Search Results

Search found 6436 results on 258 pages for 'el quick'.

Page 234/258 | < Previous Page | 230 231 232 233 234 235 236 237 238 239 240 241  | Next Page >

  • Need help with implementation of the jQuery LiveUpdate routine

    - by miCRoSCoPiC_eaRthLinG
    Hey all, Has anyone worked with the LiveUpdate function (may be a bit of a misnomer) found on this page? It's not really a live search/update function, but a quick filtering mechanism for a pre-existing list, based on the pattern you enter in a text field. For easier reference, I'm pasting the entire function in here: jQuery.fn.liveUpdate = function(list){ list = jQuery(list); if ( list.length ) { var rows = list.children('li'), cache = rows.map(function(){ return this.innerHTML.toLowerCase(); }); this .keyup(filter).keyup() .parents('form').submit(function(){ return false; }); } return this; function filter(){ var term = jQuery.trim( jQuery(this).val().toLowerCase() ), scores = []; if ( !term ) { rows.show(); } else { rows.hide(); cache.each(function(i){ var score = this.score(term); if (score > 0) { scores.push([score, i]); } }); jQuery.each(scores.sort(function(a, b){return b[0] - a[0];}), function(){ jQuery(rows[ this[1] ]).show(); }); } } }; I have this list, with members as the ID. And a text field with say, qs as ID. I tried binding the function in the following manner: $( '#qs' ).liveUpdate( '#members' ); But when I do this, the function is called only ONCE when the page is loaded (I put in some console.logs in the function) but never after when text is keyed into the text field. I also tried calling the routine from the keyup() function of qs. $( '#qs' ).keyup( function() { $( this ).liveUpdate( '#members' ); }); This ends up going into infinite loops (almost) and halting with "Too much recursion" errors. So can anyone please shed some light on how I am supposed to actually implement this function? Also while you are at it, can someone kindly explain this line to me: var score = this.score(term); What I want to know is where this member method score() is coming from? I didn't find any such method built into JS or jQuery. Thanks for all the help, m^e

    Read the article

  • Alternatives to requiring users to register for an account?

    - by jamieb
    I'm working on a side project to build a new web app idea of mine. For the sake of discussion, let's say this app displays a random photograph of a famous work of art. On a scale of 1 to 5, users are asked to rate how well they like each piece of art, and then are shown the next photo. Eventually, the app is able to get an sense of the person's style and is able to recommend artwork that he/she may find pleasing. The whole concept is similar to Netflix. I understand how to do all the preference matching logic (although not as sophisticated as Netflix). But I'd like to find a way to do this without requiring that users create an account first. This is a novelty website that a typical user might use only a handful of times. Requiring registration is overkill and will likely drastically reduce it's utility. I'd like to allow people to begin rating artwork within five seconds of their initial pageview, yet maintain the integrity of the voting (since recommendations are predicated on how other people have rated the various pieces of artwork). Can it be done? Some ideas: OpenID. The perfect solution except for the fact that it's not wildly used and my target audience isn't the most technically adept demographic. Text message. User inputs phone number and is texted a four digit code to key into the web app. Quick, easy, and great way to limit abuse. However, privacy concerns abound... people are probably even less likely to give me their phone number than their email address. Facebook login. I personally don't have a Facebook account due to privacy concerns. And I'd really hate to support such a proprietary platform. Hash code/Bookmark. Vistor's initial pageview generates a 5 or 6 digit alphanumeric code that is embedded in each subsequent URL. They can bookmark any page to save their state. Good: Very simple system that doesn't require any user action. Bad: Very easy to stuff the ballot box, might be difficult to account for users sharing the link containing their ID code via email or social networking sites.

    Read the article

  • Marshal.PtrToStructure (and back again) and generic solution for endianness swapping

    - by cgyDeveloper
    I have a system where a remote agent sends serialized structures (from and embedded C system) for me to read and store via IP/UDP. In some cases I need to send back the same structure types. I thought I had a nice setup using Marshal.PtrToStructure (receive) and Marshal.StructureToPtr (send). However, a small gotcha is that the network big endian integers need to be converted to my x86 little endian format to be used locally. When I'm sending them off again, big endian is the way to go. Here are the functions in question: private static T BytesToStruct<T>(ref byte[] rawData) where T: struct { T result = default(T); GCHandle handle = GCHandle.Alloc(rawData, GCHandleType.Pinned); try { IntPtr rawDataPtr = handle.AddrOfPinnedObject(); result = (T)Marshal.PtrToStructure(rawDataPtr, typeof(T)); } finally { handle.Free(); } return result; } private static byte[] StructToBytes<T>(T data) where T: struct { byte[] rawData = new byte[Marshal.SizeOf(data)]; GCHandle handle = GCHandle.Alloc(rawData, GCHandleType.Pinned); try { IntPtr rawDataPtr = handle.AddrOfPinnedObject(); Marshal.StructureToPtr(data, rawDataPtr, false); } finally { handle.Free(); } return rawData; } And a quick example structure that might be used like this: byte[] data = this.sock.Receive(ref this.ipep); Request request = BytesToStruct<Request>(ref data); Where the structure in question looks like: [StructLayout(LayoutKind.Sequential, CharSet = CharSet.Ansi, Pack = 1)] private struct Request { public byte type; public short sequence; [MarshalAs(UnmanagedType.ByValArray, SizeConst = 5)] public byte[] address; } What (generic) way can I swap the endianness when marshalling the structures? My need is such that the locally stored 'public short sequence' in this example will be little-endian for displaying to the user. I don't want to have to swap the endianness on a structure-specific way. My first thought was to use Reflection, but I'm not very familiar with that feature. Also, I hoped that there would be a better solution out there that somebody could point me towards. Thanks in advance :)

    Read the article

  • Team matchups for Dota Bot

    - by Dan
    I have a ghost++ bot that hosts games of Dota (a warcraft 3 map that is played 5 players versus 5 players) and I'm trying to come up with good formulas to balance the players going into a match based on their records (I have game history for several thousand games). I'm familear with some of the concepts required to match up players, like confidence based on sample size of the number of games they played, and also perameter approximation and degrees of freedom and thus throwing out any variables that don't contribute enough to the r^2. My bot collects quite a few variables for each player from each game: The Important ones: Win/Lose/Game did not finish # of Player Kills # of Player Deaths # of Kills player assisted The not so important ones: # of enemy creep kills # of creep sneak attacks # of neutral creep kills # of Tower kills # of Rax kills # of courier kills Quick explination: The kills/deaths don't determine who wins, but the gold gained and lost from this usually is enough to tilt the game. Tower/Rax kills are what the goal of the game is (once a team looses all their towers/rax their thrown can be attacked if that is destroyed they lose), but I don't really count these as important because it is pretty random who gets the credit for the tower kill, and chances are if you destroy a tower it is only because some other player is doing well and distracting the otherteam elsewhere on the map. I'm getting a bit confused when trying to deal with the fact that 5 players are on a team, so ultimately each individual isn't that responsible for the team winner or losing. Take a player that is really good at killing and has 40 kills and only 10 deaths, but in their 5 games they've only won 1. Should I give him extra credit for such a high kill score despite losing? (When losing it is hard to keep a positive kill/death ratio) Or should I dock him for losing assuming that despite the nice kill/death ratio he probably plays in a really greedy way only looking out for himself and not helping the team? Ultimately I don't think I have to guess at questions like this because I have so much data... but I don't really know how to look at the data to answer questions like this. Can anyone help me come up with formulas to help team balance and predict the outcome? Thanks, Dan

    Read the article

  • How to merge a "branch" that isn't really a branch (wasn't created by an svn copy)

    - by MatrixFrog
    I'm working on a team with lots of people who are pretty unfamiliar with the concepts of version control systems, and are just kind of doing whatever seems to work, by trial and error. Someone created a "branch" from the trunk that is not ancestrally related to the trunk. My guess is it went something like this: They created a folder in branches. They checked out all the code from the trunk to somewhere on their desktop. They added all that code to the newly created folder as though it was a bunch of brand new files. So the repository isn't aware that all that code is actually just a copy of the trunk. When I look at the history of that branch in TortoiseSVN, and uncheck the "Stop on copy/rename" box, there is no revision that has the trunk (or any other path) under the "Copy from path" column. Then they made lots of changes on their "branch". Meanwhile, others were making lots of changes on the trunk. We tried to do a merge and of course it doesn't work. Because, the trunk and the fake branch are not ancestrally related. I can see only two ways to resolve this: Go through the logs on the "branch", look at every change that was made, and manually apply each change to the trunk. Go through the logs on the trunk, look at every change that was made between revision 540 (when the "branch" was created) and HEAD, and manually apply each change to the "branch". This involves 7 revisions one way or 11 revisions the other way, so neither one is really that terrible. But is there any way to cause the repository to "realize" that the branch really IS ancestrally related even though it was created incorrectly, so that we can take advantage of the built-in merging functionality in Eclipse/TortoiseSVN? (You may be wondering: Why did your company hire these people and allow them to access the SVN repository without making sure they knew how to use it properly first?! We didn't -- this is a school assignment, which is a collaboration between two different classes -- the ones in the lower class were given a very quick hand-wavey "overview" of SVN which didn't really teach them anything. I've asked everyone in the group to please PLEASE read the svn book, and I'll make sure we (the slightly more experienced half of the team) keep a close eye on the repository to ensure this doesn't happen again.)

    Read the article

  • QT vs. Net - REAL comparisons for R.A.D. projects

    - by Pirate for Profit
    Man in all these Qt vs. .NET discussions 90% these people argue about the dumbest crap. Trying to get a real comparison chart here, because I know a little about both frameworks but I don't know everything. I believe Qt and .NET both have strengths and weaknesses. This is to make a comparison that highlights these so people can make more informed decisions before embarking on a project, in the spirit of R.A.D. Event Handling In Qt the event handling system is very simple. You just emit signals when something cool happens and then catch them in slots. ie. // run some calculations, then emit valueChanged(30, false, 20.2); and then catching it, any object can make a slot to recieve that message easily void MyObj::valueChanged(int percent, bool ok, float timeRemaining). It's easy to "block" an event or "disconnect" when needed, and works seamlessly across threads... once you get the hang of it, it just seems a lot more natural and intuitive than the way the .NET event handling is set up (you know, void valueChanged(object sender, CustomEventArgs e). And I'm not just talking about syntax, because in the end the .NET anonymous delegates are the bomb. I'm also talking about in more than just reflection (because, yes, .NET obviously has much stronger reflection capabilities). I'm talking about in the way the system feels to a human being. Qt wins hands down for the simplest yet still flexible event handling system ever i m o. Plugins and such I do love some of the ease of C# compared to C++, as well as .NET's assembly architecture, even though it leads to a bunch of .dll's (there's ways to combine everything into a single exe though). That is a big bonus for modular projects, which are a PITA to import stuff in C++ as far as RAD is concerned. Database Ease of Doing Crap Also what about datasets and database manipulations. I think .net wins here but I'm not sure. Threading/Conccurency How do you guys think of the threading? In .NET, all I've ever done is make like a list of master worker threads with locks. I like QConcurrentFramework, you don't worry about locks or anything, and with the ease of the signal slot system across threads it's nice to get notified about the progress of things. QConcurrent is the simplest threading mechanism I've ever played with. Memory Usage Also what do you think of the overall memory usage comparison. Is the .NET garbage collector pretty on the ball and quick compared to the instantaneous nature of native memory management? Or does it just let programs leak up a storm and lag the computer then clean it up when it's about to really lag? Doesn't the just-in-time compiler make native code that is pretty good, like and that only happens the first time the program is run? However, I am a n00b who doesn't know what I'm talking about, please school me on the subject.

    Read the article

  • Getting XML data from a external page and parsing it with PHP

    - by James P
    I'm trying to create a database of World of Warcraft gems. If I go to this page: http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=purple&searchType=items And go to View Source in Firefox, I see a tonne of XML data which is exactly what I want. I wrote up this quick script to try and parse some of it: <?php $gemUrls = array( 'Blue' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=blue&searchType=items', 'Red' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=red&searchType=items', 'Yellow' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=yellow&searchType=items', 'Meta' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=meta&searchType=items', 'Green' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=green&searchType=items', 'Orange' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=orange&searchType=items', 'Purple' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=purple&searchType=items', 'Prismatic' => 'http://www.wowarmory.com/search.xml?fl[source]=all&fl[type]=gems&fl[subTp]=purple&searchType=items' ); // Get blue gems $blueGems = file_get_contents($gemUrls['Blue']); $xml = new SimpleXMLElement($blueGems); echo $xml->items[0]->item; ?> But I get a load of errors like this: Warning: SimpleXMLElement::__construct() [simplexmlelement.--construct]: Entity: line 20: parser error : xmlParseEntityRef: no name in C:\xampp\htdocs\WoW\index.php on line 19 Warning: SimpleXMLElement::__construct() [simplexmlelement.--construct]: if(Browser.iphone && Number(getcookie2("mobIntPageVisits")) < 3 && getcookie2( in C:\xampp\htdocs\WoW\index.php on line 19 I'm not sure what's wrong. I think file_get_contents() is bringing back data that isn't XML, maybe some Javascript files judging by the iPhone parts in the errors. Is there any way to just get back the XML from that page? Without any HTML or anything? Thanks :)

    Read the article

  • file.createNewFile() creates files with last-modified time before actual creation time

    - by Kaleb Pederson
    I'm using JPoller to detect changes to files in a specific directory, but it's missing files because they end up with a timestamp earlier than their actual creation time. Here's how I test: public static void main(String [] files) { for (String file : files) { File f = new File(file); if (f.exists()) { System.err.println(file + " exists"); continue; } try { // find out the current time, I would hope to assume that the last-modified // time on the file will definitely be later than this System.out.println("-----------------------------------------"); long time = System.currentTimeMillis(); // create the file System.out.println("Creating " + file + " at " + time); f.createNewFile(); // let's see what the timestamp actually is (I've only seen it <time) System.out.println(file + " was last modified at: " + f.lastModified()); // well, ok, what if I explicitly set it to time? f.setLastModified(time); System.out.println("Updated modified time on " + file + " to " + time + " with actual " + f.lastModified()); } catch (IOException e) { System.err.println("Unable to create file"); } } } And here's what I get for output: ----------------------------------------- Creating test.7 at 1272324597956 test.7 was last modified at: 1272324597000 Updated modified time on test.7 to 1272324597956 with actual 1272324597000 ----------------------------------------- Creating test.8 at 1272324597957 test.8 was last modified at: 1272324597000 Updated modified time on test.8 to 1272324597957 with actual 1272324597000 ----------------------------------------- Creating test.9 at 1272324597957 test.9 was last modified at: 1272324597000 Updated modified time on test.9 to 1272324597957 with actual 1272324597000 The result is a race condition: JPoller records time of last check as xyz...123 File created at xyz...456 File last-modified timestamp actually reads xyz...000 JPoller looks for new/updated files with timestamp greater than xyz...123 JPoller ignores newly added file because xyz...000 is less than xyz...123 I pull my hair out for a while I tried digging into the code but both lastModified() and createNewFile() eventually resolve to native calls so I'm left with little information. For test.9, I lose 957 milliseconds. What kind of accuracy can I expect? Are my results going to vary by operating system or file system? Suggested workarounds? NOTE: I'm currently running Linux with an XFS filesystem. I wrote a quick program in C and the stat system call shows st_mtime as truncate(xyz...000/1000).

    Read the article

  • Prolog: Sentence Parser Problem

    - by Devon
    Hey guys, Been sat here for hours now just staring at this code and have no idea what I'm doing wrong. I know what's happening from tracing the code through (it is going on an eternal loop when it hits verbPhrase). Any tips are more then welcome. Thank you. % Knowledge-base det(the). det(a). adjective(quick). adjective(brown). adjective(orange). adjective(sweet). noun(cat). noun(mat). noun(fox). noun(cucumber). noun(saw). noun(mother). noun(father). noun(family). noun(depression). prep(on). prep(with). verb(sat). verb(nibbled). verb(ran). verb(looked). verb(is). verb(has). % Sentece Structures sentence(Phrase) :- append(NounPhrase, VerbPhrase, Phrase), nounPhrase(NounPhrase), verbPhrase(VerbPhrase). sentence(Phrase) :- verbPhrase(Phrase). nounPhrase([]). nounPhrase([Head | Tail]) :- det(Head), nounPhrase2(Tail). nounPhrase(Phrase) :- nounPhrase2(Phrase). nounPhrase(Phrase) :- append(NP, PP, Phrase), nounPhrase(NP), prepPhrase(PP). nounPhrase2([]). nounPhrase2(Word) :- noun(Word). nounPhrase2([Head | Tail]) :- adjective(Head), nounPhrase2(Tail). prepPhrase([]). prepPhrase([Head | Tail]) :- prep(Head), nounPhrase(Tail). verbPhrase([]). verbPhrase(Word) :- verb(Word). verbPhrase([Head | Tail]) :- verb(Head), nounPhrase(Tail). verbPhrase(Phrase) :- append(VP, PP, Phrase), verbPhrase(VP), prepPhrase(PP).

    Read the article

  • Any tool(s) for knowing the layout (segments) of running process in Windows?

    - by claws
    I've always been curious about How exactly the process looks in memory? What are the different segments(parts) in it? How exactly will be the program (on the disk) & process (in the memory) are related? My previous question: http://stackoverflow.com/questions/1966920/more-info-on-memory-layout-of-an-executable-program-process In my quest, I finally found a answer. I found this excellent article that cleared most of my queries: http://www.linuxforums.org/articles/understanding-elf-using-readelf-and-objdump_125.html In the above article, author shows how to get different segments of the process (LINUX) & he compares it with its corresponding ELF file. I'm quoting this section here: Courious to see the real layout of process segment? We can use /proc//maps file to reveal it. is the PID of the process we want to observe. Before we move on, we have a small problem here. Our test program runs so fast that it ends before we can even dump the related /proc entry. I use gdb to solve this. You can use another trick such as inserting sleep() before it calls return(). In a console (or a terminal emulator such as xterm) do: $ gdb test (gdb) b main Breakpoint 1 at 0x8048376 (gdb) r Breakpoint 1, 0x08048376 in main () Hold right here, open another console and find out the PID of program "test". If you want the quick way, type: $ cat /proc/`pgrep test`/maps You will see an output like below (you might get different output): [1] 0039d000-003b2000 r-xp 00000000 16:41 1080084 /lib/ld-2.3.3.so [2] 003b2000-003b3000 r--p 00014000 16:41 1080084 /lib/ld-2.3.3.so [3] 003b3000-003b4000 rw-p 00015000 16:41 1080084 /lib/ld-2.3.3.so [4] 003b6000-004cb000 r-xp 00000000 16:41 1080085 /lib/tls/libc-2.3.3.so [5] 004cb000-004cd000 r--p 00115000 16:41 1080085 /lib/tls/libc-2.3.3.so [6] 004cd000-004cf000 rw-p 00117000 16:41 1080085 /lib/tls/libc-2.3.3.so [7] 004cf000-004d1000 rw-p 004cf000 00:00 0 [8] 08048000-08049000 r-xp 00000000 16:06 66970 /tmp/test [9] 08049000-0804a000 rw-p 00000000 16:06 66970 /tmp/test [10] b7fec000-b7fed000 rw-p b7fec000 00:00 0 [11] bffeb000-c0000000 rw-p bffeb000 00:00 0 [12] ffffe000-fffff000 ---p 00000000 00:00 0 Note: I add number on each line as reference. Back to gdb, type: (gdb) q So, in total, we see 12 segment (also known as Virtual Memory Area--VMA). But I want to know about Windows Process & PE file format. Any tool(s) for getting the layout (segments) of running process in Windows? Any other good resources for learning more on this subject? EDIT: Are there any good articles which shows the mapping between PE file sections & VA segments?

    Read the article

  • How to merge an improperly created "branch" that isn't really a branch (wasn't created by an svn cop

    - by MatrixFrog
    I'm working on a team with lots of people who are pretty unfamiliar with the concepts of version control systems, and are just kind of doing whatever seems to work, by trial and error. Someone created a "branch" from the trunk that is not ancestrally related to the trunk. My guess is it went something like this: They created a folder in branches. They checked out all the code from the trunk to somewhere on their desktop. They added all that code to the newly created folder as though it was a bunch of brand new files. So the repository isn't aware that all that code is actually just a copy of the trunk. When I look at the history of that branch in TortoiseSVN, and uncheck the "Stop on copy/rename" box, there is no revision that has the trunk (or any other path) under the "Copy from path" column. Then they made lots of changes on their "branch". Meanwhile, others were making lots of changes on the trunk. We tried to do a merge and of course it doesn't work. Because, the trunk and the fake branch are not ancestrally related. I can see only two ways to resolve this: Go through the logs on the "branch", look at every change that was made, and manually apply each change to the trunk. Go through the logs on the trunk, look at every change that was made between revision 540 (when the "branch" was created) and HEAD, and manually apply each change to the "branch". This involves 7 revisions one way or 11 revisions the other way, so neither one is really that terrible. But is there any way to cause the repository to "realize" that the branch really IS ancestrally related even though it was created incorrectly, so that we can take advantage of the built-in merging functionality in Eclipse/TortoiseSVN? (You may be wondering: Why did your company hire these people and allow them to access the SVN repository without making sure they knew how to use it properly first?! We didn't -- this is a school assignment, which is a collaboration between two different classes -- the ones in the lower class were given a very quick hand-wavey "overview" of SVN which didn't really teach them anything. I've asked everyone in the group to please PLEASE read the svn book, and I'll make sure we (the slightly more experienced half of the team) keep a close eye on the repository to ensure this doesn't happen again.)

    Read the article

  • Optimizing Haskell code

    - by Masse
    I'm trying to learn Haskell and after an article in reddit about Markov text chains, I decided to implement Markov text generation first in Python and now in Haskell. However I noticed that my python implementation is way faster than the Haskell version, even Haskell is compiled to native code. I am wondering what I should do to make the Haskell code run faster and for now I believe it's so much slower because of using Data.Map instead of hashmaps, but I'm not sure I'll post the Python code and Haskell as well. With the same data, Python takes around 3 seconds and Haskell is closer to 16 seconds. It comes without saying that I'll take any constructive criticism :). import random import re import cPickle class Markov: def __init__(self, filenames): self.filenames = filenames self.cache = self.train(self.readfiles()) picklefd = open("dump", "w") cPickle.dump(self.cache, picklefd) picklefd.close() def train(self, text): splitted = re.findall(r"(\w+|[.!?',])", text) print "Total of %d splitted words" % (len(splitted)) cache = {} for i in xrange(len(splitted)-2): pair = (splitted[i], splitted[i+1]) followup = splitted[i+2] if pair in cache: if followup not in cache[pair]: cache[pair][followup] = 1 else: cache[pair][followup] += 1 else: cache[pair] = {followup: 1} return cache def readfiles(self): data = "" for filename in self.filenames: fd = open(filename) data += fd.read() fd.close() return data def concat(self, words): sentence = "" for word in words: if word in "'\",?!:;.": sentence = sentence[0:-1] + word + " " else: sentence += word + " " return sentence def pickword(self, words): temp = [(k, words[k]) for k in words] results = [] for (word, n) in temp: results.append(word) if n > 1: for i in xrange(n-1): results.append(word) return random.choice(results) def gentext(self, words): allwords = [k for k in self.cache] (first, second) = random.choice(filter(lambda (a,b): a.istitle(), [k for k in self.cache])) sentence = [first, second] while len(sentence) < words or sentence[-1] is not ".": current = (sentence[-2], sentence[-1]) if current in self.cache: followup = self.pickword(self.cache[current]) sentence.append(followup) else: print "Wasn't able to. Breaking" break print self.concat(sentence) Markov(["76.txt"]) -- module Markov ( train , fox ) where import Debug.Trace import qualified Data.Map as M import qualified System.Random as R import qualified Data.ByteString.Char8 as B type Database = M.Map (B.ByteString, B.ByteString) (M.Map B.ByteString Int) train :: [B.ByteString] -> Database train (x:y:[]) = M.empty train (x:y:z:xs) = let l = train (y:z:xs) in M.insertWith' (\new old -> M.insertWith' (+) z 1 old) (x, y) (M.singleton z 1) `seq` l main = do contents <- B.readFile "76.txt" print $ train $ B.words contents fox="The quick brown fox jumps over the brown fox who is slow jumps over the brown fox who is dead."

    Read the article

  • Stack and queue operations on the same array.

    - by Passonate Learner
    Hi. I've been thinking about a program logic, but I cannot draw a conclusion to my problem. Here, I've implemented stack and queue operations to a fixed array. int A[1000]; int size=1000; int top; int front; int rear; bool StackIsEmpty() { return (top==0); } bool StackPush( int x ) { if ( top >= size ) return false; A[top++] = x; return true; } int StackTop( ) { return A[top-1]; } bool StackPop() { if ( top <= 0 ) return false; A[--top] = 0; return true; } bool QueueIsEmpty() { return (front==rear); } bool QueuePush( int x ) { if ( rear >= size ) return false; A[rear++] = x; return true; } int QueueFront( ) { return A[front]; } bool QueuePop() { if ( front >= rear ) return false; A[front++] = 0; return true; } It is presumed(or obvious) that the bottom of the stack and the front of the queue is pointing at the same location, and vice versa(top of the stack points the same location as rear of the queue). For example, integer 1 and 2 is inside an array in order of writing. And if I call StackPop(), the integer 2 will be popped out, and if I call QueuePop(), the integer 1 will be popped out. My problem is that I don't know what happens if I do both stack and queue operations on the same array. The example above is easy to work out, because there are only two values involved. But what if there are more than 2 values involved? For example, if I call StackPush(1); QueuePush(2); QueuePush(4); StackPop(); StackPush(5); QueuePop(); what values will be returned in the order of bottom(front) from the final array? I know that if I code a program, I would receive a quick answer. But the reason I'm asking this is because I want to hear a logical explanations from a human being, not a computer.

    Read the article

  • In Rails, how should I implement a Status field for a Tasks app - integer or enum?

    - by Doug
    For a Rails 3.0 Todo app, I have a Tasks model with a Status field. What's the best way to store the Status field data (field type) and still display a human-readable version in a view (HTML table)? Status can be: 0 = Normal 1 = Active 2 = Completed Right now I have this: Rails Schema Here: create_table "tasks", :force = true do |t| t.integer "status", :limit = 1, :default = 0, :null = false Rails Model Here: class Task < ActiveRecord::Base validates_inclusion_of :status, :in => 0..2, :message => "{{value}} must be 0, 1, or 2" Rails View Here: <h1>Listing tasks</h1> <table> <tr> <th>Status</th> <th>Name</th> <th></th> <th></th> <th></th> </tr> <% @tasks.each do |task| %> <tr> <td><%= task.status %></td> <td><%= task.name %></td> <td><%= link_to 'Show', task %></td> <td><%= link_to 'Edit', edit_task_path(task) %></td> <td><%= link_to 'Delete', task, :confirm => 'Are you sure?', :method => :delete %></td> </tr> <% end %> </table> Requirements Store a Task's status in the db such that the values are easily localizable, i.e. I'm not sure I want to store "normal", "active", "completed" as a string field. Solution must work with Rails 3.0. Questions: Should I store the field as an integer (see above)? If so, how do I display the correct human readable status in an HTML table in my Rails view, e.g. show "Active" instead of "1" in the HTML table. Should I use an enum? If so, is this easy to localize later? Should I use straight strings, e.g. "Normal", "Active", "Completed" Can you provide a quick code sample of the view helper, controller or view code to make this work?

    Read the article

  • Qt vs .NET - a few comparisons [closed]

    - by Pirate for Profit
    Event Handling In Qt the event handling system you just emit signals when something cool happens and then catch them in slots, for instance emit valueChanged(int percent, bool something); and void MyCatcherObj::valueChanged(int p, bool ok){} blocking them and disconnecting them when needed, doing it across threads... once you get the hang of it, it just seems a lot more natural and intuitive than the way the .NET event handling is set up (you know, object sender, CustomEventArgs e). And I'm not just talking about syntax, because in the end the .NET delegate crap is the bomb. I'm also talking about in more than just reflection (because, yes, .NET obviously has much stronger reflection capabilities). I'm talking about in the way the system feels to a human being. Qt wins hands down i m o. Basically, the footprints make more sense and you can visualize the project easier without the clunky event handling system. I wish I could it explain it better. The only thing is, I do love some of the ease of C# compared to C++ and .NET's assembly architecture. That is a big bonus for modular projects, which are a PITA to do in C++. Database Ease of Doing Crap Also what about datasets and database manipulations. I think .net wins here but I'm not sure. Threading/Conccurency How do you guys think of the threading? In .NET, all I've ever done is make like a list of master worker threads with locks. I like QConcurrentFramework, you don't worry about locks or anything, and with the ease of the signal slot system across threads it's nice to get notified about the progress of things. Memory Usage Also what do you think of the overall memory usage comparison. Is the .NET garbage collector pretty on the ball and quick compared to the instantaneous nature of native memory management? Or does it just let programs leak up a storm and lag the computer then clean it up when it's about to really lag? However, I am a n00b who doesn't know what I'm talking about, please school me on the subject.

    Read the article

  • What's the correct way to use Cakephp urls?

    - by Pichan
    Hello all, it's my first post here :) I'm having some difficulties with dealing with urls and parameters. I've gone through the router class api documentation over and over again and found nothing useful. First of all, I'd like to know if there is any 'universal' format in CakePHP(1.3) for handling urls. I'm currently handling all my urls as simple arrays(in the format that Router::url and $html-link accepts) and it's easy as long as I only need to pass them as arguments to cake's own methods. It usually gets tricky if I need something else. Mainly I'm having problems with converting string urls to the basic array format. Let's say I want to convert $arrayUrl to string and than again into url: $arrayUrl=array('controller'=>'SomeController','action'=>'someAction','someValue'); $url=Router::url($arrayUrl); //$url is now '/path/to/site/someController/someAction/someValue' $url=Router::normalize($url); //remove '/path/to/site' $parsed=Router::parse($url); /*$parsed is now Array( [controller] => someController [action] => someAction [named] => Array() [pass] => Array([0] => someValue) [plugin] => ) */ That seems an awful lot of code to do something as simple as to convert between 2 core formats. Also, note that $parsed is still not in the same as $arrayUrl. Of course I could tweak $parsed manually and actually I've done that a few times as a quick patch but I'd like to get to the bottom of this. I also noticed that when using prefix routing, $this-params in controller has the prefix embedded in the action(i.e. [action] = 'admin_edit') and the result of Router::parse() does not. Both of course have the prefix in it's own key. To summarize, how do I convert an url between any of these 3(or 4, if you include the prefix thing) mentioned formats the right way? Of course it would be easy to hack my way through this, but I'd still like to believe that cake is being developed by a bunch of people who have a lot more experience and insight than me, so I'm guessing there's a good reason for this "perceived misbehavior". I've tried to present my problem as good as I can, but due to my rusty english skills, I had to take a few detours :) I'll explain more if needed.

    Read the article

  • Armchair Linguists: 'code' vs. 'codes'--or why I write 'code' and my manager asks for 'codes'

    - by Ukko
    I wanted to tap into the collective wisdom here to see if I can get some insight into one of my pet peeves, people who thread "code" as a countable noun. Let me also preface this by saying that I am not talking about anyone who speaks english as a second language, this is a native phenomenon. For those of us who slept through grammar class there are two classes of nouns which basically refer to things that are countable and non-countable (sometimes referred to as count and noncount). For instance 'sand' is a non-count noun and 'apple' is count. You can talk about "two apples" but "two sands" does not parse. The bright students then would point out a word like "beer" where is looks like this is violated. Beer as a substance is certainly a non-count noun, but I can ask for "two beers" without offending the grammar police. The reason is that there are actually two words tied up in that one utterance, Definition #1 is a yummy golden substance and Definition #2 is a colloquial term for a container of said substance. #1 is non-count and #2 is countable. This gets to my problem with "codes" as a countable noun. In my mind the code that we programmers write is non-count, "I wrote some code today." When used in the plural like "Have you got the codes" I can only assume that you are asking if I have the cryptographically significant numbers for launching a missile or the like. Every time my peer in marketing asks about when we will have the new codes ready I have a vision of rooms of code breakers going over the latest Enigma coded message. I corrected the usage in all the documents I am asked to review, but then I noticed that our customer was also using the work "codes" when they meant "code". At this point I have realized that there is a significant sub-population that uses "codes" and they seem to be impervious to what I see as the dominant "correct" usage. This is the part I want some help on, has anyone else noticed this phenomenon? Do you know what group it is associated with, old Fortran programmer perhaps? Is it a regionalism? I have become quick to change my terms when I notice a customer's usage, but it would be nice to know if I am sending a proposal somewhere what style they expect. I would hate to get canned with a review of "Ha, these guy's must be morons they don't even know 'code' is plural!"

    Read the article

  • Dependency property does not work within a geometry in a controltemplate

    - by Erik Bongers
    I have a DepencencyProperty (a boolean) that works fine on an Ellipse, but not on an ArcSegment. Am I doing something that is not possible? Here's part of the xaml. Both the TemplateBindings of Origin and LargeArc do not work in the geometry. But the LargeArc DependencyProperty does work in the Ellipse, so my DependencyProperty seems to be set up correctly. <ControlTemplate TargetType="{x:Type nodes:TestCircle}"> <Canvas Background="AliceBlue"> <Ellipse Height="10" Width="10" Fill="Yellow" Visibility="{TemplateBinding LargeArc, Converter={StaticResource BoolToVisConverter}}"/> <Path Canvas.Left="0" Canvas.Top="0" Stroke="Black" StrokeThickness="3"> <Path.Data> <GeometryGroup> <PathGeometry> <PathFigure IsClosed="True" StartPoint="{TemplateBinding Origin}"> <LineSegment Point="150,100" /> <ArcSegment Point="140,150" IsLargeArc="{TemplateBinding LargeArc}" Size="50,50" SweepDirection="Clockwise"/> </PathFigure> </PathGeometry> </GeometryGroup> </Path.Data> </Path> </Canvas> </ControlTemplate> What I'm trying to build is a (sort of) pie-shaped usercontrol where the shape of the Pie is defined by DependencyProperties and the actual graphics used are in a template, so they can be replaced or customized. In other words: I would like the code-behind to be visual-free (which, I assume, is good separation). SOLUTION--------------------------(I'm not allowed to answer my own questions yet) I found the answer myself, and this can be useful for others encountering the same issue. This is why the TemplateBinding on the Geometry failed: A TemplateBinding will only work when binding a DependencyProperty to another DependencyProperty. Following article set me on the right track: http://blogs.msdn.com/b/liviuc/archive/2009/12/14/wpf-templatebinding-vs-relativesource-templatedparent.aspx The ArcSegment properties are no DependencyProperties. Thus, the solution to the above problem is to replace <ArcSegment Point="140,150" IsLargeArc="{TemplateBinding LargeArc}" with <ArcSegment Point="140,150" IsLargeArc="{Binding RelativeSource={RelativeSource TemplatedParent}, Path=LargeArc}" Colin, your working example where an 'ordinary' binding was used in the geometry set me on the right track. BTW, love the infographics and the construction of your UserControl in your blogpost. And, hey, that quick tip on code snippets, and especially on that DP attribute and the separation of those DPs into a partial class file is pure gold!

    Read the article

  • Why does SFINAE not apply to this?

    - by Simon Buchan
    I'm writing some simple point code while trying out Visual Studio 10 (Beta 2), and I've hit this code where I would expect SFINAE to kick in, but it seems not to: template<typename T> struct point { T x, y; point(T x, T y) : x(x), y(y) {} }; template<typename T, typename U> struct op_div { typedef decltype(T() / U()) type; }; template<typename T, typename U> point<typename op_div<T, U>::type> operator/(point<T> const& l, point<U> const& r) { return point<typename op_div<T, U>::type>(l.x / r.x, l.y / r.y); } template<typename T, typename U> point<typename op_div<T, U>::type> operator/(point<T> const& l, U const& r) { return point<typename op_div<T, U>::type>(l.x / r, l.y / r); } int main() { point<int>(0, 1) / point<float>(2, 3); } This gives error C2512: 'point<T>::point' : no appropriate default constructor available Given that it is a beta, I did a quick sanity check with the online comeau compiler, and it agrees with an identical error, so it seems this behavior is correct, but I can't see why. In this case some workarounds are to simply inline the decltype(T() / U()), to give the point class a default constructor, or to use decltype on the full result expression, but I got this error while trying to simplify an error I was getting with a version of op_div that did not require a default constructor*, so I would rather fix my understanding of C++ rather than to just do what works. Thanks! *: the original: template<typename T, typename U> struct op_div { static T t(); static U u(); typedef decltype(t() / u()) type; }; Which gives error C2784: 'point<op_div<T,U>::type> operator /(const point<T> &,const U &)' : could not deduce template argument for 'const point<T> &' from 'int', and also for the point<T> / point<U> overload.

    Read the article

  • Ajax to read updated values from XML

    - by punit
    I am creating file upload progress bar. I have a CGI script which copies the data, and here I increment the progress bar value by ONE after certain iterations. I am storing the incremented value in XML file (I also tried using plain text file). On the other side I have ajax reading incremented value from xml and depending on that it increments the DIV element. However, what happens here is, it seems to me that although the ajax reads all the incremented values but it processes it after the CGI has finished execution. That is progress bar starts execution once the file copying and other stuff in CGI is completed. My code is: AJAX:::: function polling_start() { //GETS CALLED WHEN USER HITS FILE UPLOAD BUTTON intervalID = window.setInterval(send_request,100); } window.onload = function (){ request = initXMLHttpClient(); request.overrideMimeType('text/xml'); progress = document.getElementById('progress'); } function initXMLHttpClient() { if (window.XMLHttpRequest){ // code for IE7+, Firefox, Chrome, Opera, Safari xmlhttp = new XMLHttpRequest(); } else{ // code for IE6, IE5 xmlhttp = new ActiveXObject("Microsoft.XMLHTTP"); } return xmlhttp } function send_request() { request.open("GET","progress_bar.xml",true); request.onreadystatechange = request_handler; request.send(); } function request_handler() { if (request.readyState == 4 && request.status == 200) { var level=request.responseXML.getElementsByTagName('PROGRESS')[0].firstChild; progress.style.width = progress.innerHTML = level.nodeValue + '%'; progress.style.backgroundColor = "green"; } } /****ON SERVER SIDE*********/ char xmlDat1[] = "<DOCUMENT><PROGRESS>"; char xmlDat2[] = "</PROGRESS></DOCUMENT>"; fptr = fopen("progress_bar.xml", "w"); .........OTHER STUFF.............................. ................................. if(i == inc && j<=100) { fprintf(fptr, "%s\n", "\n\n\n]"); //fprintf(fptr, "%s\n", ""); fprintf(fptr, "%s", xmlDat1); // fprintf(fptr, "%d" ,j); fprintf(fptr, "%s" ,xmlDat2); fseek(fptr, 0, SEEK_SET); /*fprintf(fptr, "%d" ,j); fseek(fptr, 0, SEEK_SET);*/ i = 0; //sleep(1); j++; } (I also tried to write in .text, but same response) Any quick response would be appreciable.

    Read the article

  • Available Coroutine Libraries in Java

    - by JUST MY correct OPINION
    I would like to do some stuff in Java that would be clearer if written using concurrent routines, but for which full-on threads are serious overkill. The answer, of course, is the use of coroutines, but there doesn't appear to be any coroutine support in the standard Java libraries and a quick Google on it brings up tantalising hints here or there, but nothing substantial. Here's what I've found so far: JSIM has a coroutine class, but it looks pretty heavyweight and conflates, seemingly, with threads at points. The point of this is to reduce the complexity of full-on threading, not to add to it. Further I'm not sure that the class can be extracted from the library and used independently. Xalan has a coroutine set class that does coroutine-like stuff, but again it's dubious if this can be meaningfully extracted from the overall library. It also looks like it's implemented as a tightly-controlled form of thread pool, not as actual coroutines. There's a Google Code project which looks like what I'm after, but if anything it looks more heavyweight than using threads would be. I'm basically nervous of something that requires software to dynamically change the JVM bytecode at runtime to do its work. This looks like overkill and like something that will cause more problems than coroutines would solve. Further it looks like it doesn't implement the whole coroutine concept. By my glance-over it gives a yield feature that just returns to the invoker. Proper coroutines allow yields to transfer control to any known coroutine directly. Basically this library, heavyweight and scary as it is, only gives you support for iterators, not fully-general coroutines. The promisingly-named Coroutine for Java fails because it's a platform-specific (obviously using JNI) solution. And that's about all I've found. I know about the native JVM support for coroutines in the Da Vinci Machine and I also know about the JNI continuations trick for doing this. These are not really good solutions for me, however, as I would not necessarily have control over which VM or platform my code would run on. (Indeed any bytecode manipulation system would suffer similar problems -- it would be best were this pure Java if possible. Runtime bytecode manipulation would restrict me from using this on Android, for example.) So does anybody have any pointers? Is this even possible? If not, will it be possible in Java 7? Edited to add: Just to ensure that confusion is contained, this is a related question to my other one, but not the same. This one is looking for an existing implementation in a bid to avoid reinventing the wheel unnecessarily. The other one is a question relating to how one would go about implementing coroutines in Java should this question prove unanswerable. The intent is to keep different questions on different threads.

    Read the article

  • Choosing the right .NET architecture. WCF? WPF/Forms, ASP.NET (MVC)?

    - by Tommy Jakobsen
    I’m in the situation that I have to design and implement a rather big system from the bottom. I’m having some (a lot actually) questions about the architecture that I would like your comments and thoughts on. I don’t hope that I’ve written too much here, but I wanted to give you all an idea of what the system is. Quick info about the applications, read if you want: I can’t share much detail about the project, but basically it’s a system where we offer our customer a service to manage their users. We have a hotline where the users call and our hotline uses an (windows) application (intranet) to manage the user’s data, etc. The customer also has a web application where they can see reports, information about their business and users, and the ability to modify their data. Modifying data is not just user data like address and so, but also information about the products/services the user has, which can be complicated. The applications will be built on Microsoft .NET Framework 4, with a MS SQL Server 2008 database. There will be a few applications that have to access this database, such as: Intranet application (used by us and our hotline) Customer web application type 1 Customer web application type 2 Customer web application type n different applications) … Now my big problem is what .NET parts I should use for such a system. For the “backend” I’ve considered using Windows Communication Foundation: Would WCF be a good choice? The intranet application will be an application that has to edit lots of records in the database. It has to be easy to navigate using the keyboard (fast to work with). Has a feature such as “find customer, find that, lookup this, choose this and update that”. What would be the best choice to develop this application in? Would it be WPF or good old Windows Forms? I don’t need all of the fancy graphics features in WPF, like 3D, but the application has to look nice (maybe something like the new Visual Studio/Office tools). And the same question goes for the web pages. They have much the same work to do, but not as many features as the intranet application, and not the same amount of data (much less). That is my questions for now. I’m hoping to get a discussion going that will open my eyes to some of these technologies, helping me decide architecture to go with. I would like to say thanks in advance, and let you all know that any thoughts will be much appreciated.

    Read the article

  • Looking for Fiddler2 help. connection to gateway refused? Just got rid of a virus

    - by John Mackey
    I use Fiddler2 for facebook game items, and it's been a great success. I accessed a website to download some dat files I needed. I think it was eshare, ziddu or megaupload, one of those. Anyway, even before the rar file had downloaded, I got this weird green shield in the bottom right hand corner of my computer. It said a Trojan was trying to access my computer, or something to that extent. It prompted me to click the shield to begin anti-virus scanning. It turns out this rogue program is called Antivirus System Pro and is pretty hard to get rid of. After discovering the rogue program, I tried using Fiddler and got the following error: [Fiddler] Connection to Gateway failed.Exception Text: No connection could be made because the target machine actively refused it 127.0.0.1:5555 I ended up purchasing SpyDoctor + Antivirus, which I'm told is designed specifically for getting rid of these types of programs. Anyway, I did a quick-scan last night with spydoctor and malware bytes. Malware picked up 2 files, and Spydoctor found 4. Most were insignificant, but it did find a worm called Worm.Alcra.F, which was labeled high-priority. I don’t know if that’s the Anti-Virus Pro or not, but SpyDoctor said it got rid of all of those successfully. I tried to run Fiddler again before leaving home, but was still getting the "gateway failed" error. Im using the newest version of firefox. When I initially set up the Fiddler 2.2.8.6, I couldn’t get it to run at first, so I found this faq on the internet that said I needed to go through ToolsOptionsSettings and set up an HTTP Proxy to 127.0.0.1 and my Port to 8888. Once I set that up and downloaded this fiddler helper as a firefox add-on, it worked fine. When I turn on fiddler, it automatically takes my proxy setting from no proxy (default) to the 127.0.0.1 with Port 8888 set up. It worked fine until my computer detected this virus. Anyway, hopefully I've given you sufficient information to offer me your best advice here. Like I said, Spydoctor says the bad stuff is gone, so maybe the rogue program made some type of change in my fiddler that I could just reset or uncheck or something like that? Or will I need to completely remove fiddler and those dat files and rar files I downloaded? Any help would be greatly appreciated. Thanks for your time.

    Read the article

  • HTML wrapper div over embedded flash object cannot be "clickable" by jQuery

    - by Michael Mao
    Hi all: I've been trying to do as the client requested : redirect to campaign page then to destination page once a customer clicks on the top banner in swf format. You can check what's been done at :http://ausdcf.org If you are using Firefox, Chrome or Safari, I suspect you can reach the destination page. However, if you are using IE or Opera, I doubt it. I think to cause of such a weird problem is: The swf ojbects don't have a link to url, SO I have to hack the theme template file like this : <div id="header">'; /* * A quick and dirty way to put some swf into PHP, and rotate among them ... */ //available banners $banners = array( 'http://localhost/smf/flash/banner_fertalign_1.swf', 'http://localhost/smf/flash/banner_fertalign_2.swf', 'http://localhost/smf/flash/banner_fertalign_3.swf' ); //get random banner srand((double) microtime() * 1000000); $rand = rand(0,count($banners)-1); echo '<div id="top_banner_clickable">'; echo '<div id="top_banner_wrapper">'; echo '<object width="400" height="60">'; echo '<param name="wmode" value="transparent">'; echo '<embed wmode="transparent" src="'.$banners[$rand].'" '; echo 'width="400" height="60" type="application/x-shockwave-flash"'; echo 'pluginspage="http://www.macromedia.com/shockwave/download/index.cgi?P1_Prod_Version=ShockwaveFlash" />'; echo '</object>'; echo '</div>'; echo '</div>'; And the related jQuery code is like this: /* master.js */ $(document).ready(function() { $("#top_banner_clickable").click(function() { window.location ="http://ausdcf.org/campaign/"; }); }); I absolutely know nothing about Flash or embedded objects. I guess that's the cause of this problem. Plus, I don't know why it works with some browsers but not all... I even tried to add a z-index to the wrapper div in css like this: #top_banner_clickable { z-index : 100; } No this wouldn't do, either... Is there a way to go around this issue? Many thanks in advance.

    Read the article

  • Multiprogramming in Django, writing to the Database

    - by Marcus Whybrow
    Introduction I have the following code which checks to see if a similar model exists in the database, and if it does not it creates the new model: class BookProfile(): # ... def save(self, *args, **kwargs): uniqueConstraint = {'book_instance': self.book_instance, 'collection': self.collection} # Test for other objects with identical values profiles = BookProfile.objects.filter(Q(**uniqueConstraint) & ~Q(pk=self.pk)) # If none are found create the object, else fail. if len(profiles) == 0: super(BookProfile, self).save(*args, **kwargs) else: raise ValidationError('A Book Profile for that book instance in that collection already exists') I first build my constraints, then search for a model with those values which I am enforcing must be unique Q(**uniqueConstraint). In addition I ensure that if the save method is updating and not inserting, that we do not find this object when looking for other similar objects ~Q(pk=self.pk). I should mention that I ham implementing soft delete (with a modified objects manager which only shows non-deleted objects) which is why I must check for myself rather then relying on unique_together errors. Problem Right thats the introduction out of the way. My problem is that when multiple identical objects are saved in quick (or as near as simultaneous) succession, sometimes both get added even though the first being added should prevent the second. I have tested the code in the shell and it succeeds every time I run it. Thus my assumption is if say we have two objects being added Object A and Object B. Object A runs its check upon save() being called. Then the process saving Object B gets some time on the processor. Object B runs that same test, but Object A has not yet been added so Object B is added to the database. Then Object A regains control of the processor, and has allready run its test, even though identical Object B is in the database, it adds it regardless. My Thoughts The reason I fear multiprogramming could be involved is that each Object A and Object is being added through an API save view, so a request to the view is made for each save, thus not a single request with multiple sequential saves on objects. It might be the case that Apache is creating a process for each request, and thus causing the problems I think I am seeing. As you would expect, the problem only occurs sometimes, which is characteristic of multiprogramming or multiprocessing errors. If this is the case, is there a way to make the test and set parts of the save() method a critical section, so that a process switch cannot happen between the test and the set?

    Read the article

< Previous Page | 230 231 232 233 234 235 236 237 238 239 240 241  | Next Page >