection of events but maintain ring bufferALTER EVENT SESSION track_sprocs ON SERVERDROP EVENT sqlserver.module_end,DROP EVENT sqlserver.sp_statement_completedGO Aside: Altering the session to drop the events is a neat little trick that allows me to stop collection of events while keeping in-memory targets such as the ring buffer available for use. If you stop the session the in-memory target data is lost. Now that we’ve collected some events related to running the stored procedure, we need to do some processing of the data. I’m going to do this in multiple steps using temporary tables so you can see what’s going on; kind of like having to “show your work” on a math test. The first step is to just cast the target data into XML so I can work with it. After that you can pull out the interesting columns, for our purposes I’m going to limit the output to just the event name, object name, stack and sql text. You can see that I’ve don a second CAST, this time of the tsql_stack column, so that I can further process this data. -- Store the XML data to a temp tableSELECT CAST( t.target_data AS XML) xml_dataINTO #xml_event_dataFROM sys.dm_xe_sessions s INNER JOIN sys.dm_xe_session_targets t    ON s.address = t.event_session_addressWHERE s.name = 'track_sprocs' SELECT * FROM #xml_event_data -- Parse the column data out of the XML blockSELECT    event_xml.value('(./@name)', 'varchar(100)') as [event_name],    event_xml.value('(./data[@name="object_name"]/value)[1]', 'varchar(255)') as [object_name],    CAST(event_xml.value('(./action[@name="tsql_stack"]/value)[1]','varchar(MAX)') as XML) as [stack_xml],    event_xml.value('(./action[@name="sql_text"]/value)[1]', 'varchar(max)') as [sql_text]INTO #event_dataFROM #xml_event_data    CROSS APPLY xml_data.nodes('//event') n (event_xml) SELECT * FROM #event_data event_name object_name stack_xml sql_text sp_statement_completed NULL <frame level="1" handle="0x03000500D0057C1403B79600669D00000100000000000000" line="4" offsetStart="94" offsetEnd="172" /><frame level="2" handle="0x01000500CF3F0331B05EC084000000000000000000000000" line="1" offsetStart="0" offsetEnd="-1" /> EXEC sp_multiple_statements sp_statement_completed NULL <frame level="1" handle="0x03000500D0057C1403B79600669D00000100000000000000" line="6" offsetStart="174" offsetEnd="-1" /><frame level="2" handle="0x01000500CF3F0331B05EC084000000000000000000000000" line="1" offsetStart="0" offsetEnd="-1" /> EXEC sp_multiple_statements module_end sp_multiple_statements <frame level="1" handle="0x03000500D0057C1403B79600669D00000100000000000000" line="0" offsetStart="0" offsetEnd="0" /><frame level="2" handle="0x01000500CF3F0331B05EC084000000000000000000000000" line="1" offsetStart="0" offsetEnd="-1" /> EXEC sp_multiple_statements After parsing the columns it’s easier to see what is recorded. You can see that I got back two sp_statement_completed events, which makes sense given the test procedure I’m running, and I got back a single module_end for the entire statement. As described, the sql_text isn’t telling me what I really want to know for the first two events so a little extra effort is required. -- Parse the tsql stack information into columnsSELECT    event_name,    object_name,    frame_xml.value('(./@level)', 'int') as [frame_level],    frame_xml.value('(./@handle)', 'varchar(MAX)') as [sql_handle],    frame_xml.value('(./@offsetStart)', 'int') as [offset_start],    frame_xml.value('(./@offsetEnd)', 'int') as [offset_end]INTO #stack_data    FROM #event_data        CROSS APPLY    stack_xml.nodes('//frame') n (frame_xml)    SELECT * from #stack_data event_name object_name frame_level sql_handle offset_start offset_end sp_statement_completed NULL 1 0x03000500D0057C1403B79600669D00000100000000000000 94 172 sp_statement_completed NULL 2 0x01000500CF3F0331B05EC084000000000000000000000000 0 -1 sp_statement_completed NULL 1 0x03000500D0057C1403B79600669D00000100000000000000 174 -1 sp_statement_completed NULL 2 0x01000500CF3F0331B05EC084000000000000000000000000 0 -1 module_end sp_multiple_statements 1 0x03000500D0057C1403B79600669D00000100000000000000 0 0 module_end sp_multiple_statements 2 0x01000500CF3F0331B05EC084000000000000000000000000 0 -1 Parsing out the stack information doubles the fun and I get two rows for each event. If you examine the stack from the previous table, you can see that each stack has two frames and my query is parsing each event into frames, so this is expected. There is nothing magic about the two frames, that’s just how many I get for this example, it could be fewer or more depending on your statements. The key point here is that I now have a sql_handle and the offset values for those handles, so I can use dm_exec_sql_statement to get the actual statement. Just a reminder, this DMV can only return what is in the cache – if you have old data it’s possible your statements have been ejected from the cache. “Old” is a relative term when talking about caches and can be impacted by server load and how often your statement is actually used. As with most things in life, your mileage may vary. SELECT    qs.*,     SUBSTRING(st.text, (qs.offset_start/2)+1,         ((CASE qs.offset_end          WHEN -1 THEN DATALENGTH(st.text)         ELSE qs.offset_end         END - qs.offset_start)/2) + 1) AS statement_textFROM #stack_data AS qsCROSS APPLY sys.dm_exec_sql_text(CONVERT(varbinary(max),sql_handle,1)) AS st event_name object_name frame_level sql_handle offset_start offset_end statement_text sp_statement_completed NULL 1 0x03000500D0057C1403B79600669D00000100000000000000 94 172 SELECT 'This is the first statement' sp_statement_completed NULL 1 0x03000500D0057C1403B79600669D00000100000000000000 174 -1 SELECT 'this is the second statement' module_end sp_multiple_statements 1 0x03000500D0057C1403B79600669D00000100000000000000 0 0 C Now that looks more like what we were after, the statement_text field is showing the actual statement being run when the sp_statement_completed event occurs. You’ll notice that it’s back down to one row per event, what happened to frame 2? The short answer is, “I don’t know.” In SQL Server 2008 nothing is returned from dm_exec_sql_statement for the second frame and I believe this to be a bug; this behavior has changed in the next major release and I see the actual statement run from the client in frame 2. (In other words I see the same statement that is returned by the sql_text action  or DBCC INPUTBUFFER) There is also something odd going on with frame 1 returned from the module_end event; you can see that the offset values are both 0 and only the first letter of the statement is returned. It seems like the offset_end should actually be –1 in this case and I’m not sure why it’s not returning this correctly. This behavior is being investigated and will hopefully be corrected in the next major version. You can workaround this final oddity by ignoring the offsets and just returning the entire cached statement. SELECT    event_name,    sql_handle,    ts.textFROM #stack_data    CROSS APPLY sys.dm_exec_sql_text(CONVERT(varbinary(max),sql_handle,1)) as ts event_name sql_handle text sp_statement_completed 0x0300070025999F11776BAF006F9D00000100000000000000 CREATE PROCEDURE sp_multiple_statements AS SELECT 'This is the first statement' SELECT 'this is the second statement' sp_statement_completed 0x0300070025999F11776BAF006F9D00000100000000000000 CREATE PROCEDURE sp_multiple_statements AS SELECT 'This is the first statement' SELECT 'this is the second statement' module_end 0x0300070025999F11776BAF006F9D00000100000000000000 CREATE PROCEDURE sp_multiple_statements AS SELECT 'This is the first statement' SELECT 'this is the second statement' Obviously this gives more than you want for the sp_statement_completed events, but it’s the right information for module_end. I leave it to you to determine when this information is needed and use the workaround when appropriate. Aside: You might think it’s odd that I’m showing apparent bugs with my samples, but you’re going to see this behavior if you use this method, so you need to know about it.I’m all about transparency. Happy Eventing- Mike Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Tales of a corrupt SQL log

    - by guybarrette
    Warning: I’m a simple dev, not an all powerful DBA with godly powers. This morning, one of my sites was down and DNN reported a problem with the database.  A quick series of tests revealed that the culprit was a corrupted log file. Easy fix I said, I have daily backups so it’s just a mater of restoring a good copy of the database and log files.  Well, I found out that’s not exactly true.  You see, for this database, I have daily file backups and these are not database backups created by SQL Server. So I restored a set of files from a couple of days ago, stopped the SQL service, copied the files over the bad ones, restarted the service only to find out that SQL doesn’t like when you do that.  It suspects something fishy and marks the database as suspect.  A database marked as suspect can’t be accessed at all.  So now what? I searched throughout the tubes of the InterWeb and found that you can restore from a corrupted log file by creating a new database with the same name as the defective one, then copy the restored database file (the one with data) over the newly created one.  Sweet!  But you still end up with SQL marking the database as suspect but at least, the newly created log is OK.  Well not true, it’s not corrupted but the lack of data makes it not OK for SQL so you need to rebuild the log.  How can you do that when SQL blocks any action the database?  First, you need to change the database status from suspect to emergency.  Then you need to set the database for single access only.  After that, you need to repair the log with DBCC and do the DBA dance.  If you dance long enough, SQL should repair the log file.  Now you need to set the access back to multi user.  Here’s the T-SQL script: use master GO EXEC sp_resetstatus 'MyDatabase' ALTER DATABASE MyDatabase SET EMERGENCY Alter database MyDatabase set Single_User DBCC checkdb('MyDatabase') ALTER DATABASE MyDatabase SET SINGLE_USER WITH ROLLBACK IMMEDIATE DBCC CheckDB ('MyDatabase', REPAIR_ALLOW_DATA_LOSS) ALTER DATABASE MyDatabase SET MULTI_USER So I guess that I would have been a lot easier to restore a SQL backup.  I can’t really say but the InterWeb seems to say so.  Anyway, lessons learned: Vive la différence: File backups are different then SQL backups. Don’t touch me: SQL doesn’t like when you restore a file over a corrupted one. The more the merrier: You should do both SQL and file backups. WTF?: The InterWeb provides you with dozens of way to deal with the problem but many are SQL 2000 or SQL 2005 only, many are confusing and many are written in strange dialects only DBAs understand. var addthis_pub="guybarrette";

    Read the article

  • C# Process Exited event not firing from within webservice

    - by davidpizon
    I am attempting to wrap a 3rd party command line application within a web service. If I run the following code from within a console application: Process process= new System.Diagnostics.Process(); process.StartInfo.FileName = "some_executable.exe"; // Do not spawn a window for this process process.StartInfo.CreateNoWindow = true; process.StartInfo.ErrorDialog = false; // Redirect input, output, and error streams process.StartInfo.UseShellExecute = false; process.StartInfo.RedirectStandardOutput = true; process.StartInfo.RedirectStandardError = true; process.StartInfo.RedirectStandardInput = true; process.EnableRaisingEvents = true; process.ErrorDataReceived += (sendingProcess, eventArgs) => { // Make note of the error message if (!String.IsNullOrEmpty(eventArgs.Data)) if (this.WarningMessageEvent != null) this.WarningMessageEvent(this, new MessageEventArgs(eventArgs.Data)); }; process.OutputDataReceived += (sendingProcess, eventArgs) => { // Make note of the message if (!String.IsNullOrEmpty(eventArgs.Data)) if (this.DebugMessageEvent != null) this.DebugMessageEvent(this, new MessageEventArgs(eventArgs.Data)); }; process.Exited += (object sender, EventArgs e) => { // Make note of the exit event if (this.DebugMessageEvent != null) this.DebugMessageEvent(this, new MessageEventArgs("The command exited")); }; process.Start(); process.StandardInput.Close(); process.BeginOutputReadLine(); process.BeginErrorReadLine(); process.WaitForExit(); int exitCode = process.ExitCode; process.Close(); process.Dispose(); if (this.DebugMessageEvent != null) this.DebugMessageEvent(this, new MessageEventArgs("The command exited with code: " + exitCode)); All events, including the "process.Exited" event fires as expected. However, when this code is invoked from within a web service method, all events EXCEPT the "process.Exited" event fire. The execution appears to hang at the line: process.WaitForExit(); Would anyone be able to shed some light as to what I might be missing?

    Read the article

  • jQuery click event on IE7-8, does not execute on the div, only on its text

    - by user3665301
    I have a problem using the jQuery click event with IE7-8-9. I apply the event on a div. But on these two IE versions, I have to click on the text contained within the div to make the event work. I don't understand because it was still normally working on these versions until I made a few changes (like adding the font css properties) but when I try to delete these changes it stil does not work as I want; Here is a jsfiddle illustrating the situation and its full screen result. http://jsfiddle.net/rC632/ function clickEvent(){ $('.answerDiv').click(function(){ $( "div:animated" ).stop(); if ( idPreviousClick === $(this)[0].id) { } else { if (idPreviousClick != -1) { $("#"+idPreviousClick).css({height:'100px', width:'100px', top:'0', 'line-height': '100px'}); $("#"+idPreviousClick).parent().css({height:'100px', width:'100px', top:'0'}); } $(this).animate({height:'120px', width:'120px', 'line-height': '120px'}); $(this).parent().animate({height:'120px', width:'120px', top:'-10px'}); idPreviousClick = $(this)[0].id; } }); } $(document).ready(function(){ clickEvent(); }); var idPreviousClick = -1; http://jsfiddle.net/rC632/embedded/result/ Could you have any idea of what is missing ? Thanks

    Read the article

  • Binding event handlers to specific elements, using bubbling (JavaScript/jQuery)

    - by Bungle
    I'm working on a project that approximates the functionality of Firebug's inspector tool. That is, when mousing over elements on the page, I'd like to highlight them (by changing their background color), and when they're clicked, I'd like to execute a function that builds a CSS selector that can be used to identify them. However, I've been running into problems related to event bubbling, and have thoroughly confused myself. Rather than walk you down that path, it might make sense just to explain what I'm trying to do and ask for some help getting started. Here are some specs: I'm only interested in elements that contain a text node (or any descendant elements with text nodes). When the mouse enters such an element, change its background color. When the mouse leaves that element, change its background color back to what it was originally. When an element is clicked, execute a function that builds a CSS selector for that element. I don't want a mouseover on an element's margin area to count as a mouseover for that element, but for the element beneath (I think that's default browser behavior anyway?). I can handle the code that highlights/unhighlights, and builds the CSS selector. What I'm primarily having trouble with is efficiently binding event handlers to the elements that I want to be highlightable/clickable, and avoiding/stopping bubbling so that mousing over a (<p>) element doesn't also execute the handler function on the <body>, for example. I think the right way to do this is to bind event handlers to the document element, then somehow use bubbling to only execute the bound function on the topmost element, but I don't have any idea what that code looks like, and that's really where I could use help. I'm using jQuery, and would like to rely on that as much as possible. Thanks in advance for any guidance!

    Read the article

  • In C#, are event handler arguments covariant?

    - by Roger Lipscombe
    Maybe covariant's not the word, but if I have a class that raises an event, with (e.g.) FrobbingEventArgs, am I allowed to handle it with a method that takes EventArgs? Here's some code: class Program { static void Main(string[] args) { Frobber frobber = new Frobber(); frobber.Frobbing += FrobberOnFrobbing; frobber.Frob(); } private static void FrobberOnFrobbing(object sender, EventArgs e) { // Do something interesting. Note that the parameter is 'EventArgs'. } } internal class Frobber { public event EventHandler<FrobbingEventArgs> Frobbing; public event EventHandler<FrobbedEventArgs> Frobbed; public void Frob() { OnFrobbing(); // Frob. OnFrobbed(); } private void OnFrobbing() { var handler = Frobbing; if (handler != null) handler(this, new FrobbingEventArgs()); } private void OnFrobbed() { var handler = Frobbed; if (handler != null) handler(this, new FrobbedEventArgs()); } } internal class FrobbedEventArgs : EventArgs { } internal class FrobbingEventArgs : EventArgs { } The reason I ask is that ReSharper seems to have a problem with (what looks like) the equivalent in XAML, and I'm wondering if it's a bug in ReSharper, or a mistake in my understanding of C#.

    Read the article

  • C++ wxWidgets Event Handling

    - by Wallter
    Q1:I am using wxWidgets (C++) and have come accost a problem that i can not locate any help. I have created several wxTextCtrl boxes and would like the program to update the simple calculations in them when the user 'kills the focus.' I could not find any documentation on this subject on the wxWidgets webpage and Googling it only brought up wxPython. The two events i have found are: EVT_COMMAND_KILL_FOCUS - EVT_KILL_FOCUS for neither of which I could find any snippet for. Could anyone give me a short example or lead me to a page that would be helpful? Q2:Would i have to create an event to handle the focus being killed for each of my 8 wxTextCtrl boxes? In the case that i have to create a different event: How would i get each event to differentiate from each other? I know i will have to create new wxID's for each of the wxTextCtrl boxes but how do I get the correct one to be triggered? class BasicPanel : public wxPanel { ... wxTextCtrl* one; wxTextCtrl* two; wxTextCtrl* three; wxTextCtrl* four; ... }

    Read the article

  • In C#, are event handler arguments contravariant?

    - by Roger Lipscombe
    If I have a class that raises an event, with (e.g.) FrobbingEventArgs, am I allowed to handle it with a method that takes EventArgs? Here's some code: class Program { static void Main(string[] args) { Frobber frobber = new Frobber(); frobber.Frobbing += FrobberOnFrobbing; frobber.Frob(); } private static void FrobberOnFrobbing(object sender, EventArgs e) { // Do something interesting. Note that the parameter is 'EventArgs'. } } internal class Frobber { public event EventHandler<FrobbingEventArgs> Frobbing; public event EventHandler<FrobbedEventArgs> Frobbed; public void Frob() { OnFrobbing(); // Frob. OnFrobbed(); } private void OnFrobbing() { var handler = Frobbing; if (handler != null) handler(this, new FrobbingEventArgs()); } private void OnFrobbed() { var handler = Frobbed; if (handler != null) handler(this, new FrobbedEventArgs()); } } internal class FrobbedEventArgs : EventArgs { } internal class FrobbingEventArgs : EventArgs { } The reason I ask is that ReSharper seems to have a problem with (what looks like) the equivalent in XAML, and I'm wondering if it's a bug in ReSharper, or a mistake in my understanding of C#.

    Read the article

  • Assign delegate event handler from dynamically added child control

    - by mickyjtwin
    I have a control that handles commenting. In this control, I have set a delegate event handler for sending an email. I then have various types of controls, e.g. blog, articles etc, each of which may or may not have the commenting control added (which is done dynamically with me not knowing the id's), i.e. the commenting control is added outside this control. Each of these controls handles it's emailing differently(hence the event). What I'm trying to determine, is how to assign the event in the parent control. At the moment, I'm having to recursively search through all the controls on the page until I find the comment control, and set it that way. Example below explains: COMMENTING CONTROL public delegate void Commenting_OnSendEmail(); public partial class Commenting : UserControl { public Commenting_OnSendEmail OnComment_SendEmail(); private void Page_Load(object sender, EventArgs e) { if(OnComment_SendEmail != null) { OnComment_SendEmail(); } } } PARENT CONTROL public partial class Blog : UserControl { private void Page_Load(object sender, EventArgs e) { Commenting comControl = (Commenting)this.FindControl<Commenting>(this); if(comControl != null) { comCtrol.OnComment_SendEmail += new Commenting_OnSendMail(Blog_Comment_OnSendEmail); } } } Is there an easier way?

    Read the article

  • "pushModalScreen called by a non-event thread" thrown on event thread

    - by JGWeissman
    I am trying to get my Blackberry application to display a custom modal dialog, and have the opening thread wait until the user closes the dialog screen. final Screen dialog = new FullScreen(); ...// Fields are added to dialog Application.getApplication().invokeAndWait(new Runnable() { public void run() { Application.getUiApplication().pushModalScreen(dialog); } }); This is throwing an Exception which says "pushModalScreen called by a non-event thread" despite the fact that I am using invokeAndWait to call pushModalScreen from the event thread. Any ideas about what the real problem is? Here is the code to duplicate this problem: package com.test; import net.rim.device.api.ui.*; import net.rim.device.api.ui.component.*; import net.rim.device.api.ui.container.*; public class Application extends UiApplication { public static void main(String[] args) { new Application(); } private Application() { new Thread() { public void run() { Application.this.enterEventDispatcher(); } }.start(); final Screen dialog = new FullScreen(); final ButtonField closeButton = new ButtonField("Close Dialog"); closeButton.setChangeListener(new FieldChangeListener() { public void fieldChanged(Field field, int context) { Application.getUiApplication().popScreen(dialog); } }); dialog.add(closeButton); Application.getApplication().invokeAndWait(new Runnable() { public void run() { try { Application.getUiApplication().pushModalScreen(dialog); } catch (Exception e) { // To see the Exception in the debugger throw new RuntimeException(e.getMessage()); } } }); System.exit(0); } } I am using Component Package version 4.5.0.

    Read the article

  • Adding jQuery event handler to expand div when a link is clicked

    - by hollyb
    I'm using a bit of jQuery to expand/hide some divs. I need to add an event handler so that when a link is clicked, it opens a corrisponding div. Each toggled div will have a unique class assigned to it. I am looking for some advice about how to build the event handler. The jQuery $(document).ready(function(){ $(".toggle_container:not(:first)").hide(); $(".toggle_container:first").show(); $("h6.trigger:first").addClass("active"); $("h6.trigger").click(function(){ $(this).toggleClass("active"); $(this).next(".toggle_container").slideToggle(300); }); The css: // uses a background image with an on (+) and off (-) state stacked on top of each other h6.trigger { background: url(buttonBG.gif) no-repeat;height: 46px;line-height: 46px;width: 300px;font-size: 2em;font-weight: normal;} h6.trigger a {color: #fff;text-decoration: none; display: block;} h6.active {background-position: left bottom;} .toggle_container { overflow: hidden; } .toggle_container .block {padding: 20px;} The html has a list of links, such as: <a href="#">One</a> <a href="#">Two</a> and the coorisponding divs to open: <h6 class="trigger">Container one</h6> <div class="toggle_container"> div one </div> <h6 class="trigger">Container two</h6> <div class="toggle_container Open"> div one </div> As I mentioned, I will be assigning a unique class to facilitate this. Any advice? Thanks! To clarify, i'm looking for some advice to build an event handler to toggle open a specific div when a link is clicked from a different part of the page, from a nav for instance.

    Read the article

  • Attaching event on the fly with prototype

    - by andreas
    Hi, I've been scratching my head over this for an hour... I have a list of todo which has done button on each item. Every time a user clicks on that, it's supposed to do an Ajax call to the server, calling the marking function then updates the whole list. The problem I'm facing is that the new list has done buttons as well on each item. And it doesn't attach the click event anymore after first update. How do you attach an event handler on newly updated element with prototype? Note: I've passed evalScripts: true to the updater wrapper initial event handler attach <script type="text/javascript"> document.observe('dom:loaded', function() { $$('.todo-done a').each(function(a) { $(a).observe('click', function(e) { var todoId = $(this).readAttribute('href').substr(1); new Ajax.Updater('todo-list', $.utils.getPath('/todos/ajax_mark/done'), { onComplete: function() { $.utils.updateList({evalScripts: true}); } }); }); }) }); </script> updateList: <script type="text/javascript"> var Utils = Class.create({ updateList: function(options) { if(typeof options == 'undefined') { options = {}; } new Ajax.Updater('todo-list', $.utils.getPath('/todos/ajax_get'), $H(options).merge({ onComplete: function() { $first = $('todo-list').down(); $first.hide(); Effect.Appear($first, {duration: 0.7}); new Effect.Pulsate($('todo-list').down(), {pulses: 1, duration: 0.4, queue: 'end'}); } }) ); } }) $.utils = new Utils('globals'); </script> any help is very much appreciated, thank you :)

    Read the article

  • Event Dispatching, void pointer and alternatives

    - by PeeS
    i have my event dispatching / handling functionality working fine, but there is one issue that i need to resolve. Long story short, here are the details. // The event structure struct tEventMessage { // Type of the event int Type; // (void*) Allows those to be casted into per-Type objects void *pArgument1; void *pArgument2; }; I am sending events from different modules in my engine by using the above structure, which requires a pointer to an argument. All messages are queued, and then dispatched on the next ::Tick(). It works fine, untill i try to send something that doesn't exist in next ::Tick, for example: When a mouse click is being handled, it calculates the click coordinates in world space. This is being sent with a pointer to a vector representing that position, but after my program quits that method, this pointer gets invalid obviously, cause local CVector3 is destructed: CVector2 vScreenSpacePosition = vAt; CVector3 v3DPositionA = CVector3(0,0,0); CVector3 v3DPositionB = CVector3(0,0,0); // Screen space to World space calculation for depth zNear v3DPositionA = CMath::UnProject(vScreenSpacePosition, m_vScreenSize, m_Level.GetCurrentCamera()->getViewMatrix(), m_Level.GetCurrentCamera()->getProjectionMatrix(), -1.0 ); // Screen space to World space calculation for depth zFar v3DPositionB = CMath::UnProject(vScreenSpacePosition, m_vScreenSize, m_Level.GetCurrentCamera()->getViewMatrix(), m_Level.GetCurrentCamera()->getProjectionMatrix(), 1.0); // Send zFar position and ScreenSpace position to the handlers // Obviously both vectors won't be valid after this method quits.. CEventDispatcher::Get()->SendEvent(CIEventHandler::EVENT_SYSTEM_FINGER_DOWN, static_cast<void*>(&v3DPositionB), static_cast<void*>(&vScreenSpacePosition)); What i want to ask is, if there is any chance i could make my tEventMessage more 'template', so i can handle sending objects like in the above situation + use what is already implemented? Can't figure it out at the moment.. What else can be done here to allow me to pass some locally available data ? Please can somebody shed a bit of light on this please?

    Read the article

  • jquery how to determine event triggered with .trigger() or with .event()

    - by Tony_M
    I have a selectbox. <select onchange="javascript:changePropertyDropdown('3','0','0','1',this.value,'80','50'); hideCart()" size="1" class="inputbox" id="property_id_prd_3_0_1" name="property_id_prd_3_0_1[]"> <option selected="selected" value="0">Select an option</option> <option value="1">38</option> <option value="2">40</option> <option value="3">42</option> <option value="4">43</option> </select> and some button which triggered change event for selectbox : $('div.attribute_wrapper select').bind('cascadeSelect',function(e, pAttr){ $(this).val(pAttr); }); Call it like this (prodAttr come with ajax): $('div.productImgGallery img').click(function(){ $('div.attribute_wrapper select').trigger('change'); }; $('div.attribute_wrapper select').change(function(){ $(this).trigger('cascadeSelect',prodAttr); }); When i call it like this, hideCart() function fires too. I need to call function changePropertyDropdown() only whith .trigger(), and changePropertyDropdown() + hideCart() on change event. How can i do this ?

    Read the article

  • INotifyPropertyChanged event listener within respective class not firing on client side (silverlight)

    - by Rob
    I'm trying to work out if there's a simple way to internalize the handling of a property change event on a Custom Entity as I need to perform some bubbling of the changes to a child collection within the class when the Background Property is changed via a XAML binding: public class MyClass : INotifyPropertyChanged { [Key] public int MyClassId { get; set; } [DataMember] public ObservableCollection<ChildMyClass> MyChildren { get; set; } public string _backgroundColor; [DataMember] public string BackgroundColor { get { return this._backgroundColor; } set { this._backgroundColor = value; if (this.PropertyChanged != null) { this.PropertyChanged(this, new PropertyChangedEventArgs("BackgroundColor")); } } } public MyClass() { this.BackgroundColor = "#FFFFFFFF"; this.PropertyChanged += MyClass_PropertyChanged; } public event PropertyChangedEventHandler PropertyChanged; void MyClass_PropertyChanged(object sender, System.ComponentModel.PropertyChangedEventArgs e) { //Do something here - fires client side but not server side } } I can listen to the event by externalizing it without any problems but it's an ugly way to handle something that want to set and forget inside my class e.g.: public class SomeOtherClass { public SomeOtherClass() { MyClass mc = new MyClass(); mc.PropertyChanged += MyClass_PropertyChanged; } void MyClass_PropertyChanged(object sender, System.ComponentModel.PropertyChangedEventArgs e) { MyClass mc = (MyClass)sender; mc.UpdateChildren(); } }

    Read the article

  • SQL SERVER – SHRINKFILE and TRUNCATE Log File in SQL Server 2008

    - by pinaldave
    Note: Please read the complete post before taking any actions. This blog post would discuss SHRINKFILE and TRUNCATE Log File. The script mentioned in the email received from reader contains the following questionable code: “Hi Pinal, If you could remember, I and my manager met you at TechEd in Bangalore. We just upgraded to SQL Server 2008. One of our jobs failed as it was using the following code. The error was: Msg 155, Level 15, State 1, Line 1 ‘TRUNCATE_ONLY’ is not a recognized BACKUP option. The code was: DBCC SHRINKFILE(TestDBLog, 1) BACKUP LOG TestDB WITH TRUNCATE_ONLY DBCC SHRINKFILE(TestDBLog, 1) GO I have modified that code to subsequent code and it works fine. But, are there other suggestions you have at the moment? USE [master] GO ALTER DATABASE [TestDb] SET RECOVERY SIMPLE WITH NO_WAIT DBCC SHRINKFILE(TestDbLog, 1) ALTER DATABASE [TestDb] SET RECOVERY FULL WITH NO_WAIT GO Configuration of our server and system is as follows: [Removed not relevant data]“ An email like this that suddenly pops out in early morning is alarming email. Because I am a dead, busy mind, so I had only one min to reply. I wrote down quickly the following note. (As I said, it was a single-minute email so it is not completely accurate). Here is that quick email shared with all of you. “Hi Mr. DBA [removed the name] Thanks for your email. I suggest you stop this practice. There are many issues included here, but I would list two major issues: 1) From the setting database to simple recovery, shrinking the file and once again setting in full recovery, you are in fact losing your valuable log data and will be not able to restore point in time. Not only that, you will also not able to use subsequent log files. 2) Shrinking file or database adds fragmentation. There are a lot of things you can do. First, start taking proper log backup using following command instead of truncating them and losing them frequently. BACKUP LOG [TestDb] TO  DISK = N'C:\Backup\TestDb.bak' GO Remove the code of SHRINKING the file. If you are taking proper log backups, your log file usually (again usually, special cases are excluded) do not grow very big. There are so many things to add here, but you can call me on my [phone number]. Before you call me, I suggest for accuracy you read Paul Randel‘s two posts here and here and Brent Ozar‘s Post here. Kind Regards, Pinal Dave” I guess this post is very much clear to you. Please leave your comments here. As mentioned, this is a very huge subject; I have just touched a tip of the ice-berg and have tried to point to authentic knowledge. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, SQL, SQL Authority, SQL Backup and Restore, SQL Data Storage, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL SERVER – Fix : Error : 3117 : The log or differential backup cannot be restored because no files

    - by pinaldave
    I received the following email from one of my readers. Dear Pinal, I am new to SQL Server and our regular DBA is on vacation. Our production database had some problem and I have just restored full database backup to production server. When I try to apply log back I am getting following error. I am sure, this is valid log backup file. Screenshot is attached. [Few other details regarding server/ip address removed] Msg 3117, Level 16, State 1, Line 1 The log or differential backup cannot be restored because no files are ready to roll forward. Msg 3013, Level 16, State 1, Line 1 RESTORE LOG is terminating abnormally. Screenshot attached. [Removed as it contained live IP address] Please help immediately. Well I have answered this question in my earlier post, 2 years ago, over here SQL SERVER – Fix : Error : Msg 3117, Level 16, State 4 The log or differential backup cannot be restored because no files are ready to rollforward. However, I will try to explain it a little more this time. For SQL Server database to be used it should in online state. There are multiple states of SQL Server Database. ONLINE (Available – online for data) OFFLINE RESTORING RECOVERING RECOVERY PENDING SUSPECT EMERGENCY (Limited Availability) If the database is online, it means it is active and in operational mode. It will not make sense to apply further log from backup if the operations have continued on this database. The common practice during the backup restore process is to specify the keyword RECOVERY when the database is restored. When RECOVERY keyword is specified, the SQL Server brings back the database online and will not accept any further log backups. However, if you want to restore more than one backup files, i.e. after restoring the full back up if you want to apply further differential or log backup you cannot do that when database is online and already active. You need to have your database in the state where it can further accept the backup data and not the online data request. If the SQL Server is online and also accepts database backup file, then there can be data inconsistency. This is the reason that when there are more than one database backup files to be restored, one has to restore the database with NO RECOVERY keyword in the RESTORE operation. I suggest you all to read one more post written by me earlier. In this post, I explained the time line with image and graphic SQL SERVER – Backup Timeline and Understanding of Database Restore Process in Full Recovery Model. Sample Code for reference: RESTORE DATABASE AdventureWorks FROM DISK = 'C:\AdventureWorksFull.bak' WITH NORECOVERY; RESTORE DATABASE AdventureWorks FROM DISK = 'C:\AdventureWorksDiff.bak' WITH RECOVERY; In this post, I am not trying to cover complete backup and recovery. I am just attempting to address one type of error and its resolution. Please test these scenarios on the development server. Playing with live database backup and recovery is always very crucial and needs to be properly planned. Leave a comment here if you need help with this subject. Similar Post: SQL SERVER – Restore Sequence and Understanding NORECOVERY and RECOVERY Note: We will cover Standby Server maintenance and Recovery in another blog post and it is intentionally, not covered this post. Reference : Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, Readers Question, SQL, SQL Authority, SQL Backup and Restore, SQL Error Messages, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • Why does "commit" appear in the mysql slow query log?

    - by Tom
    In our MySQL slow query logs I often see lines that just say "COMMIT". What causes a commit to take time? Another way to ask this question is: "How can I reproduce getting a slow commit; statement with some test queries?" From my investigation so far I have found that if there is a slow query within a transaction, then it is the slow query that gets output into the slow log, not the commit itself. Testing In mysql command line client: mysql begin; Query OK, 0 rows affected (0.00 sec) mysql UPDATE members SET myfield=benchmark(9999999, md5('This is to slow down the update')) WHERE id = 21560; Query OK, 0 rows affected (2.32 sec) Rows matched: 1 Changed: 0 Warnings: 0 At this point (before the commit) the UPDATE is already in the slow log. mysql commit; Query OK, 0 rows affected (0.01 sec) The commit happens fast, it never appeared in the slow log. I also tried a UPDATE which changes a large amount of data but again it was the UPDATE that was slow not the COMMIT. However, I can reproduce a slow ROLLBACK that takes 46s and gets output to the slow log: mysql begin; Query OK, 0 rows affected (0.00 sec) mysql UPDATE members SET myfield=CONCAT(myfield,'TEST'); Query OK, 481446 rows affected (53.31 sec) Rows matched: 481446 Changed: 481446 Warnings: 0 mysql rollback; Query OK, 0 rows affected (46.09 sec) I understand why rollback has a lot of work to do and therefore takes some time. But I'm still struggling to understand the COMMIT situation - i.e. why it might take a while.

    Read the article

< Previous Page | 14 15 16 17 18 19 20 21 22 23 24 25  | Next Page >