Search Results

Search found 5623 results on 225 pages for 'inline assembly'.

Page 28/225 | < Previous Page | 24 25 26 27 28 29 30 31 32 33 34 35  | Next Page >

  • What makes signed integers behave differently?

    - by 000
    In this example of x86_64 hex/disassembled code I see: 48B80000000000000000 mov rax, 0x0 Signed Byte 52 Unsigned Byte 52 Signed Short 14388 Unsigned Short 14388 Signed Int 943863860 Unsigned Int 943863860 Signed Int64 3472328296363079732 Unsigned Int64 3472328296363079732 Float 4.630555e-05 Double 1.39804332763832e-76 String 48B80000000000000000 which to me appears to have the same functionality as: 48C7C000000000 mov rax, 0x0 48C7C000000000 Signed Byte 52 Unsigned Byte 52 Signed Short 14388 Unsigned Short 14388 Signed Int 927152180 Unsigned Int 927152180 Signed Int64 3472328377950746676 Unsigned Int64 3472328377950746676 Float 1.163599e-05 Double 1.39806836023098e-76 String 48C7C000000000 How is the first example treated differently from the second example?

    Read the article

  • Implementing a bit shift using AND, NOT, ADD [closed]

    - by fdart17
    I'm implementing a 16-bit left bit shift by r bits, and I only have access to AND, NOT and ADD. There are 3 condition codes, negative, zero and positive, which are set when you use any of these operations. How I went about it was : (1) And the number with 1000 0000 0000 0000 to set condition codes to positive if the most significant bit is 1. (2) Add the number with itself. This shifts bits one to the left. (3) If the MSB was 1, add 1 to the result. (4) Loop threw (1)-(3) r times. I'm wondering if anyone has any hints to some more efficient methods? Thanks!

    Read the article

  • How Byte loading/storing is implemented By the CPU?

    - by AlexDan
    I know that in 32bit machine, cpu read from memory 32bits at a time. since the registers in this case is 32bit in size too, I can understand how this works. What I don't understand is how the cpu implement load instructions of 1 byte. does it load the whole word where the single byte is located to the register, then perform some kind of "byte shifting", or does the cpu can load a single byte, in this case when does the byte masking happen, is it until the byte got loaded in the register, or it happen when byte is send through the data bus ? P.S. The cpu Im using is MIPS, the instructions Im talking about are: lb or lbu

    Read the article

  • please look my first project [closed]

    - by gökhan
    good night my project is about that Write a program that uses a loop to calculate the first 15 values of the Fibonacci number sequence and store them in an array, call it FibArr, of size 15 DWORDS. Fibonacci number sequence is described by the following formula: Fib(1) = 1, Fib(2) = 1, Fib(n) = Fib(n -1) + Fib(n - 2). Place each value in the EAX register and display it with a call DumpRegs statement.So I dont have an idea about this topic please help me thanks

    Read the article

  • django crispy-forms inline forms

    - by abolotnov
    I'm trying to adopt crispy-forms and bootstrap and use as much of their functionality as possible instead of inventing something over and over again. Is there a way to have inline forms functionality with crispy-forms/bootstrap like django-admin forms have? Here is an example: class NewProjectForm(forms.Form): name = forms.CharField(required=True, label=_(u'???????? ???????'), widget=forms.TextInput(attrs={'class':'input-block-level'})) group = forms.ModelChoiceField(required=False, queryset=Group.objects.all(), label=_(u'?????? ????????'), widget=forms.Select(attrs={'class':'input-block-level'})) description = forms.CharField(required=False, label=_(u'???????? ???????'), widget=forms.Textarea(attrs={'class':'input-block-level'})) class Meta: model = Project fields = ('name','description','group') def __init__(self, *args, **kwargs): self.helper = FormHelper() self.helper.form_class = 'horizontal-form' self.helper.form_action = 'submit_new_project' self.helper.layout = Layout( Field('name', css_class='input-block-level'), Field('group', css_class='input-block-level'), Field('description',css_class='input-block-level'), ) self.helper.add_input(Submit('submit',_(u'??????? ??????'))) self.helper.add_input(Submit('cancel',_(u'? ?????????'))) super(NewProjectForm, self).__init__(*args, **kwargs) it will display a decent form: How do I go about adding a form that basically represents this model: class Link(models.Model): name = models.CharField(max_length=255, blank=False, null=False, verbose_name=_(u'????????')) url = models.URLField(blank=False, null=False, verbose_name=_(u'??????')) project = models.ForeignKey('Project') So there will be a project and name/url links and way to add many, like same thing is done in django-admin where you are able to add extra 'rows' with data related to your main model. On the sreenshot below you are able to fill out data for 'Question' object and below that you are able to add data for QuestionOption objects -you are able to click the '+' icon to add as many QuestionOptions as you want. I'm not looking for a way to get the forms auto-generated from models (that's nice but not the most important) - is there a way to construct a form that will let you add 'rows' of data like django-admin does?

    Read the article

  • The ORDER BY clause is invalid in views, inline functions, derived tables, subqueries, and common ta

    - by zurna
    I get "The ORDER BY clause is invalid in views, inline functions, derived tables, subqueries, and common table expressions, unless TOP or FOR XML is also specified." error with the following code. I initially had two tables, ADSAREAS & CATEGORIES. I started receiving this error when I removed CATEGORIES table. Select Case SIDX Case "ID" : SQLCONT1 = " AdsAreasID" Case "Page" : SQLCONT1 = " AdsAreasName" Case Else : SQLCONT1 = " AdsAreasID" End Select Select Case SORD Case "asc" : SQLCONT2 = " ASC" Case "desc" : SQLCONT2 = " DESC" Case Else : SQLCONT2 = " ASC" End Select ''# search feature ---> Select Case SEARCHFIELD Case "ID" : SQLSFIELD = "AND AdsAreasID" Case "Ads Areas" : SQLSFIELD = "AND AdsAreasName" Case Else : SQLSFIELD = "" End Select Select Case SEARCHOPER Case "eq" : SQLSOPER = " = " & SEARCHSTRING Case "ne" : SQLSOPER = " <> " & SEARCHSTRING Case "lt" : SQLSOPER = " <" & SEARCHSTRING Case "le" : SQLSOPER = " <= " & SEARCHSTRING Case "gt" : SQLSOPER = " >" & SEARCHSTRING Case "ge" : SQLSOPER = " >= " & SEARCHSTRING Case "bw" : SQLSOPER = " LIKE '" & SEARCHSTRING & "%' " Case "ew" : SQLSOPER = " LIKE '%" & SEARCHSTRING & "' " Case "cn" : SQLSOPER = " LIKE '%" & SEARCHSTRING & "%' " Case Else : SQLSOPER = "" End Select ''# search feature ---> SQL = "SELECT * FROM ( SELECT A.AdsAreasID, A.AdsAreasName, ROW_NUMBER() OVER (ORDER BY A.AdsAreasID) As Row" SQL = SQL & " FROM ADSAREAS A" SQL = SQL & " WHERE Row > ("& RecordsPageSize - RecordsPerPage &") AND Row <= ("& RecordsPageSize &") ORDER BY" & SQLCONT1 & SQLCONT2 Set objXML = objConn.Execute(SQL)

    Read the article

  • Variable-width inline underline effects in CSS

    - by sidereal
    I need to simulate the look of a typical paper form in CSS. It consists of a two-column table of fields. Each field consists of a field name (of variable width) followed by an underline that continues to the end of the column. The field might be populated, in which case there is some text centered above the line, or it may be blank. If that isn't clear, he's a rough idea in manky ASCII art: Name: _______Foo_______ Age: _____17______ Location: __Melbourne__ Handedness: _Left_ (except that the underline would continue under any text) To implement the underline without text, I assume I should use a border-bottom rather than a text-decoration: underline. Additionally, I need the bordered element to take up the full available space. Both of those argue for a block-level element. However, I can't find any way to get the block level element (either a div, an li, or a span set to display: block or inline-block) to remain on the same line as the label. As soon as I give it a width: 100%, it newlines. I've tried various combinations of floats, and I'm not inclined to do anything ridiculous with absolute positioning. Any recommendations?

    Read the article

  • Formatting inline many-to-many related models presented in django admin

    - by Jonathan
    I've got two django models (simplified): class Product(models.Model): name = models.TextField() price = models.IntegerField() class Invoice(models.Model): company = models.TextField() customer = models.TextField() products = models.ManyToManyField(Product) I would like to see the relevant products as a nice table (of product fields) in an Invoice page in admin and be able to link to the individual respective Product pages. My first thought was using the admin's inline - but django used a select box widget per related Product. This isn't linked to the Product pages, and also as I have thousands of products, and each select box independently downloads all the product names, it quickly becomes unreasonably slow. So I turned to using ModelAdmin.filter_horizontal as suggested here, which used a single instance of a different widget, where you have a list of all Products and another list of related Products and you can add\remove products in the later from the former. This solved the slowness, but it still doesn't show the relevant Product fields, and it ain't linkable. So, what should I do? tweak views? override ModelForms? I Googled around and couldn't find any example of such code...

    Read the article

  • Microsoft C Compiler: Inline variable declaration?

    - by Rosarch
    I'm writing C in Visual Studio 2010. The compiler doesn't seem to want to let me use inline variable declarations. The following code produces an error: unsigned int fibonacci_iterative(unsigned int n) { if (n == 0) { return 0; } if (n == 1) { return 1; } unsigned int prev_prev = 0; // error unsigned int prev = 1; // error unsigned int next = 0; // error for (int term_number = 0; term_number < n; term_number++) { unsigned int temp = prev_prev + prev; prev = next; prev_prev = prev; next = temp; } return next; } Error: error C2143: syntax error : missing ';' before 'type' error C2143: syntax error : missing ';' before 'type' error C2143: syntax error : missing ';' before 'type' Why is this happening? Is there a setting to make the compiler not so strict?

    Read the article

  • Symfony 1.3: how to force as inline the choices generated by renderLabel()

    - by user248959
    Hi, this line: <li><?php echo $form['genero']->renderLabel() ?></li> is generating <li> <label for="usuario_genero">Genero</label> <ul class="radio_list"> <li> <!-- this li doesn't have any id--> <input type="radio" checked="checked" id="usuario_genero_0" value="0" name="usuario[genero]">&nbsp;<label for="usuario_genero_0">Chico</label> </li> <li> <!-- this li doesn't have any id--> <input type="radio" id="usuario_genero_1" value="1" name="usuario[genero]">&nbsp;<label for="usuario_genero_1">Chica</label> </li> </ul> </li> I'd like to force the choices as inline, but the li's generated don't have the 'id' attribute. What should i do? Regards Javi

    Read the article

  • Closing the gap between 2 inline elements

    - by insanepaul
    I have a simple div element that contains 2 inline tags. I've associated a onmouseout event to the div element. The problem is that the onmouseout event is fired when the user hovers their mouse between the two tags in the div and also after the end of the second tag. What I want to do is allow the user to hover their mouse across the whole of the div tag and only fire the onmouseout event when the mouse pointer is outside the div element (which is what I assumed from what I've done). I increased padding to close the gap between the 2 tags. This works but where they meet in IE7 at least the event is fired!!! I must be doing something wrong can someone please help. <div id="Div1" onmouseover="hideDiv()"> <a id="A1" href="HTMLNew.htm">ARTICLES</a> <a id="A2" href="HTMLNew.htm">COURSES & CASES</a> </div>

    Read the article

  • jQuery > Update inline script on form submission

    - by Andrew Kirk
    I am using the ChemDoodle Web Components to display molecules on a web page. Basically, I can just insert the following script in my page and it will create an HTML5 canvas element to display the molecule. <script> var transform1 = new TransformCanvas('transform1', 200, 200, true); transform1.specs.bonds_useJMOLColors = true; transform1.specs.bonds_width_2D = 3; transform1.specs.atoms_useJMOLColors = true; transform1.specs.atoms_circles_2D = true; transform1.specs.backgroundColor = 'black'; transform1.specs.bonds_clearOverlaps_2D = true; transform1.loadMolecule(readPDB(molecule)); </script> In this example, "molecule" is a variable that I have defined in an external script by using the jQuery.ajax() function to load a PDB file. This is all fine and good. Now, I would like to include a form on the page that will allow a user to paste in a PDB molecule definition. Upon submitting the form, I want to update the "molecule" variable with the form data so that the ChemDoodle Web Components script will work its magic and display molecule defined by the PDB definition pasted into the form. I am using the following jQuery code to process the the form submission. $(".button").click(function() { // validate and process form here //hide previous errors $('.error').hide(); //validate pdb textarea field var pdb = $("textarea#pdb").val(); if (pdb == "") { $("label#pdb_error").show(); $("textarea#pdb").focus(); return false; } molecule = pdb; }); This code is setting the "molecule" variable upon the form submission but it is not being passed back to the inline script as I had hoped. I have tried a number of variations on this but can't seem to get it right. Any clues as to where I might be going wrong would be much appreciated.

    Read the article

  • SPAN vs DIV (inline-block)

    - by blackjid
    Hi, Is there any reason to use a <div style="display:inline-block"> instead of a <span> to layout a webpage? Can I put content nested inside the span? What is valid and what isn't? Thanks! It's ok to use this to make a 3x2 table like layout? <div> <span> content1(divs,p, spans, etc) </span> <span> content2(divs,p, spans, etc) </span> <span> content3(divs,p, spans, etc) </span> </div> <div> <span> content4(divs,p, spans, etc) </span> <span> content5(divs,p, spans, etc) </span> <span> content6(divs,p, spans, etc) </span> </div>

    Read the article

  • How to provide an inline model field with a queryset choices without losing field value for inline r

    - by Judith Boonstra
    The code displayed below is providing the choices I need for the app field, and the choices I need for the attr field when using Admin. I am having a problem with the attr field on the inline form for already saved records. The attr selected for these saved does show in small print above the field, but not within the field itself. # MODELS: Class Vocab(models.Model): entity = models.Charfield, max_length = 40, unique = True) Class App(models.Model): name = models.ForeignKey(Vocab, related_name = 'vocab_appname', unique = True) app = SelfForeignKey('self, verbose_name = 'parent', blank = True, null = True) attr = models.ManyToManyField(Vocab, related_name = 'vocab_appattr', through ='AppAttr' def parqs(self): a method that provides a queryset consisting of available apps from vocab, excluding self and any apps within the current app's dependent line. def attrqs(self): a method that provides a queryset consisting of available attr from vocab excluding those already selected by current app, 2) those already selected by any apps within the current app's parent line, and 3) those selected by any apps within the current app's dependent line. Class AppAttr(models.Model): app = models.ForeignKey(App) attr = models.ForeignKey(Vocab) # FORMS: from models import AppAttr def appattr_form_callback(instance, field, *args, **kwargs) if field.name = 'attr': if instance: return field.formfield(queryset = instance.attrqs(), *kwargs) return field.formfield(*kwargs) # ADMIN: necessary imports class AppAttrInline(admin.TabularInline): model = AppAttr def get_formset(self, request, obj = None, **kwargs): kwargs['formfield_callback'] = curry(appattr_form_callback, obj) return super(AppAttrInline, self).get_formset(request, obj, **kwargs) class AppForm(forms.ModelForm): class Meta: model = App def __init__(self, *args, **kwargs): super(AppForm, self).__init__(*args, **kwargs) if self.instance.id is None: working = App.objects.all() else: thisrec = App.objects.get(id = self.instance.id) working = thisrec.parqs() self.fields['par'].queryset = working class AppAdmin(admin.ModelAdmin): form = AppForm inlines = [AppAttrInline,] fieldsets = .......... necessary register statements

    Read the article

  • Inline text box labels with WPF

    - by Douglas
    I'm trying to reproduce the layout of some paper forms in a WPF application. Labels for text boxes are to be "inline" with the content of the text boxes, rather than "outside" like normal Windows forms. So, with an Xxxxxx label: +-----------------------------+ | Xxxxxx: some text written | | in the multiline input. | | | | another paragraph continues | | without indentation. | | | | | +-----------------------------+ The Xxxxxx cannot be editable, if the user selects all the content of the text box, the label must remain unselected, I need to be able to style the text colour/formatting of the label separately, when there is no text in the text box, but it has focus, the caret should flash just after the label, and I need the baselines of the text in the text box and the label to line up. One solution I tried was putting a textblock partially over the input, then using text indent to indent the editable text, though this caused problems with following paragraphs, since they were indented too. I'm not sure how to indent just the first paragraph. It required some fiddling to get the text to line up - a more reliable setup would be ideal. So, any suggestions on how to set this up? Thanks

    Read the article

  • C++ using typedefs in non-inline functions

    - by ArunSaha
    I have a class like this template< typename T > class vector { public: typedef T & reference; typedef T const & const_reference; typedef size_t size_type; const_reference at( size_t ) const; reference at( size_t ); and later in the same file template< typename T > typename vector<T>::const_reference // Line X vector<T>::at( size_type i ) const { rangecheck(); return elems_[ i ]; } template< typename T > reference // Line Y vector<T>::at( size_type i ) { rangecheck(); return elems_[ i ]; } Line X compiles fine but Line Y does not compile. The error message from g++ (version 4.4.1) is: foo.h:Y: error: expected initializer before 'vector' From this I gather that, if I want to have non-inline functions then I have to fully qualify the typedef name as in Line X. (Note that, there is no problem for size_type.) However, at least to me, Line X looks clumsy. Is there any alternative approach?

    Read the article

  • Inline horizontal spacer in HTML

    - by LeafStorm
    I am making a Web page with a slideshow, using the same technique used on http://zine.pocoo.org/. The person I'm making the site for wants the slideshow to be centered. However, some of the photos are portrait layout and some are landscape. (This was not my choice.) I need a position: absolute to get the li's containing the items in the right place, so centering them does not work. (At least, not by normal methods.) So, I thought that it might work to insert a 124-pixel "spacer" before the image on the portrait pictures. I tried it with a <span style="width: 124px;">&nbsp;</span>, but it only inserts a single space, not the full 124 pixels. The slideshow fades in and out OK, though, so I think that it would work if I could get the proper spacing. My question is this: does anyone know a way to have 124px of space inline in HTML (preferably without using images), or another way to center the pictures in the li items?

    Read the article

  • Use external inline script as local function

    - by Aidan
    Had this closed once as a duplicate, yet the so-called duplicate DID NOT actually address my whole question. I have found this script that, when run inline, returns your IP. <script type="text/javascript" src="http://l2.io/ip.js"></script> http://l2.io/ip.js Has nothing more than a line of code that says document.write('123.123.123.123'); (But obviously with the user's IP address) I want to use this IP address as a return string for a function DEFINED EXTERNALLY, BUT STILL ON MY DOMAIN. That is, I have a "scripts.js" that contains all the scripts I wish to use, and I would like to include it in that list as a local function that calls to the 12.io function, but javascript won't allow the < tags, so I am unsure as to how to do this. I.e. function getIP() { return (THAT SCRIPT'S OUTPUT); } This is the topic this was supposedly a duplicate of, and it is very similar. Get ip address with javascript However, this DOES NOT address defining as a forwarded script it in my own script file.

    Read the article

  • using ILMerge with .NET 4 libraries

    - by Sarah Vessels
    I'm having trouble using ILMerge in my post-build after upgrading from .NET 3.5/Visual Studio 2008 to .NET 4/Visual Studio 2010. I have a Solution with several projects whose target framework is set to ".NET Framework 4". I use the following ILMerge command to merge the individual project DLLs into a single DLL: if not $(ConfigurationName) == Debug if exist "C:\Program Files (x86)\Microsoft\ILMerge\ILMerge.exe" "C:\Program Files (x86)\Microsoft\ILMerge\ILMerge.exe" /lib:"C:\Windows\Microsoft.NET\Framework64\v4.0.30319" /lib:"C:\Program Files (x86)\Microsoft Visual Studio 10.0\Common7\IDE\PublicAssemblies" /keyfile:"$(SolutionDir)$(SolutionName).snk" /targetplatform:v4 /out:"$(SolutionDir)bin\development\$(SolutionName).dll" "$(SolutionDir)Connection\$(OutDir)Connection.dll" ...other project DLLs... /xmldocs If I leave off specifying the location of the .NET 4 framework directory, I get an "Unresolved assembly reference not allowed: System" error from ILMerge. If I leave off specifying the location of the MSTest directory, I get an "Unresolved assembly reference not allowed: Microsoft.VisualStudio.QualityTools.UnitTestFramework" error. The ILMerge command above works and produces a DLL. When I reference that DLL in another .NET 4 C# project, however, and try to use code within it, I get the following warning: The primary reference "MyILMergedDLL" could not be resolved because it has an indirect dependency on the .NET Framework assembly "mscorlib, Version=4.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" which has a higher version "4.0.65535.65535" than the version "4.0.0.0" in the current target framework. If I then remove the /targetplatform:v4 flag and try to use MyILMergedDLL.dll, I get the following error: The type 'System.Xml.Serialization.IXmlSerializable' is defined in an assembly that is not referenced. You must add a reference to assembly 'System.Xml, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089'. It doesn't seem like I should have to do that. Whoever uses my MyILMergedDLL.dll API should not have to add references to whatever libraries it references. How can I get around this?

    Read the article

  • Visual Studio 2010 error: Type universe cannot resolve assembly

    - by dthrasher
    I've loaded a WPF project initially created in Visual Studio 2008 into Visual Studio 2010. The conversion process goes smoothly, but on certain XAML files the VS2010 designer throws several errors related to project references, including this one: System.Reflection.Adds.UnresolvedAssemblyException Type universe cannot resolve assembly: GalaSoft.MvvmLight, Version=3.0.0.31869, Culture=neutral, PublicKeyToken=3e875cdb3903c512. This assembly reference works just fine in the Expression Blend 4 designer, but not in VS2010. I can build and run the solution successfully. My solution targets the .Net Framework 3.5 SP1.

    Read the article

  • Resolving MSB3247 - Found conflicts between different versions of the same dependent assembly

    - by David Gardiner
    A .NET 3.5 solution ended up with this warning when compiling with msbuild. Sometimes NDepend might help out but in this case it didn't give any further details. Like Bob I ended up having to resort to opening each assembly in ILDASM until I found the one that was referencing an older version of the dependant assembly. I did try using MSBUILD from VS 2010 Beta 2 (as the Connect article indicated this was fixed in the next version of the CLR) but that didn't provide any more detail either (maybe fixed post Beta 2) Is there a better (more automated) approach?

    Read the article

  • reporting services: use a custom assembly with a local (RDLC) report

    - by JMarsch
    Hello: I am designing a report that will be used in local mode (an RDLC file) in a Winform app. I have a custom assembly with a static class that has some functions that I want to use inside of the report (as expressions). I have found all sorts of help for doing this with RDL reports, but I'm running into a permissions problem with my RDLC report. I get the following error at runtime: "The report references the code module (my module), which is not a trusted assembly". I know that this is some kind of a code security issue, but I'm not sure what to do to fix it. The documentation that I have seen online is aimed at RDL reports, and it instructs me to edit a SQL Server-specific policy file. I'm using RDLC, so there is no sql server involved. What do I need to do to acquire the appropriate permissions?

    Read the article

  • Enumerating .NET assembly resources at runtime

    - by Alex_P
    I have a resource assembly with image files in it that are built using Resource or Content build action. This makes these files accessible using the Uris. However I cannot find the way to enumerate such resources. If I set the build action to Embedded Resource it becomes possible to enumerate the files with the following code: string[] resources = Assembly.GetExecutingAssembly().GetManifestResourceNames(); but it in turn makes these files inaccessible using Uris. The question is - how to enumerate resources that are compiled with either Resource or Content build action? NOTE: As Thomas Levesque pointed out it is possible to enumerate such resources by leveraging the AssemblyAssociatedContentFileAttribute, but it seems to only work for WPF Application assemblies and not for class library ones. So the question is still open.

    Read the article

  • Override the neutral language of a specific resource file within an assembly

    - by Sandor Drieënhuizen
    I have an assembly that contains several resource files. Most of them have the neutral language 'nl' (Dutch, specified on the assembly as the neutral language), so I don't specify the 'nl' in their filenames. However, I'm putting strings in the English language in some other resource files (they are internal error messages) and I will never provide Dutch translations of them. If I name those resource files something like 'Errors.en.resx', no designer class is generated (breaks the build) because there is no 'Errors.resx'. This is annoying because now I have to put 'en' strings into a 'nl'-implied resource file and I really don't want to translate those strings to 'nl' or provide empty strings just to satisfy the compiler. Is there a way to override the neutral language on a specific resource file or perhaps somehow have the 'Errors.en.resx' build a designer class?

    Read the article

  • WCF XmlSerializer assembly not speeding up first request

    - by Matt Dearing
    I am generating proxy classes to a clients java webservice wsdls and xsd files with svcutil. The first call made to each service proxy class takes a very long time. I was hoping to speed this up by generating the XmlSerializers assembly myself (based on the article How to: Improve the Startup Time of WCF Client Applications using the XmlSerializer), but when I do the first call to each service still takes the same amount of time. Here are the steps I am following: //generate strong name key file sn -k Blah.snk //generate the proxy class file svcutil blah.wsdl blah2.wsdl blah3.wsdl ... base.xsd blah.xsd ... /UseSerializerForFaults /ser:XmlSerializer /n:*,SomeNamespace /out:Blah.cs //compile the class into an assembly signing it with the strong name key file csc /target:library /keyfile:Blah.snk /out:Blah.dll Blah.cs //generate the XmlSerializer code this will give us Blah.XmlSerializers.dll.cs svcutil /t:xmlSerializer Blah.dll //compile the xmlserializer code into its own dll using the same key to sign it and referencing the original dll csc /target:library /keyfile:Blah.snk /out:Blah.XmlSerializers.dll Blah.XmlSerializers.dll.cs /r:Blah.dll I then create a standard Console application that references both Blah.dll and Blah.XmlSerializers.dll. I will then try something like: //BlahProxy is one of the generated service proxy classes BlahProxy p = new BlahProxy(); //this call takes 30ish seconds p.SomeMethod(); BlahProxy p2 = new BlahProxy(); //this call takes < 1 second p2.SomeMethod(); //BlahProx2y is one of the generated service proxy classes BlahProxy2 p3 = new BlahProxy2(); //this call takes 30ish seconds p3.SomeMethod(); BlahProxy2 p4 = new BlahProxy2(); //this call takes < 1 second p4.SomeMethod(); I know that the problem is not server side because I don't see the request made in Fiddler until around 29 seconds. Subsequent calls to each service take < 1 second, so thats why I was hoping the main slow down was the .net runtime generating the xmlserializer code itself, compiling it and loading the assembly. I figured this would be the reason the first call to each service is slow and the rest are fast. Unfortunatley, me generating the code myself is not speeding anything up. Does anyone see what I am doing wrong?

    Read the article

< Previous Page | 24 25 26 27 28 29 30 31 32 33 34 35  | Next Page >