Search Results

Search found 1806 results on 73 pages for 'opinion'.

Page 67/73 | < Previous Page | 63 64 65 66 67 68 69 70 71 72 73  | Next Page >

  • HowTo stick QDialog to Screen Borders like Skype do?

    - by mosg
    Hello. A long time ago I tried to find method how to stick QDialog window to screen borders for my small projects like Skype windows do it, but I failed. May be I was looking this code not in the right place, so now I'm looking the solution here, on stack! :) So, does any one have a deal with some kind of such code, links, samples? In my opinion, we have to reimplement QDialog moveEvent function, like below, but that code does not working: void CDialog::moveEvent(QMoveEvent * event) { QRect wndRect; int leftTaskbar = 0, rightTaskbar = 0, topTaskbar = 0, bottomTaskbar = 0; // int top = 0, left = 0, right = 0, bottom = 0; wndRect = this->frameGeometry(); // Screen resolution int screenWidth = QApplication::desktop()->width(); int screenHeight = QApplication::desktop()->height(); int wndWidth = wndRect.right() - wndRect.left(); int wndHeight = wndRect.bottom() - wndRect.top(); int posX = event->pos().x(); int posY = event->pos().y(); // Snap to screen border // Left border if (posX >= -m_nXOffset + leftTaskbar && posX <= leftTaskbar + m_nXOffset) { //left = leftTaskbar; this->move(leftTaskbar, posY); return; } // Top border if (posY >= -m_nYOffset && posY <= topTaskbar + m_nYOffset) { //top = topTaskbar; this->move(posX, topTaskbar); return; } // Right border if (posX + wndWidth <= screenWidth - rightTaskbar + m_nXOffset && posX + wndWidth >= screenWidth - rightTaskbar - m_nXOffset) { //right = screenWidth - rightTaskbar - wndWidth; this->move(screenWidth - rightTaskbar - wndWidth, posY); return; } // Bottom border if (posY + wndHeight <= screenHeight - bottomTaskbar + m_nYOffset && posY + wndHeight >= screenHeight - bottomTaskbar - m_nYOffset) { //bottom = screenHeight - bottomTaskbar - wndHeight; this->move(posX, screenHeight - bottomTaskbar - wndHeight); return; } QDialog::moveEvent(event); } Thanks.

    Read the article

  • Is it ok to use <strong> in place of <b> blindly ?

    - by metal-gear-solid
    Note: I know <b> is presentational and <span style="font-weight:bold> is better way and <strong> and <em> for emphasis but my question is not regarding this. Should we convert every <b> to <strong> blindly ? many people do this they think <b> is not good as per web standards so they convert every <b> to <strong> upon site redesign, content re-population, new site design and people suggest this to others also. Dreamweaver has also given option to convert all <b> and <i> to <strong> and <em> on code paste in design view and when we use B and I Which people use blindly. And Dreamweaver(if above option is checked) and many online WYSIWYG editor give output in <strong> and <em> while button shows B and I. In my opinion it's creating misconception about <strong> and <b> When we get content from client we don't know where client want to give emphasize and where he just want to use text bold for presentation purpose.What we should do in this condition? no one has time client and we both to give time to think for each bold text, it should be in <b> or in <strong>, <i> or in <em> What are pros and cons to convert every <b> and <i> into <strong> and <em> blindly if we are saying our site is accessible? Update: remember <b> and <i> are not deprecated there are in HTML 5 specification

    Read the article

  • How build my own Application Setting

    - by adisembiring
    I want to build a ApplicationSetting for my application. The ApplicationSetting can be stored in a properties file or in a database table. The settings are stored in key-value pairs. E.g. ftp.host = blade ftp.username = dummy ftp.pass = pass content.row_pagination = 20 content.title = How to train your dragon. I have designed it as follows: Application settings reader: interface IApplicationSettingReader { read(); } DatabaseApplicationSettingReader { dao appSettingDao; AppSettings read() { List<AppSettingEntity> listEntity = appSettingsDao.findAll(); Map<String, String> map = new HaspMap<String, String>(); foreach (AppSettingEntity entity : listEntity) { map.put(entity.getConfigName(), entity.getConfigValue()); } return new AppSettings(map); } } DatabaseApplicationSettingReader { dao appSettingDao; AppSettings read() { //read from some properties file return new AppSettings(map); } } Application settings class: AppSettings { private static AppSettings instance; private Map map; Public AppSettings(Map map) { this.map = map; } public static AppSettings getInstance() { if (instance == null) { throw new RuntimeException("Object not configure yet"); } return instance; } public static configure(IApplicationSettingReader reader) { instance = reader.read(); } public String getFtpSetting(String param) { return map.get("ftp." + param); } public String getContentSetting(String param) { return map.get("content." + param); } } Test class: AppSettingsTest { IApplicationSettingReader reader; @Before public void setUp() throws Exception { reader = new DatabaseApplicationSettingReader(); } @Test public void getContentSetting_should_get_content_title() { AppSettings.configure(reader); Instance settings = AppSettings.getInstance(); String title = settings.getContentSetting("title"); assertNotNull(title); Sysout(title); } } My questions are: Can you give your opinion about my code, is there something wrong? I configure my application setting once, while the application start, I configure the application setting with appropriate reader (DbReader or PropertiesReader), I make it singleton. The problem is, when some user edit the database or file directly to database or file, I can't get the changed values. Now, I want to implement something like ApplicationSettingChangeListener. So if the data changes, I will refresh my application settings. Do you have any suggestions how this can be implemented?

    Read the article

  • Why does my ko computed observable not update bound UI elements when its value changes?

    - by Allen
    I'm trying to wrap a cookie in a computed observable (which I'll later turn into a protectedObservable) and I'm having some problems with the computed observable. I was under the opinion that changes to the computed observable would be broadcast to any UI elements that have been bound to it. I've created the following fiddle JavaScript: var viewModel = {}; // simulating a cookie store, this part isnt as important var cookie = function () { // simulating a value stored in cookies var privateZipcode = "12345"; return { 'write' : function (val) { privateZipcode = val; }, 'read': function () { return privateZipcode; } } }(); viewModel.zipcode = ko.computed({ read: function () { return cookie.read(); }, write: function (value) { cookie.write(value); }, owner: viewModel }); ko.applyBindings(viewModel);? HTML: zipcode: <input type='text' data-bind="value: zipcode"> <br /> zipcode: <span data-bind="text: zipcode"></span>? I'm not using an observable to store privateZipcode since that's really just going to be in a cookie. I'm hoping that the ko.computed will provide the notifications and binding functionality that I need, though most of the examples I've seen with ko.computed end up using a ko.observable underneath the covers. Shouldn't the act of writing the value to my computed observable signal the UI elements that are bound to its value? Shouldn't these just update? Workaround I've got a simple workaround where I just use a ko.observable along side of my cookie store and using that will trigger the required updates to my DOM elements but this seems completely unnecessary, unless ko.computed lacks the signaling / dependency type functionality that ko.observable has. My workaround fiddle, you'll notice that the only thing that changes is that I added a seperateObservable that isn't used as a store, its only purpose is to signal to the UI that the underlying data has changed. // simulating a cookie store, this part isnt as important var cookie = function () { // simulating a value stored in cookies var privateZipcode = "12345"; // extra observable that isnt really used as a store, just to trigger updates to the UI var seperateObservable = ko.observable(privateZipcode); return { 'write' : function (val) { privateZipcode = val; seperateObservable(val); }, 'read': function () { seperateObservable(); return privateZipcode; } } }(); This makes sense and works as I'd expect because viewModel.zipcode depends on seperateObservable and updates to that should (and does) signal the UI to update. What I don't understand, is why doesn't a call to the write function on my ko.computed signal the UI to update, since that element is bound to that ko.computed? I suspected that I might have to use something in knockout to manually signal that my ko.computed has been updated, and I'm fine with that, that makes sense. I just haven't been able to find a way to accomplish that.

    Read the article

  • What does these FindBug messages show?

    - by Hans Klock
    Not every description from from http://findbugs.sourceforge.net/bugDescriptions.html is clear to me. Sure, I can study the implementation but if somebody is more experienced then me, some explanation and examples would be great. Do you have some examples for UI_INHERITANCE_UNSAFE_GETRESOURCE when this is getting a problem? In BX_UNBOXED_AND_COERCED_FOR_TERNARY_OPERATOR I don't see the problem either. If one type is "bigger" then the other, for example int and float, then the result is float. If its Integer and Float its the wrapper Float too. That's what I expect. Does the GC_UNRELATED_TYPES really help to find errors? Isn't it the job of the compiler to check, if--taking the given example--Foo can't go into a Collection<String>. Does HE_SIGNATURE_DECLARES_HASHING_OF_UNHASHABLE_CLASS mean something like bla(Foo f){hashtable.put(f);}, where ´Foo´ is not hashable? Does FingBugs "see" the subclasses too? NP_GUARANTEED_DEREF_ON_EXCEPTION_PATH is stronger "wrong" then NP_ALWAYS_NULL_EXCEPTION? Why two error cases and with NP_NULL_ON_SOME_PATH_EXCEPTION even one more? Sounds very similar to me. What is an example of SIO_SUPERFLUOUS_INSTANCEOF? Something like foo(String s){if (s intenceof String) .... This does a null check too, but this is not the test here... NN_NAKED_NOTIFY. I my opinion the description is not clear. A change of the state is not necessary. If I use new Object() to wait and notify on I don't change the object state. Or is state the lock-state? I don't get it. SP_SPIN_ON_FIELD. Can this really happen that a compiler will move this outside from a loop? This doesn't make sense to me because from outside a Thread can always change the values. And if the variable is volatile the JVM can't cache the value. So what's the meaning? That is the difference between STCAL_STATIC_CALENDAR_INSTANCE and STCAL_INVOKE_ON_STATIC_CALENDAR_INSTANCE or STCAL_INVOKE_ON_STATIC_DATE_FORMAT_INSTANCE/STCAL_STATIC_SIMPLE_DATE_FORMAT_INSTANCE? Why is XXXX.class in WL_USING_GETCLASS_RATHER_THAN_CLASS_LITERAL better then getClass()? A getClass() in a superclass called from the subclass will always return the Class object from the subclass which is good I think. What exactly does EQ_UNUSUAL do? It should check that the argument is of the same type of the class itself but it does't? Did you ever had problems with breaks? Is there real value with SF_SWITCH_FALLTHROUGH? Sounds to strong for me. No idea what TQ_EXPLICIT_UNKNOWN_SOURCE_VALUE_REACHES_ALWAYS_SINK and TQ_EXPLICIT_UNKNOWN_SOURCE_VALUE_REACHES_NEVER_SINK could be.

    Read the article

  • How does the rsync algorithm correctly identify repeating blocks?

    - by Kai
    I'm on a personal quest to learn how the rsync algorithm works. After some reading and thinking, I've come up with a situation where I think the algorithm fails. I'm trying to figure out how this is resolved in an actual implementation. Consider this example, where A is the receiver and B is the sender. A = abcde1234512345fghij B = abcde12345fghij As you can see, the only change is that 12345 has been removed. Now, to make this example interesting, let's choose a block size of 5 bytes (chars). Hashing the values on the sender's side using the weak checksum gives the following values list. abcde|12345|fghij abcde -> 495 12345 -> 255 fghij -> 520 values = [495, 255, 520] Next we check to see if any hash values differ in A. If there's a matching block we can skip to the end of that block for the next check. If there's a non-matching block then we've found a difference. I'll step through this process. Hash the first block. Does this hash exist in the values list? abcde -> 495 (yes, so skip) Hash the second block. Does this hash exist in the values list? 12345 -> 255 (yes, so skip) Hash the third block. Does this hash exist in the values list? 12345 -> 255 (yes, so skip) Hash the fourth block. Does this hash exist in the values list? fghij -> 520 (yes, so skip) No more data, we're done. Since every hash was found in the values list, we conclude that A and B are the same. Which, in my humble opinion, isn't true. It seems to me this will happen whenever there is more than one block that share the same hash. What am I missing?

    Read the article

  • What is the best practice for adding persistence to an MVC model?

    - by etheros
    I'm in the process of implementing an ultra-light MVC framework in PHP. It seems to be a common opinion that the loading of data from a database, file etc. should be independent of the Model, and I agree. What I'm unsure of is the best way to link this "data layer" into MVC. Datastore interacts with Model //controller public function update() { $model = $this->loadModel('foo'); $data = $this->loadDataStore('foo', $model); $data->loadBar(9); //loads data and populates Model $model->setBar('bar'); $data->save(); //reads data from Model and saves } Controller mediates between Model and Datastore Seems a bit verbose and requires the model to know that a datastore exists. //controller public function update() { $model = $this->loadModel('foo'); $data = $this->loadDataStore('foo'); $model->setDataStore($data); $model->getDataStore->loadBar(9); //loads data and populates Model $model->setBar('bar'); $model->getDataStore->save(); //reads data from Model and saves } Datastore extends Model What happens if we want to save a Model extending a database datastore to a flatfile datastore? //controller public function update() { $model = $this->loadHybrid('foo'); //get_class == Datastore_Database $model->loadBar(9); //loads data and populates $model->setBar('bar'); $model->save(); //saves } Model extends datastore This allows for Model portability, but it seems wrong to extend like this. Further, the datastore cannot make use of any of the Model's methods. //controller extends model public function update() { $model = $this->loadHybrid('foo'); //get_class == Model $model->loadBar(9); //loads data and populates $model->setBar('bar'); $model->save(); //saves } EDIT: Model communicates with DAO //model public function __construct($dao) { $this->dao = $dao; } //model public function setBar($bar) { //a bunch of business logic goes here $this->dao->setBar($bar); } //controller public function update() { $model = $this->loadModel('foo'); $model->setBar('baz'); $model->save(); } Any input on the "best" option - or alternative - is most appreciated.

    Read the article

  • Selecting one row when working with typed datasets.

    - by Wodzu
    I have a typed dataset in my project. I would like to populate a datatable with only one row instead of all rows. The selected row must be based on the primary key column. I know I could modify the designer code to achive this functionality however if I change the code in the designer I risk that this code will be deleted when I update my datased via designer in the future. So I wanted to alter the SelectCommand not in the designer but just before firing up MyTypedTableAdapter.Fill method. The strange thing is that the designer does not create a SelectCommand! It creates all other commands but not this one. If it would create SelectCommand I could alter it in this way: this.operatorzyTableAdapter.Adapter.SelectCommand.CommandText += " WHERE MyColumn = 1"; It is far from perfection but atleast I would not have to modify the designer's work. unfortunately as I said earlier the SelectCommand is not created. Instead designer creates something like this: [global::System.Diagnostics.DebuggerNonUserCodeAttribute()] private void InitCommandCollection() { this._commandCollection = new global::System.Data.SqlClient.SqlCommand[1]; this._commandCollection[0] = new global::System.Data.SqlClient.SqlCommand(); this._commandCollection[0].Connection = this.Connection; this._commandCollection[0].CommandText = "SELECT Ope_OpeID, Ope_Kod, Ope_Haslo, Ope_Imie, Ope_Nazwisko FROM dbo.Operatorzy"; this._commandCollection[0].CommandType = global::System.Data.CommandType.Text; } It doesn't make sense in my opinion. Why to create UpdateCommand, InsertCommand and DeleteCommand but do not create SelectCommand? I could bear with this but this._commandCollection is private so I cannot acces it outside of the class code. I don't know how to get into this collection without changing the designer's code. The idea which I have is to expose the collection via partial class definition. However I want to introduce many typed datasets and I really don't want to create partial class definition for each of them. Please note that I am using .NET 3.5. I've found this article about accessing private properties but it concerns .NET 4.0 Thanks for your time.

    Read the article

  • Why is the W3C box model considered better?

    - by Mel
    Why do most developers consider the W3C box-model to be better than the box-model used by Internet Explorer? I know it's very frustrating developing pages that look the way you want them on Internet Explorer, but I find the W3C box-model to be counter-intuitive. For example, if margins, padding, and border were factored into the width, I could assign fixed width values for all my columns without having to worry about how many columns I have and what changes I make to my padding and margins. Under W3C box model I have to worry about how many columns I have, and develop something akin to a mathematical formula to calculate my values. Changing them would nightmarish, especially for complex layouts. My novice opinion is that it's more restrictive. Consider this small frame-work I wrote: #content { margin:0 auto 30px auto; padding:0 30px 30px 30px; width:900px; } #content .column { float:left; margin:0 20px 20px 20px; } #content .first { margin-left:0; } #content .last { margin-right:0; } .width_1-4 { width:195px; } .width_1-3 { width:273px; } .width_1-2 { width:430px; } .width_3-4 { width:645px; } .width_1-1 { width:900px; } These values I have assigned here will falter unless there are three columns, and thus the margins at 0(first)+20+20+20+20+0(last). It would be a disaster if I wanted to add padding to my columns too, as my entire setup would have to be re calibrated. Now imagine if column width incorporated all the other elements, all I would need to do is change one associated value, and I have my layout. I'm less criticizing it and more hoping to understand why it's better, or why I'm finding it more difficult. Am I doing this whole thing wrong? I don't know very much about this topic, but it seems counter intuitive to use W3C's box-model. Some advice would be really appreciated. Thanks

    Read the article

  • What would you suggest as a high school first language?

    - by ldigas
    Edit by OA: After reading some answers I'll just update the question a little. At first I put it a little bluntly, but some of those gave me some good arguments which have to be taken into consideration while making a stand on this one. (these are mostly picked up from comments and answers below). A few things to take into account: to many pupils this is a first programming language - at this stage most of them have trouble grasping a difference between data types, variable passing, ... and whatnot, less alone pointers and similar 'low level stuff' :) they will all have to pass this to get into next grade (well, big majority of them anyway) not all of them have computers at home, not all of them are willing to learn this, less alone interested in - so the concepts have to be taught on a finite time scale in school hours (as well as practice on computers) free literature is a bonus - the teacher will make some scripts and handaways, but still ... I wouldn't like to bear the parents with the burden of buying expensive literature (also, english is not a native language here ... and although they are all learning it, their ability to read it fluently is somewhat questionable) somebody gave an argument - "a language which does not get in the way of ideas" - good one accessibility on different platforms in not expecially important at this point - although most of the suggested ones are available on win as well as linux - not many macs in this part of europe (their prices are sky high for anything but specialised usage) I will check what are the licencing issues on ms express editions about using it massively in high schools for purposes like this - if someone has any info about this, please, do not be shy with it :) A friend of mine, informatics teacher - in EU it comes as something as junior cs teacher, in a local high school asked me what I thought about what should be the first language pupils should be taught? It is a technical school (a little more oriented towards mathematics than the gymnasium, but not computer oriented totally). So I'm asking you - what do you think should be the first language pupils are exposed to in highschool? They have been teaching Pascal so far, but she's not sure that's a good course. She thought about switching to C (which I resented; considering not all pupils have interests in programming, to start with, and should be taught something higher level since they are just gripping the idea of a loop and such ... for a start), I suggested python or ruby (preferably py since it handles all paradigms). What is your opinion on this one? I looked, but didn't find a similar question on SO, so if there is one, please just point me towards it. Edit: The assumption is that none of the pupils have been exposed to any programming in junior school. See also: What is the best way to teach young kids some basic programming concepts? Best ways to teach a beginner to program How and when do you teach a kid to code What is the easiest language to start with? High School Programming

    Read the article

  • approximating log10[x^k0 + k1]

    - by Yale Zhang
    Greetings. I'm trying to approximate the function Log10[x^k0 + k1], where .21 < k0 < 21, 0 < k1 < ~2000, and x is integer < 2^14. k0 & k1 are constant. For practical purposes, you can assume k0 = 2.12, k1 = 2660. The desired accuracy is 5*10^-4 relative error. This function is virtually identical to Log[x], except near 0, where it differs a lot. I already have came up with a SIMD implementation that is ~1.15x faster than a simple lookup table, but would like to improve it if possible, which I think is very hard due to lack of efficient instructions. My SIMD implementation uses 16bit fixed point arithmetic to evaluate a 3rd degree polynomial (I use least squares fit). The polynomial uses different coefficients for different input ranges. There are 8 ranges, and range i spans (64)2^i to (64)2^(i + 1). The rational behind this is the derivatives of Log[x] drop rapidly with x, meaning a polynomial will fit it more accurately since polynomials are an exact fit for functions that have a derivative of 0 beyond a certain order. SIMD table lookups are done very efficiently with a single _mm_shuffle_epi8(). I use SSE's float to int conversion to get the exponent and significand used for the fixed point approximation. I also software pipelined the loop to get ~1.25x speedup, so further code optimizations are probably unlikely. What I'm asking is if there's a more efficient approximation at a higher level? For example: Can this function be decomposed into functions with a limited domain like log2((2^x) * significand) = x + log2(significand) hence eliminating the need to deal with different ranges (table lookups). The main problem I think is adding the k1 term kills all those nice log properties that we know and love, making it not possible. Or is it? Iterative method? don't think so because the Newton method for log[x] is already a complicated expression Exploiting locality of neighboring pixels? - if the range of the 8 inputs fall in the same approximation range, then I can look up a single coefficient, instead of looking up separate coefficients for each element. Thus, I can use this as a fast common case, and use a slower, general code path when it isn't. But for my data, the range needs to be ~2000 before this property hold 70% of the time, which doesn't seem to make this method competitive. Please, give me some opinion, especially if you're an applied mathematician, even if you say it can't be done. Thanks.

    Read the article

  • How to Work Around Limitations in Generic Type Constraints in C#?

    - by Jose
    Okay I'm looking for some input, I'm pretty sure this is not currently supported in .NET 3.5 but here goes. I want to require a generic type passed into my class to have a constructor like this: new(IDictionary<string,object>) so the class would look like this public MyClass<T> where T : new(IDictionary<string,object>) { T CreateObject(IDictionary<string,object> values) { return new T(values); } } But the compiler doesn't support this, it doesn't really know what I'm asking. Some of you might ask, why do you want to do this? Well I'm working on a pet project of an ORM so I get values from the DB and then create the object and load the values. I thought it would be cleaner to allow the object just create itself with the values I give it. As far as I can tell I have two options: 1) Use reflection(which I'm trying to avoid) to grab the PropertyInfo[] array and then use that to load the values. 2) require T to support an interface like so: public interface ILoadValues { void LoadValues(IDictionary values); } and then do this public MyClass<T> where T:new(),ILoadValues { T CreateObject(IDictionary<string,object> values) { T obj = new T(); obj.LoadValues(values); return obj; } } The problem I have with the interface I guess is philosophical, I don't really want to expose a public method for people to load the values. Using the constructor the idea was that if I had an object like this namespace DataSource.Data { public class User { protected internal User(IDictionary<string,object> values) { //Initialize } } } As long as the MyClass<T> was in the same assembly the constructor would be available. I personally think that the Type constraint in my opinion should ask (Do I have access to this constructor? I do, great!) Anyways any input is welcome.

    Read the article

  • Generic Type constraint in .net

    - by Jose
    Okay I'm looking for some input, I'm pretty sure this is not currently supported in .NET 3.5 but here goes. I want to require a generic type passed into my class to have a constructor like this: new(IDictionary<string,object>) so the class would look like this public MyClass<T> where T : new(IDictionary<string,object>) { T CreateObject(IDictionary<string,object> values) { return new T(values); } } But the compiler doesn't support this, it doesn't really know what I'm asking. Some of you might ask, why do you want to do this? Well I'm working on a pet project of an ORM so I get values from the DB and then create the object and load the values. I thought it would be cleaner to allow the object just create itself with the values I give it. As far as I can tell I have two options: 1) Use reflection(which I'm trying to avoid) to grab the PropertyInfo[] array and then use that to load the values. 2) require T to support an interface like so: public interface ILoadValues { void LoadValues(IDictionary values); } and then do this public MyClass<T> where T:new(),ILoadValues { T CreateObject(IDictionary<string,object> values) { T obj = new T(); obj.LoadValues(values); return obj; } } The problem I have with the interface I guess is philosophical, I don't really want to expose a public method for people to load the values. Using the constructor the idea was that if I had an object like this namespace DataSource.Data { public class User { protected internal User(IDictionary<string,object> values) { //Initialize } } } As long as the MyClass<T> was in the same assembly the constructor would be available. I personally think that the Type constraint in my opinion should ask (Do I have access to this constructor? I do, great!) Anyways any input is welcome.

    Read the article

  • Best of both worlds: browser and desktop game?

    - by Ricket
    When considering a platform for a game, I've decided on multi-platform (Win/Lin/Mac) but can't make up my mind as far as browser vs. desktop. As I'm not all too far in development, and now having second thoughts, I'd like your opinion! Browser-based games using Java applets: market penetration is reasonably high (for version 6, it's somewhere around 60% I believe?) using JOGL, 3D performance/quality is decent; certainly good enough to render the crappy 3D graphics that I make there's the (small?) possibility of porting something to Android great for an audience of gamers who switch computers often; can sit down at any computer, load a webpage and play it also great for casual gamers or less knowledgeable gamers who are quite happy with playing games in a browser but don't want to install more things to their computer written in a high-level language which I am more familiar with than C++ - but at the same time, I would like to improve my skills with C++ as it is probably where I am headed in the game industry once I get out of school... easier update process: reload the page. Desktop games using good ol' C++ and OpenGL 100% market penetration, assuming complete cross-platform; however, that number reduces when you consider how many people will go through downloading and installing an executable compared to just browsing to a webpage and hitting "yes" to a security warning. more trouble to maintain the cross-platform; but again, for learning purposes I would embrace the challenge and the knowledge I would gain better performance all around true full screen, whereas browser games often struggle with smooth full screen graphics (especially on Linux, in my experience) can take advantage of distribution platforms such as Steam more likely to be considered a "real" game, whereas browser and Java games are often dismissed as not being real games and therefore not played by "hardcore gamers" installer can be large; don't have to worry so much about download times Is there a way to have the best of both worlds? I love Java applets, but I also really like the reasons to write a desktop game. I don't want to constantly port everything between a Java applet project and a C++ project; that would be twice the work! Unity chose to write their own web player plugin. I don't like this, because I am one of the people that will not install their web player for anything, and I don't see myself being able to convince my audience to install a browser plugin. What are my options? Are there other examples out there besides Unity, of games that have browser and desktop versions? Did I leave out anything in the pro/con lists above?

    Read the article

  • markdown to HTML with customised WMD editor

    - by spirytus
    For my application I customized slightly the way WMD behaves so when user enters empty lines, these are reflected in HTML output as <br />'s. Now I came to a point when I should store it somewhere at backend and so after going thru SO posts for a while I'm not sure what is the best way to do it. I have few options and if you could point out which their pros/cons that would be much appreciated. send to server and store as markdown rather than HTML. To me obvious advantage would be keeping exactly same formatting as user originally entered. But then how can I convert it back to HTML for display to a client? It seems very troublesome to convert it on client side as even if it would be possible what would happen if JS would be disabled? If I wanted to do it on the server, then standard server side implementations of markup to HTML might be resource expensive. Would that be an issue in your opinion? Even if it wouldn't be the case then as I mentioned my WMD implementation is customised and those server side solutions wouldn't probably do the right conversion to markdown anyway and there always would be a risk that something would convert wrong. Send to server as converted HTML. Same as above.. conversion on client side would be difficult, server side same with possibility of getting it wrong. send original markdown and converted HTML and store both. No performance issues related to converting markdown to HTML on client side, nor on server side. Users would have always same markdown they originally entered and same HTML they originally saw in preview (possibly sanitized in php though). It would have to take twice that much storage space though and that is my biggest worry. I tend to lean towards 3rd solution as it seems simplest, but there is a worry of doubled storage space needed for this solution. Please bear in mind that my implementation of WMD is slightly modified and also I'm going with PHP/MySql server side implementation. So apart from 3 options I listed above, are there any other possible solutions to my problem? Did I miss anything important that would make one of the options above better then the rest? And what other pros/cons would apply to each solution I listed? Also how is it implemented on SO? I read somwhere that they using option 3, and so if its good enough for SO would be good enough for me :) but not sure if its true anyway, so how is it done? Also please forgive me, but at least for once I got to say that StackOverflow IS THE BEST DAMN RESOURCE ON THE WEB and I truly appreciate all the people trying to help others here! The site and users here are simply amazing!

    Read the article

  • What is fastest way to convert bool to byte?

    - by Amir Rezaei
    What is fastest way to convert bool to byte? I want this mapping: False=0, True=1 Note: I don't want to use any if statement. Update: I don't want to use conditional statement. I don't want the CPU to halt or guess next statement. I want to optimize this code: private static string ByteArrayToHex(byte[] barray) { char[] c = new char[barray.Length * 2]; byte k; for (int i = 0; i < barray.Length; ++i) { k = ((byte)(barray[i] >> 4)); c[i * 2] = (char)(k > 9 ? k + 0x37 : k + 0x30); k = ((byte)(barray[i] & 0xF)); c[i * 2 + 1] = (char)(k > 9 ? k + 0x37 : k + 0x30); } return new string(c); } Update: The length of the array is very large, it's in terabyte order! Therefore I need to do optimization if possible. I shouldn't need to explain my self. The question is still valid. Update: I'm working on a project and looking at others code. That's why I didn't provide with the function at first place. I didn't want to spend time on explaining for people when they have opinion about the code. I shouldn’y need to provide in my question the background of my work, and a function that is not written by me. I have started to optimize it part by part. If I needed help with the whole function I would asked that in another question. That is why I asked this very simple at the beginning. Unfortunately people couldn’t keep themselves to the question. So please if you want to help answer the question. Update: For dose who want to see the point of this question. This example shows how two if statement are reduced from the code. byte A = k > 9 ; //If it was possible (k>9) == 0 || 1 c[i * 2] = A * (k + 0x30) - (A - 1) * (k + 0x30);

    Read the article

  • Which style is preferable when writing this boolean expression?

    - by Jeppe Stig Nielsen
    I know this question is to some degree a matter of taste. I admit this is not something I don't understand, it's just something I want to hear others' opinion about. I need to write a method that takes two arguments, a boolean and a string. The boolean is in a sense (which will be obvious shortly) redundant, but it is part of a specification that the method must take in both arguments, and must raise an exception with a specific message text if the boolean has the "wrong" value. The bool must be true if and only if the string is not null or empty. So here are some different styles to write (hopefully!) the same thing. Which one do you find is the most readable, and compliant with good coding practice? // option A: Use two if, repeat throw statement and duplication of message string public void SomeMethod(bool useName, string name) { if (useName && string.IsNullOrEmpty(name)) throw new SomeException("..."); if (!useName && !string.IsNullOrEmpty(name)) throw new SomeException("..."); // rest of method } // option B: Long expression but using only && and || public void SomeMethod(bool useName, string name) { if (useName && string.IsNullOrEmpty(name) || !useName && !string.IsNullOrEmpty(name)) throw new SomeException("..."); // rest of method } // option C: With == operator between booleans public void SomeMethod(bool useName, string name) { if (useName == string.IsNullOrEmpty(name)) throw new SomeException("..."); // rest of method } // option D1: With XOR operator public void SomeMethod(bool useName, string name) { if (!(useName ^ string.IsNullOrEmpty(name))) throw new SomeException("..."); // rest of method } // option D2: With XOR operator public void SomeMethod(bool useName, string name) { if (useName ^ !string.IsNullOrEmpty(name)) throw new SomeException("..."); // rest of method } Of course you're welcome to suggest other possibilities too. Message text "..." would be something like "If 'useName' is true a name must be given, and if 'useName' is false no name is allowed".

    Read the article

  • itertools.product eliminating repeated reversed tuples

    - by genclik27
    I asked a question yesterday and thanks to Tim Peters, it is solved. The question is here; itertools.product eliminating repeated elements The new question is further version of this. This time I will generate tuples inside of tuples. Here is an example; lis = [[(1,2), (3,4)], [(5,2), (1,2)], [(2,1), (1,2)]] When I use it in itertools.product function this is what I get, ((1, 2), (5, 2), (2, 1)) ((1, 2), (5, 2), (1, 2)) ((1, 2), (1, 2), (2, 1)) ((1, 2), (1, 2), (1, 2)) ((3, 4), (5, 2), (2, 1)) ((3, 4), (5, 2), (1, 2)) ((3, 4), (1, 2), (2, 1)) ((3, 4), (1, 2), (1, 2)) I want to change it in a way that if a sequence has (a,b) inside of it, then it can not have (b,a). In this example if you look at this sequence ((3, 4), (1, 2), (2, 1)) it has (1,2) and (2,1) inside of it. So, this sequence ((3, 4), (1, 2), (2, 1)) should not be considered in the results. As I said, I asked similar question before, in that case it was not considering duplicate elements. I try to adapt it to my problem. Here is modified code. Changed parts in old version are taken in comments. def reverse_seq(seq): s = [] for i in range(len(seq)): s.append(seq[-i-1]) return tuple(s) def uprod(*seqs): def inner(i): if i == n: yield tuple(result) return for elt in sets[i] - reverse: #seen.add(elt) rvrs = reverse_seq(elt) reverse.add(rvrs) result[i] = elt for t in inner(i+1): yield t #seen.remove(elt) reverse.remove(rvrs) sets = [set(seq) for seq in seqs] n = len(sets) #seen = set() reverse = set() result = [None] * n for t in inner(0): yield t In my opinion this code should work but I am getting error for the input lis = [[(1,2), (3,4)], [(5,2), (1,2)], [(2,1), (1,2)]]. I could not understand where I am wrong. for i in uprod(*lis): print i Output is, ((1, 2), (1, 2), (1, 2)) Traceback (most recent call last): File "D:\Users\SUUSER\workspace tree\sequence_covering _array\denemeler_buraya.py", line 39, in <module> for i in uprod(*lis): File "D:\Users\SUUSER\workspace tree\sequence_covering _array\denemeler_buraya.py", line 32, in uprod for t in inner(0): File "D:\Users\SUUSER\workspace tree\sequence_covering _array\denemeler_buraya.py", line 22, in inner for t in inner(i+1): File "D:\Users\SUUSER\workspace tree\sequence_covering _array\denemeler_buraya.py", line 25, in inner reverse.remove(rvrs) KeyError: (2, 1) Thanks,

    Read the article

  • Creating multiple heads in remote repository

    - by Jab
    We are looking to move our team (~10 developers) from SVN to mercurial. We are trying to figure out how to manage our workflow. In particular, we are trying to see if creating remote heads is the right solution. We currently have a very large repository with multiple, related projects. They share a lot of code, but pieces of the project are deployed by different teams (3 teams) independent of other portions of the code-base. So each team is working on concurrent large features. The way we currently handles this in SVN are branches. Team1 has a branch for Feature1, same deal for the other teams. When Team1 finishes their change, it gets merged into the trunk and deployed out. The other teams follow suite when their project is complete, merging of course. So my initial thought are using Named Branches for these situations. Team1 makes a Feature1 branch off of the default branch in Hg. Now, here is the question. Should the team PUSH that branch, in it's current/half-state to the repository. This will create a second head in the core repo. My initial reaction was "NO!" as it seems like a bad idea. Handling multiple heads on our repository just sounds awful, but there are some advantages... First, the teams want to setup Continuous Integration to build this branch during their development cycle(months long). This will only work if the CI can pull this branch from the repo. This is something we do now with SVN, copy a CI build and change the branch. Easy. Second, it makes it easier for any team member to jump onto the branch and start working. Without pushing to the core repo, they would have to receive a push from a developer on that team with the changeset information. It is also possible to lose local commits to hardware failure. The chances increase a lot if it's a branch by a single developer who has followed the "don't push until finished" approach. And lastly is just for ease of use. The developers can easily just commit and push on their branch at any time without consequence(as they do today, in their SVN branches). Is there a better way to handle this scenario that I may be missing? I just want a veteran's opinion before moving forward with the strategy. For bug fixes we like the general workflow of mecurial, anonymous branches that only consist of 1-2 commits. The simplicity is great for those cases. By the way, I've read this , great article which seems to favor Named branches.

    Read the article

  • Thinking about introducing PHP/MySQL into a .NET/SQL Server environment. Thoughts?

    - by abszero
    I posted this over at reddit but it didn't gain any momentum. So here is what is going on: our company was recently purchased by another web shop and I was promoted to head of development here in our office. Our office is completely .NET/SQL Server and the company who purchased us is a *nix/PHP/MySQL shop. Now several of our large clients who are on the .NET platform are up for complete rewrites (the sites are from '04 and are running on the 1.x framework.) While reviewing the proposal for one client with my superior I came across a pretty extensive module which would require several hundred man hours to complete and voiced some concern about it in relation to the quote. One of the guys from the PHP group happen to hear this and told me of a module that they (PHP Group) use in Drupal that does exactly what the proposal in front of me was describing and it only took, at most, 8 hours to completely setup / configure. My superior suggested that I take a look at Drupal and the module in question over the weekend but stressed that we should only go that route if it really made sense. So this weekend I spun up a CentOS instance in VirtualBox and started playing around with Drupal. I am still fleshing it out so don't have a solid opinion on it just yet. Anyway I have some questions / fears that I was hoping progit could help me out in! Has anyone had experience doing this and, if so, how did it turn out? I am completely ignorant to what IDE's (if any) are available to for PHP. The last time I worked with PHP it was in Notepad and that was less than intuitive. So is there are more intuitive IDE out there for PHP dev? I don't want to scare my .NET guys. Since the merger all of our new business clients that have had relatively small websites have gone on Drupal with the larger sites going on .NET. My concern is that if they see a large site go onto Drupal that they might start getting anxious and start handing out their resumes. For the foreseeable future there are no plans to liquidate the .NET platform and really we can't just from a support standpoint. What would be the best way to approach this? Any other helpful info? Thanks!

    Read the article

  • Setting default values for inherited property without using accessor in Objective-C?

    - by Ben Stock
    I always see people debating whether or not to use a property's setter in the -init method. I don't know enough about the Objective-C language yet to have an opinion one way or the other. With that said, lately I've been sticking to ivars exclusively. It seems cleaner in a way. I don't know. I digress. Anyway, here's my problem … Say we have a class called Dude with an interface that looks like this: @interface Dude : NSObject { @private NSUInteger _numberOfGirlfriends; } @property (nonatomic, assign) NSUInteger numberOfGirlfriends; @end And an implementation that looks like this: @implementation Dude - (instancetype)init { self = [super init]; if (self) { _numberOfGirlfriends = 0; } } @end Now let's say I want to extend Dude. My subclass will be called Playa. And since a playa should have mad girlfriends, when Playa gets initialized, I don't want him to start with 0; I want him to have 10. Here's Playa.m: @implementation Playa - (instancetype)init { self = [super init]; if (self) { // Attempting to set the ivar directly will result in the compiler saying, // "Instance variable `_numberOfGirlfriends` is private." // _numberOfGirlfriends = 10; <- Can't do this. // Thus, the only way to set it is with the mutator: self.numberOfGirlfriends = 10; // Or: [self setNumberOfGirlfriends:10]; } } @end So what's a Objective-C newcomer to do? Well, I mean, there's only one thing I can do, and that's set the property. Unless there's something I'm missing. Any ideas, suggestions, tips, or tricks? Sidenote: The other thing that bugs me about setting the ivar directly — and what a lot of ivar-proponents say is a "plus" — is that there are no KVC notifications. A lot of times, I want the KVC magic to happen. 50% of my setters end in [self setNeedsDisplay:YES], so without the notification, my UI doesn't update unless I remember to manually add -setNeedsDisplay. That's a bad example, but the point stands. I utilize KVC all over the place, so without notifications, things can act wonky. Anyway, any info is much appreciated. Thanks!

    Read the article

  • C++ Class Access Specifier Verbosity

    - by PolyTex
    A "traditional" C++ class (just some random declarations) might resemble the following: class Foo { public: Foo(); explicit Foo(const std::string&); ~Foo(); enum FooState { Idle, Busy, Unknown }; FooState GetState() const; bool GetBar() const; void SetBaz(int); private: struct FooPartialImpl; void HelperFunction1(); void HelperFunction2(); void HelperFunction3(); FooPartialImpl* m_impl; // smart ptr FooState m_state; bool m_bar; int m_baz; }; I always found this type of access level specification ugly and difficult to follow if the original programmer didn't organize his "access regions" neatly. Taking a look at the same snippet in a Java/C# style, we get: class Foo { public: Foo(); public: explicit Foo(const std::string&); public: ~Foo(); public: enum FooState { Idle, Busy, Unknown }; public: FooState GetState() const; public: bool GetBar() const; public: void SetBaz(int); private: struct FooPartialImpl; private: void HelperFunction1(); private: void HelperFunction2(); private: void HelperFunction3(); private: FooPartialImpl* m_impl; // smart ptr private: FooState m_state; private: bool m_bar; private: int m_baz; }; In my opinion, this is much easier to read in a header because the access specifier is right next to the target, and not a bunch of lines away. I found this especially true when working with header-only template code that wasn't separated into the usual "*.hpp/*.inl" pair. In that scenario, the size of the function implementations overpowered this small but important information. My question is simple and stems from the fact that I've never seen anyone else actively do this in their C++ code. Assuming that I don't have a "Class View" capable IDE, are there any obvious drawbacks to using this level of verbosity? Any other style recommendations are welcome!

    Read the article

  • Where do you put your unit test?

    - by soulmerge
    I have found several conventions to housekeeping unit tests in a project and I'm not sure which approach would be suitable for our next PHP project. I am trying to find the best convention to encourage easy development and accessibility of the tests when reviewing the source code. I would be very interested in your experience/opinion regarding each: One folder for productive code, another for unit tests: This separates unit tests from the logic files of the project. This separation of concerns is as much a nuisance as it is an advantage: Someone looking into the source code of the project will - so I suppose - either browse the implementation or the unit tests (or more commonly: the implementation only). The advantage of unit tests being another viewpoint to your classes is lost - those two viewpoints are just too far apart IMO. Annotated test methods: Any modern unit testing framework I know allows developers to create dedicated test methods, annotating them (@test) and embedding them in the project code. The big drawback I see here is that the project files get cluttered. Even if these methods are separated using a comment header (like UNIT TESTS below this line) it just bloats the class unnecessarily. Test files within the same folders as the implementation files: Our file naming convention dictates that PHP files containing classes (one class per file) should end with .class.php. I could imagine that putting unit tests regarding a class file into another one ending on .test.php would render the tests much more present to other developers without tainting the class. Although it bloats the project folders, instead of the implementation files, this is my favorite so far, but I have my doubts: I would think others have come up with this already, and discarded this option for some reason (i.e. I have not seen a java project with the files Foo.java and FooTest.java within the same folder.) Maybe it's because java developers make heavier use of IDEs that allow them easier access to the tests, whereas in PHP no big editors have emerged (like eclipse for java) - many devs I know use vim/emacs or similar editors with little support for PHP development per se. What is your experience with any of these unit test placements? Do you have another convention I haven't listed here? Or am I just overrating unit test accessibility to reviewers?

    Read the article

  • Model Binding with Parent/Child Relationship

    - by user296297
    I'm sure this has been answered before, but I've spent the last three hours looking for an acceptable solution and have been unable to find anything, so I apologize for what I'm sure is a repeat. I have two domain objects, Player and Position. Player's have a Position. My domain objects are POCOs tied to my database with NHibernate. I have an Add action that takes a Player, so I'm using the built in model binding. On my view I have a drop down list that lets a user select the Position for the Player. The value of the drop down list is the Id of the position. Everything gets populated correctly except that my Position object fails validation (ModelState.IsValid) because at the point of model binding it only has an Id and none of it's other required attributes. What is the preferred solution for solving this with ASP.NET MVC 2? Solutions I've tried... Fetch the Position from the database based on the Id before ModelState.IsValid is called in the Add action of my controller. I can't get the model to run the validation again, so ModelState.IsValid always returns false. Create a custom ModelBinder that inherits from the default binder and fetch the Position from the database after the base binder is called. The ModelBinder seems to be doing the validation so if I use anything from the default binder I'm hosed. Which means I have to completely roll my own binder and grab every value from the form...this seems really wrong and inefficient for such a common use-case. Solutions I think might work, I just can't figure out how to do... Turn off the validation for the Position class when used in Player. Write a custom ModelBinder leverages the default binder for most of the property binding, but lets me get the Position from the database BEFORE the default binder runs validation. So, how do the rest of you solve this? Thanks, Dan P.S. In my opinion having a PositionId on Player just for this case is not a good solution. There has to be solvable in a more elegant fashion.

    Read the article

  • Hidden divs for "lazy javascript" loading? Possible security/other issues?

    - by xyld
    I'm curious about people's opinion's and thoughts about this situation. The reason I'd like to lazy load javascript is because of performance. Loading javascript at the end of the body reduces the browser blocking and ends up with much faster page loads. But there is some automation I'm using to generate the html (django specifically). This automation has the convenience of allowing forms to be built with "Widgets" that output content it needs to render the entire widget (extra javascript, css, ...). The problem is that the widget wants to output javascript immediately into the middle of the document, but I want to ensure all javascript loads at the end of the body. When the following widget is added to a form, you can see it renders some <script>...</script> tags: class AutoCompleteTagInput(forms.TextInput): class Media: css = { 'all': ('css/jquery.autocomplete.css', ) } js = ( 'js/jquery.bgiframe.js', 'js/jquery.ajaxQueue.js', 'js/jquery.autocomplete.js', ) def render(self, name, value, attrs=None): output = super(AutoCompleteTagInput, self).render(name, value, attrs) page_tags = Tag.objects.usage_for_model(DataSet) tag_list = simplejson.dumps([tag.name for tag in page_tags], ensure_ascii=False) return mark_safe(u'''<script type="text/javascript"> jQuery("#id_%s").autocomplete(%s, { width: 150, max: 10, highlight: false, scroll: true, scrollHeight: 100, matchContains: true, autoFill: true }); </script>''' % (name, tag_list,)) + output What I'm proposing is that if someone uses a <div class=".lazy-js">...</div> with some css (.lazy-js { display: none; }) and some javascript (jQuery('.lazy-js').each(function(index) { eval(jQuery(this).text()); }), you can effectively force all javascript to load at the end of page load: class AutoCompleteTagInput(forms.TextInput): class Media: css = { 'all': ('css/jquery.autocomplete.css', ) } js = ( 'js/jquery.bgiframe.js', 'js/jquery.ajaxQueue.js', 'js/jquery.autocomplete.js', ) def render(self, name, value, attrs=None): output = super(AutoCompleteTagInput, self).render(name, value, attrs) page_tags = Tag.objects.usage_for_model(DataSet) tag_list = simplejson.dumps([tag.name for tag in page_tags], ensure_ascii=False) return mark_safe(u'''<div class="lazy-js"> jQuery("#id_%s").autocomplete(%s, { width: 150, max: 10, highlight: false, scroll: true, scrollHeight: 100, matchContains: true, autoFill: true }); </div>''' % (name, tag_list,)) + output Nevermind all the details of my specific implementation (the specific media involved), I'm looking for a consensus on whether the method of using lazy-loaded javascript through hidden a hidden tags can pose issues whether security or other related? One of the most convenient parts about this is that it follows the DRY principle rather well IMO because you don't need to hack up a specific lazy-load for each instance in the page. It just "works". UPDATE: I'm not sure if django has the ability to queue things (via fancy template inheritance or something?) to be output just before the end of the </body>?

    Read the article

< Previous Page | 63 64 65 66 67 68 69 70 71 72 73  | Next Page >