Search Results

Search found 11509 results on 461 pages for 'description logic'.

Page 358/461 | < Previous Page | 354 355 356 357 358 359 360 361 362 363 364 365  | Next Page >

  • Nesting Linq-to-Objects query within Linq-to-Entities query –what is happening under the covers?

    - by carewithl
    var numbers = new int[] { 1, 2, 3, 4, 5 }; var contacts = from c in context.Contacts where c.ContactID == numbers.Max() | c.ContactID == numbers.FirstOrDefault() select c; foreach (var item in contacts) Console.WriteLine(item.ContactID); Linq-to-Entities query is first translated into Linq expression tree, which is then converted by Object Services into command tree. And if Linq-to-Entities query nests Linq-to-Objects query, then this nested query also gets translated into an expression tree. a) I assume none of the operators of the nested Linq-to-Objects query actually get executed, but instead data provider for particular DB (or perhaps Object Services) knows how to transform the logic of Linq-to-Objects operators into appropriate SQL statements? b) Data provider knows how to create equivalent SQL statements only for some of the Linq-to-Objects operators? c) Similarly, data provider knows how to create equivalent SQL statements only for some of the non-Linq methods in the Net Framework class library? EDIT: I know only some Sql so I can't be completely sure, but reading Sql query generated for the above code it seems data provider didn't actually execute numbers.Max method, but instead just somehow figured out that numbers.Max should return the maximum value and then proceed to include in generated Sql query a call to TSQL's build-in MAX function. It also put all the values held by numbers array into a Sql query. SELECT CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN '0X0X' ELSE '0X1X' END AS [C1], [Extent1].[ContactID] AS [ContactID], [Extent1].[FirstName] AS [FirstName], [Extent1].[LastName] AS [LastName], [Extent1].[Title] AS [Title], [Extent1].[AddDate] AS [AddDate], [Extent1].[ModifiedDate] AS [ModifiedDate], [Extent1].[RowVersion] AS [RowVersion], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[CustomerTypeID] END AS [C2], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[InitialDate] END AS [C3], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[PrimaryDesintation] END AS [C4], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[SecondaryDestination] END AS [C5], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[PrimaryActivity] END AS [C6], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[SecondaryActivity] END AS [C7], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[Notes] END AS [C8], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[RowVersion] END AS [C9], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[BirthDate] END AS [C10], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[HeightInches] END AS [C11], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[WeightPounds] END AS [C12], CASE WHEN (([Project1].[C1] = 1) AND ([Project1].[C1] IS NOT NULL)) THEN [Project1].[DietaryRestrictions] END AS [C13] FROM [dbo].[Contact] AS [Extent1] LEFT OUTER JOIN (SELECT [Extent2].[ContactID] AS [ContactID], [Extent2].[BirthDate] AS [BirthDate], [Extent2].[HeightInches] AS [HeightInches], [Extent2].[WeightPounds] AS [WeightPounds], [Extent2].[DietaryRestrictions] AS [DietaryRestrictions], [Extent3].[CustomerTypeID] AS [CustomerTypeID], [Extent3].[InitialDate] AS [InitialDate], [Extent3].[PrimaryDesintation] AS [PrimaryDesintation], [Extent3].[SecondaryDestination] AS [SecondaryDestination], [Extent3].[PrimaryActivity] AS [PrimaryActivity], [Extent3].[SecondaryActivity] AS [SecondaryActivity], [Extent3].[Notes] AS [Notes], [Extent3].[RowVersion] AS [RowVersion], cast(1 as bit) AS [C1] FROM [dbo].[ContactPersonalInfo] AS [Extent2] INNER JOIN [dbo].[Customers] AS [Extent3] ON [Extent2].[ContactID] = [Extent3].[ContactID]) AS [Project1] ON [Extent1].[ContactID] = [Project1].[ContactID] LEFT OUTER JOIN (SELECT TOP (1) [c].[C1] AS [C1] FROM (SELECT [UnionAll3].[C1] AS [C1] FROM (SELECT [UnionAll2].[C1] AS [C1] FROM (SELECT [UnionAll1].[C1] AS [C1] FROM (SELECT 1 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable1] UNION ALL SELECT 2 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable2]) AS [UnionAll1] UNION ALL SELECT 3 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable3]) AS [UnionAll2] UNION ALL SELECT 4 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable4]) AS [UnionAll3] UNION ALL SELECT 5 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable5]) AS [c]) AS [Limit1] ON 1 = 1 LEFT OUTER JOIN (SELECT TOP (1) [c].[C1] AS [C1] FROM (SELECT [UnionAll7].[C1] AS [C1] FROM (SELECT [UnionAll6].[C1] AS [C1] FROM (SELECT [UnionAll5].[C1] AS [C1] FROM (SELECT 1 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable6] UNION ALL SELECT 2 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable7]) AS [UnionAll5] UNION ALL SELECT 3 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable8]) AS [UnionAll6] UNION ALL SELECT 4 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable9]) AS [UnionAll7] UNION ALL SELECT 5 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable10]) AS [c]) AS [Limit2] ON 1 = 1 CROSS JOIN (SELECT MAX([UnionAll12].[C1]) AS [A1] FROM (SELECT [UnionAll11].[C1] AS [C1] FROM (SELECT [UnionAll10].[C1] AS [C1] FROM (SELECT [UnionAll9].[C1] AS [C1] FROM (SELECT 1 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable11] UNION ALL SELECT 2 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable12]) AS [UnionAll9] UNION ALL SELECT 3 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable13]) AS [UnionAll10] UNION ALL SELECT 4 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable14]) AS [UnionAll11] UNION ALL SELECT 5 AS [C1] FROM (SELECT 1 AS X) AS [SingleRowTable15]) AS [UnionAll12]) AS [GroupBy1] WHERE [Extent1].[ContactID] IN ([GroupBy1].[A1], (CASE WHEN ([Limit1].[C1] IS NULL) THEN 0 ELSE [Limit2].[C1] END)) Based on this, is it possible that Linq2Entities provider indeed doesn't execute non-Linq and Linq-to-Object methods, but instead creates equivalent SQL statements for some of them ( and for others it throws an exception )? Thank you in advance

    Read the article

  • Removing Duplicate Data From SQL Query Output For Display On A Web Page [migrated]

    - by doubleJ
    I had asked a similar question on stackoverflow but didn't really get anywhere. This page shows the output that I'm currently getting from my MSSQL server. I have a table of venue information (name, address, etc...) that our events happen on. Separately, I have a table of the actual events that are scheduled (an event may happen multiple times in one day and/or over multiple days). I join those tables with this query: <?php try { $dbh = new PDO("sqlsrv:Server=localhost;Database=Sermons", "", ""); $dbh->setAttribute(PDO::ATTR_ERRMODE, PDO::ERRMODE_EXCEPTION); $sql = "SELECT TOP (100) PERCENT dbo.TblSermon.Day, dbo.TblSermon.Date, dbo.TblSermon.Time, dbo.TblSermon.Speaker, dbo.TblSermon.Series, dbo.TblSermon.Sarasota, dbo.TblSermon.NonFlc, dbo.TblJoinSermonLocation.MeetingName, dbo.TblLocation.Location, dbo.TblLocation.Pastors, dbo.TblLocation.Address, dbo.TblLocation.City, dbo.TblLocation.State, dbo.TblLocation.Zip, dbo.TblLocation.Country, dbo.TblLocation.Phone, dbo.TblLocation.Email, dbo.TblLocation.WebAddress FROM dbo.TblLocation RIGHT OUTER JOIN dbo.TblJoinSermonLocation ON dbo.TblLocation.ID = dbo.TblJoinSermonLocation.Location RIGHT OUTER JOIN dbo.TblSermon ON dbo.TblJoinSermonLocation.Sermon = dbo.TblSermon.ID WHERE (dbo.TblSermon.Date >= { fn NOW() }) ORDER BY dbo.TblSermon.Date, dbo.TblSermon.Time"; $stmt = $dbh->prepare($sql); $stmt->execute(); $stmt->setFetchMode(PDO::FETCH_ASSOC); foreach ($stmt as $row) { echo "<pre>"; print_r($row); echo "</pre>"; } unset($row); $dbh = null; } catch(PDOException $e) { echo $e->getMessage(); } ?> So, as it loops through the query results, it creates an array for each record and ends up like this: Array ( [Day] => Tuesday [Date] => 2012-10-30 00:00:00.000 [Time] => 07:00 PM [Speaker] => Keith Moore [Location] => The Ark Church [Pastors] => Alan & Joy Clayton [Address] => 450 Humble Tank Rd. [City] => Conroe [State] => TX [Zip] => 77305.0 [Phone] => (936) 756-1988 [Email] => [email protected] [WebAddress] => http://www.thearkchurch.org ) Array ( [Day] => Wednesday [Date] => 2012-10-31 00:00:00.000 [Time] => 07:00 PM [Speaker] => Keith Moore [Location] => The Ark Church [Pastors] => Alan & Joy Clayton [Address] => 450 Humble Tank Rd. [City] => Conroe [State] => TX [Zip] => 77305.0 [Phone] => (936) 756-1988 [Email] => [email protected] [WebAddress] => http://www.thearkchurch.org ) Array ( [Day] => Tuesday [Date] => 2012-11-06 00:00:00.000 [Time] => 07:00 PM [Speaker] => Keith Moore [Location] => Fellowship Of Faith Christian Center [Pastors] => Michael & Joan Kalstrup [Address] => 18999 Hwy. 59 [City] => Oakland [State] => IA [Zip] => 51560.0 [Phone] => (712) 482-3455 [Email] => [email protected] [WebAddress] => http://www.fellowshipoffaith.cc ) Array ( [Day] => Wednesday [Date] => 2012-11-14 00:00:00.000 [Time] => 07:00 PM [Speaker] => Keith Moore [Location] => Faith Family Church [Pastors] => Michael & Barbara Cameneti [Address] => 8200 Freedom Ave NW [City] => Canton [State] => OH [Zip] => 44720.0 [Phone] => (330) 492-0925 [Email] => [WebAddress] => http://www.myfaithfamily.com ) As you can see, The Ark Church and its associated contact information is duplicated, so when I work with those arrays and output them to the page, I see a bunch of duplicate content. I'd like to remove the duplicate information so that I get results similar to this: The Ark Church Alan & Joy Clayton 450 Humble Tank Rd. Conroe, TX 77305 (936) 756-1988 [email protected] http://www.thearkchurch.org Meetings: Tuesday, 2012-10-30 07:00 PM Wednesday, 2012-10-31 07:00 PM Fellowship Of Faith Christian Center Michael & Joan Kalstrup 18999 Hwy. 59 Oakland, IA 51560 (712) 482-3455 [email protected] http://www.fellowshipoffaith.cc Meetings: Tuesday, 2012-11-06 07:00 PM Faith Family Church Michael & Barbara Cameneti 8200 Freedom Ave NW Canton, OH 44720 (330) 492-0925 http://www.myfaithfamily.com Meetings: Wednesday, 2012-11-14 07:00 PM It doesn't necessarily have to end up like that (I'm not looking for code specific for these results, but a concept of how to not show the duplicated information). I'm assuming that an additional foreach or while will do it, but I haven't figured out any logic that says <?php if ($location == $previouslocation) echo ""; ?>.

    Read the article

  • Breaking through the class sealing

    - by Jason Crease
    Do you understand 'sealing' in C#?  Somewhat?  Anyway, here's the lowdown. I've done this article from a C# perspective, but I've occasionally referenced .NET when appropriate. What is sealing a class? By sealing a class in C#, you ensure that you ensure that no class can be derived from that class.  You do this by simply adding the word 'sealed' to a class definition: public sealed class Dog {} Now writing something like " public sealed class Hamster: Dog {} " you'll get a compile error like this: 'Hamster: cannot derive from sealed type 'Dog' If you look in an IL disassembler, you'll see a definition like this: .class public auto ansi sealed beforefieldinit Dog extends [mscorlib]System.Object Note the addition of the word 'sealed'. What about sealing methods? You can also seal overriding methods.  By adding the word 'sealed', you ensure that the method cannot be overridden in a derived class.  Consider the following code: public class Dog : Mammal { public sealed override void Go() { } } public class Mammal { public virtual void Go() { } } In this code, the method 'Go' in Dog is sealed.  It cannot be overridden in a subclass.  Writing this would cause a compile error: public class Dachshund : Dog { public override void Go() { } } However, we can 'new' a method with the same name.  This is essentially a new method; distinct from the 'Go' in the subclass: public class Terrier : Dog { public new void Go() { } } Sealing properties? You can also seal seal properties.  You add 'sealed' to the property definition, like so: public sealed override string Name {     get { return m_Name; }     set { m_Name = value; } } In C#, you can only seal a property, not the underlying setters/getters.  This is because C# offers no override syntax for setters or getters.  However, in underlying IL you seal the setter and getter methods individually - a property is just metadata. Why bother sealing? There are a few traditional reasons to seal: Invariance. Other people may want to derive from your class, even though your implementation may make successful derivation near-impossible.  There may be twisted, hacky logic that could never be second-guessed by another developer.  By sealing your class, you're protecting them from wasting their time.  The CLR team has sealed most of the framework classes, and I assume they did this for this reason. Security.  By deriving from your type, an attacker may gain access to functionality that enables him to hack your system.  I consider this a very weak security precaution. Speed.  If a class is sealed, then .NET doesn't need to consult the virtual-function-call table to find the actual type, since it knows that no derived type can exist.  Therefore, it could emit a 'call' instead of 'callvirt' or at least optimise the machine code, thus producing a performance benefit.  But I've done trials, and have been unable to demonstrate this If you have an example, please share! All in all, I'm not convinced that sealing is interesting or important.  Anyway, moving-on... What is automatically sealed? Value types and structs.  If they were not always sealed, all sorts of things would go wrong.  For instance, structs are laid-out inline within a class.  But what if you assigned a substruct to a struct field of that class?  There may be too many fields to fit. Static classes.  Static classes exist in C# but not .NET.  The C# compiler compiles a static class into an 'abstract sealed' class.  So static classes are already sealed in C#. Enumerations.  The CLR does not track the types of enumerations - it treats them as simple value types.  Hence, polymorphism would not work. What cannot be sealed? Interfaces.  Interfaces exist to be implemented, so sealing to prevent implementation is dumb.  But what if you could prevent interfaces from being extended (i.e. ban declarations like "public interface IMyInterface : ISealedInterface")?  There is no good reason to seal an interface like this.  Sealing finalizes behaviour, but interfaces have no intrinsic behaviour to finalize Abstract classes.  In IL you can create an abstract sealed class.  But C# syntax for this already exists - declaring a class as a 'static', so it forces you to declare it as such. Non-override methods.  If a method isn't declared as override it cannot be overridden, so sealing would make no difference.  Note this is stated from a C# perspective - the words are opposite in IL.  In IL, you have four choices in total: no declaration (which actually seals the method), 'virtual' (called 'override' in C#), 'sealed virtual' ('sealed override' in C#) and 'newslot virtual' ('new virtual' or 'virtual' in C#, depending on whether the method already exists in a base class). Methods that implement interface methods.  Methods that implement an interface method must be virtual, so cannot be sealed. Fields.  A field cannot be overridden, only hidden (using the 'new' keyword in C#), so sealing would make no sense.

    Read the article

  • Reference Data Management and Master Data: Are Relation ?

    - by Mala Narasimharajan
    Submitted By:  Rahul Kamath  Oracle Data Relationship Management (DRM) has always been extremely powerful as an Enterprise Master Data Management (MDM) solution that can help manage changes to master data in a way that influences enterprise structure, whether it be mastering chart of accounts to enable financial transformation, or revamping organization structures to drive business transformation and operational efficiencies, or restructuring sales territories to enable equitable distribution of leads to sales teams following the acquisition of new products, or adding additional cost centers to enable fine grain control over expenses. Increasingly, DRM is also being utilized by Oracle customers for reference data management, an emerging solution space that deserves some explanation. What is reference data? How does it relate to Master Data? Reference data is a close cousin of master data. While master data is challenged with problems of unique identification, may be more rapidly changing, requires consensus building across stakeholders and lends structure to business transactions, reference data is simpler, more slowly changing, but has semantic content that is used to categorize or group other information assets – including master data – and gives them contextual value. In fact, the creation of a new master data element may require new reference data to be created. For example, when a European company acquires a US business, chances are that they will now need to adapt their product line taxonomy to include a new category to describe the newly acquired US product line. Further, the cross-border transaction will also result in a revised geo hierarchy. The addition of new products represents changes to master data while changes to product categories and geo hierarchy are examples of reference data changes.1 The following table contains an illustrative list of examples of reference data by type. Reference data types may include types and codes, business taxonomies, complex relationships & cross-domain mappings or standards. Types & Codes Taxonomies Relationships / Mappings Standards Transaction Codes Industry Classification Categories and Codes, e.g., North America Industry Classification System (NAICS) Product / Segment; Product / Geo Calendars (e.g., Gregorian, Fiscal, Manufacturing, Retail, ISO8601) Lookup Tables (e.g., Gender, Marital Status, etc.) Product Categories City à State à Postal Codes Currency Codes (e.g., ISO) Status Codes Sales Territories (e.g., Geo, Industry Verticals, Named Accounts, Federal/State/Local/Defense) Customer / Market Segment; Business Unit / Channel Country Codes (e.g., ISO 3166, UN) Role Codes Market Segments Country Codes / Currency Codes / Financial Accounts Date/Time, Time Zones (e.g., ISO 8601) Domain Values Universal Standard Products and Services Classification (UNSPSC), eCl@ss International Classification of Diseases (ICD) e.g., ICD9 à IC10 mappings Tax Rates Why manage reference data? Reference data carries contextual value and meaning and therefore its use can drive business logic that helps execute a business process, create a desired application behavior or provide meaningful segmentation to analyze transaction data. Further, mapping reference data often requires human judgment. Sample Use Cases of Reference Data Management Healthcare: Diagnostic Codes The reference data challenges in the healthcare industry offer a case in point. Part of being HIPAA compliant requires medical practitioners to transition diagnosis codes from ICD-9 to ICD-10, a medical coding scheme used to classify diseases, signs and symptoms, causes, etc. The transition to ICD-10 has a significant impact on business processes, procedures, contracts, and IT systems. Since both code sets ICD-9 and ICD-10 offer diagnosis codes of very different levels of granularity, human judgment is required to map ICD-9 codes to ICD-10. The process requires collaboration and consensus building among stakeholders much in the same way as does master data management. Moreover, to build reports to understand utilization, frequency and quality of diagnoses, medical practitioners may need to “cross-walk” mappings -- either forward to ICD-10 or backwards to ICD-9 depending upon the reporting time horizon. Spend Management: Product, Service & Supplier Codes Similarly, as an enterprise looks to rationalize suppliers and leverage their spend, conforming supplier codes, as well as product and service codes requires supporting multiple classification schemes that may include industry standards (e.g., UNSPSC, eCl@ss) or enterprise taxonomies. Aberdeen Group estimates that 90% of companies rely on spreadsheets and manual reviews to aggregate, classify and analyze spend data, and that data management activities account for 12-15% of the sourcing cycle and consume 30-50% of a commodity manager’s time. Creating a common map across the extended enterprise to rationalize codes across procurement, accounts payable, general ledger, credit card, procurement card (P-card) as well as ACH and bank systems can cut sourcing costs, improve compliance, lower inventory stock, and free up talent to focus on value added tasks. Change Management: Point of Sales Transaction Codes and Product Codes In the specialty finance industry, enterprises are confronted with usury laws – governed at the state and local level – that regulate financial product innovation as it relates to consumer loans, check cashing and pawn lending. To comply, it is important to demonstrate that transactions booked at the point of sale are posted against valid product codes that were on offer at the time of booking the sale. Since new products are being released at a steady stream, it is important to ensure timely and accurate mapping of point-of-sale transaction codes with the appropriate product and GL codes to comply with the changing regulations. Multi-National Companies: Industry Classification Schemes As companies grow and expand across geographies, a typical challenge they encounter with reference data represents reconciling various versions of industry classification schemes in use across nations. While the United States, Mexico and Canada conform to the North American Industry Classification System (NAICS) standard, European Union countries choose different variants of the NACE industry classification scheme. Multi-national companies must manage the individual national NACE schemes and reconcile the differences across countries. Enterprises must invest in a reference data change management application to address the challenge of distributing reference data changes to downstream applications and assess which applications were impacted by a given change. References 1 Master Data versus Reference Data, Malcolm Chisholm, April 1, 2006.

    Read the article

  • SQL SERVER – Curious Case of Disappearing Rows – ON UPDATE CASCADE and ON DELETE CASCADE – Part 1 of 2

    - by pinaldave
    Social media has created an Always Connected World for us. Recently I enrolled myself to learn new technologies as a student. I had decided to focus on learning and decided not to stay connected on the internet while I am in the learning session. On the second day of the event after the learning was over, I noticed lots of notification from my friend on my various social media handle. He had connected with me on Twitter, Facebook, Google+, LinkedIn, YouTube as well SMS, WhatsApp on the phone, Skype messages and not to forget with a few emails. I right away called him up. The problem was very unique – let us hear the problem in his own words. “Pinal – we are in big trouble we are not able to figure out what is going on. Our product details table is continuously loosing rows. Lots of rows have disappeared since morning and we are unable to find why the rows are getting deleted. We have made sure that there is no DELETE command executed on the table as well. The matter of the fact, we have removed every single place the code which is referencing the table. We have done so many crazy things out of desperation but no luck. The rows are continuously deleted in a random pattern. Do you think we have problems with intrusion or virus?” After describing the problems he had pasted few rants about why I was not available during the day. I think it will be not smart to post those exact words here (due to many reasons). Well, my immediate reaction was to get online with him. His problem was unique to him and his team was all out to fix the issue since morning. As he said he has done quite a lot out in desperation. I started asking questions from audit, policy management and profiling the data. Very soon I realize that I think this problem was not as advanced as it looked. There was no intrusion, SQL Injection or virus issue. Well, long story short first - It was a very simple issue of foreign key created with ON UPDATE CASCADE and ON DELETE CASCADE.  CASCADE allows deletions or updates of key values to cascade through the tables defined to have foreign key relationships that can be traced back to the table on which the modification is performed. ON DELETE CASCADE specifies that if an attempt is made to delete a row with a key referenced by foreign keys in existing rows in other tables, all rows containing those foreign keys are also deleted. ON UPDATE CASCADE specifies that if an attempt is made to update a key value in a row, where the key value is referenced by foreign keys in existing rows in other tables, all of the foreign key values are also updated to the new value specified for the key. (Reference: BOL) In simple words – due to ON DELETE CASCASE whenever is specified when the data from Table A is deleted and if it is referenced in another table using foreign key it will be deleted as well. In my friend’s case, they had two tables, Products and ProductDetails. They had created foreign key referential integrity of the product id between the table. Now the as fall was up they were updating their catalogue. When they were updating the catalogue they were deleting products which are no more available. As the changes were cascading the corresponding rows were also deleted from another table. This is CORRECT. The matter of the fact, there is no error or anything and SQL Server is behaving how it should be behaving. The problem was in the understanding and inappropriate implementations of business logic.  What they needed was Product Master Table, Current Product Catalogue, and Product Order Details History tables. However, they were using only two tables and without proper understanding the relation between them was build using foreign keys. If there were only two table, they should have used soft delete which will not actually delete the record but just hide it from the original product table. This workaround could have got them saved from cascading delete issues. I will be writing a detailed post on the design implications etc in my future post as in above three lines I cannot cover every issue related to designing and it is also not the scope of the blog post. More about designing in future blog posts. Once they learn their mistake, they were happy as there was no intrusion but trust me sometime we are our own enemy and this is a great example of it. In tomorrow’s blog post we will go over their code and workarounds. Feel free to share your opinions, experiences and comments. Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • Source Control and SQL Development &ndash; Part 3

    - by Ajarn Mark Caldwell
    In parts one and two of this series, I have been specifically focusing on the latest version of SQL Source Control by Red Gate Software.  But I have been doing source-controlled SQL development for years, long before this product was available, and well before Microsoft came out with Database Projects for Visual Studio.  “So, how does that work?” you may wonder.  Well, let me share some of the details of how we do it where I work… The key to this approach is that everything is done via Transact-SQL script files; either natively written T-SQL, or generated.  My preference is to write all my code by hand, which forces you to become better at your SQL syntax.  But if you really prefer to use the Management Studio GUI to make database changes, you can still do that, and then you use the Generate Scripts feature of the GUI to produce T-SQL scripts afterwards, and store those in your source control system.  You can generate scripts for things like stored procedures and views by right-clicking on the database in the Object Explorer, and Choosing Tasks, Generate Scripts (see figure 1 to the left).  You can also do that for the CREATE scripts for tables, but that does not work when you have a table that is already in production, and you need to make just a simple change, such as adding a new column or index.  In this case, you can use the GUI to make the table changes, and then instead of clicking the Save button, click the Generate Change Script button (). Then, once you have saved the change script, go ahead and execute it on your development database to actually make the change.  I believe that it is important to actually execute the script rather than just click the Save button because this is your first test that your change script is working and you didn’t somehow lose a portion of the change. As you can imagine, all this generating of scripts can get tedious and tempting to skip entirely, so again, I would encourage you to just get in the habit of writing your own Transact-SQL code, and then it is just a matter of remembering to save your work, just like you are in the habit of saving changes to a Word or Excel document before you exit the program. So, now that you have all of these script files, what do you do with them?  Well, we organize ours into folders labeled ChangeScripts, Functions, Views, and StoredProcedures, and those folders are loaded into our source control system.  ChangeScripts contains all of the table and index changes, and anything else that is basically a one-time-only execution.  Of course you want to write your scripts with qualifying logic so that if a script were accidentally run more than once in a database, it would not crash nor corrupt anything; but these scripts are really intended to be run only once in a database. Once you have your initial set of scripts loaded into source control, then making changes, such as altering a stored procedure becomes a simple matter of checking out your CREATE PROCEDURE* script, editing it in SSMS, saving the change, executing the script in order to effect the change in your database, and then checking the script back in to source control.  Of course, this is where the lack of integration for source control systems within SSMS becomes an irritation, because this means that in addition to SSMS, I also have my source control client application running to do the check-out and check-in.  And when you have 800+ procedures like we do, that can be quite tedious to locate the procedure I want to change in source control, check it out, then locate the script file in my working folder, open it in SSMS, do the change, save it, and the go back to source control to check in.  Granted, it is not nearly as burdensome as, say, losing your source code and having to rebuild it from memory, or losing the audit trail that good source control systems provide.  It is worth the effort, and this is how I have been doing development for the last several years. Remember that everything that the SQL Server Management Studio does in modifying your database can also be done in plain Transact-SQL code, and this is what you are storing.  And now I have shown you how you can do it all without spending any extra money.  You already have source control, or can get free, open-source source control systems (almost seems like an oxymoron, doesn’t it) and of course Management Studio is free with your SQL Server database engine software. So, whether you spend the money on tools to make it easier, or not, you now have no excuse for not using source control with your SQL development. * In our current model, the scripts for stored procedures and similar database objects are written with an IF EXISTS…DROP… at the top, followed by the CREATE PROCEDURE… section, and that followed by a section that assigns permissions.  This allows me to run the same script regardless of whether the procedure previously existed in the database.  If the script was only an ALTER PROCEDURE, then it would fail the first time that procedure was deployed to a database, unless you wrote other code to stub it if it did not exist.  There are a few different ways you could organize your scripts for deployment, each with its own trade-offs, but I think it is absolutely critical that whichever way you organize things, you ensure that the same script is run throughout the deployment cycle, and do not allow customizations to creep in between TEST and PROD.  If you do, then you have broken the integrity of your deployment process because what you deployed to PROD was not exactly the same as what was tested in TEST, so you effectively have now released untested code into PROD.

    Read the article

  • To Bit or Not To Bit

    - by Johnm
    'Twas a long day of troubleshooting and firefighting and now, with most of the office vacant, you face a blank scripting window to create a new table in his database. Many questions circle your mind like dirty water gurgling down the bathtub drain: "How normalized should this table be?", "Should I use an identity column?", "NVarchar or Varchar?", "Should this column be NULLABLE?", "I wonder what apple blue cheese bacon cheesecake tastes like?" Well, there are times when the mind goes it's own direction. A Bit About Bit At some point during your table creation efforts you will encounter the decision of whether to use the bit data type for a column. The bit data type is an integer data type that recognizes only the values of 1, 0 and NULL as valid. This data type is often utilized to store yes/no or true/false values. An example of its use would be a column called [IsGasoline] which would be intended to contain the value of 1 if the row's subject (a car) had a gasoline engine and a 0 if the subject did not have a gasoline engine. The bit data type can even be found in some of the system tables of SQL Server. For example, the sysssispackages table in the msdb database which contains SQL Server Integration Services Package information for the packages stored in SQL Server. This table contains a column called [IsEncrypted]. A value of 1 indicates that the package has been encrypted while the value of 0 indicates that it is not. I have learned that the most effective way to disperse the crowd that surrounds the office coffee machine is to engage into SQL Server debates. The bit data type has been one of the most reoccurring, as well as the most enjoyable, of these topics. It contains a practical side and a philosophical side. Practical Consideration This data type certainly has its place and is a valuable option for database design; but it is often used in situations where the answer is really not a pure true/false response. In addition, true/false values are not very informative or scalable. Let's use the previously noted [IsGasoline] column for illustration. While on the surface it appears to be a rather simple question when evaluating a car: "Does the car have a gasoline engine?" If the person entering data is entering a row for a Jeep Liberty, the response would be a 1 since it has a gasoline engine. If the person is entering data is entering a row for a Chevrolet Volt, the response would be a 0 since it is an electric engine. What happens when a person is entering a row for the gasoline/electric hybrid Toyota Prius? Would one person's conclusion be consistent with another person's conclusion? The argument could be made that the current intent for the database is to be used only for pure gasoline and pure electric engines; but this is where the scalability issue comes into play. With the use of a bit data type a database modification and data conversion would be required if the business decided to take on hybrid engines. Whereas, alternatively, if the int data type were used as a foreign key to a reference table containing the engine type options, the change to include the hybrid option would only require an entry into the reference table. Philosophical Consideration Since the bit data type is often used for true/false or yes/no data (also called Boolean) it presents a philosophical conundrum of what to do about the allowance of the NULL value. The inclusion of NULL in a true/false or yes/no response simply violates the logical principle of bivalence which states that "every proposition is either true or false". If NULL is not true, then it must be false. The mathematical laws of Boolean logic support this concept by stating that the only valid values of this scenario are 1 and 0. There is another way to look at this conundrum: NULL is also considered to be the absence of a response. In other words, it is the equivalent to "undecided". Anyone who watches the news can tell you that polls always include an "undecided" option. This could be considered a valid option in the world of yes/no/dunno. Through out all of these considerations I have discovered one absolute certainty: When you have found a person, or group of persons, who are willing to entertain a philosophical debate of the bit data type, you have found some true friends.

    Read the article

  • Content Query Web Part and the Yes/No Field

    - by Bil Simser
    The Content Query Web Part (CQWP) is a pretty powerful beast. It allows you to do multiple site queries and aggregate the results. This is great for rolling up content and doing some summary type reporting. Here’s a trick to remember about Yes/No fields and using the CQWP. If you’re building a news style site and want to aggregate say all the announcements that people tag a certain way, up onto the home page this might be a solution. First we need to allow a way for users of all our sites to mark an announcement for inclusion on our Intranet Home Page. We’ll do this by just modifying the Announcement Content type and adding a Yes/No field to it. There are alternate ways of doing this like building a new Announcement type or stapling a feature to all sites to add our column but this is pretty low impact and only affects our current site collection so let’s go with it for now, okay? You can berate me in the comments about the proper way I should have done this part. Go to the Site Settings for the Site Collection and click on Site Content Types under the Galleries. This takes you to the gallery for this site and all subsites. Scroll down until you see the List Content Types and click on Announcements. Now we’re modifying the Announcement content type which affects all those announcement lists that are created by default if you’re building sites using the Team Site template (or creating a new Announcements list on any site for that matter). Click on Add from new site column under the Column list. This will allow us to create a new Yes/No field that users will see in Announcement items. This field will allow the user to flag the announcement for inclusion on the home page. Feel free to modify the fields as you see fit for your environment, this is just an example. Now that we’ve added the column to our Announcements Content type we can go into any site that has an announcement list, modify that announcement and flag it to be included on our home page. See the new Featured column? That was the result of modifying our Announcements Content Type on this site collection. Now we can move onto the dirty part, displaying it in a CQWP on the home page. And here is where the fun begins (and the head scratching should end). On our home page we want to drop a Content Query Web Part and aggregate any Announcement that’s been flagged as Featured by the users (we could also add the filter to handle Expires so we don’t show old content so go ahead and do that if you want). First add a CQWP to the page then modify the settings for the web part. In the first section, Query, we want the List Type to be set to Announcements and the Content type to be Announcement so set your options like this: Click Apply and you’ll see the results display all Announcements from any site in the site collection. I have five team sites created each with a unique announcement added to them. Now comes the filtering. We don’t want to include every announcement, only ones users flag using that Featured column we added. At first blush you might scroll down to the Additional Filters part of the Query options and set the Featured column to be equal to Yes: This seems correct doesn’t it? After all, the column is a Yes/No column and looking at an announcement in the site, it displays the field as Yes or No: However after applying the filter you get this result: (I have the announcements from Team Site 1 and Team Site 4 flagged as Featured) Huh? It’s BACKWARDS! Let’s confirm that. Go back in and change the Additional Filters section from Yes to No and hit Apply and you get this: Wait a minute? Shouldn’t I see Team Site 1 and 4 if the logic is backwards? Why am I seeing the same thing as before. What gives… For whatever reason, unknown to me, a Yes/No field (even though it displays as such) really uses 1 and 0 behind the scenes. Yeah, someone was stuck on using integer values for booleans when they wrote SharePoint (probably after a long night of white boarding ways to mess with developers heads) and came up with this. The solution is pretty simple but not very discoverable. Set the filter to include your flagged items like so: And it will filter the items marked as Featured correctly giving you this result: This kind of solution could also be extended and enhanced. Here are a few suggestions and ideas: Modify the ItemStyle.xsl file to add a new style for this aggregation which would include the first few paragraphs of the body (or perhaps add another field to the Content type called Excerpt or Summary and display that instead) Add an Image column to the Announcement Content type to include a Picture field and display it in the summary Add a Category choice field (Employee News, Current Events, Headlines, etc.) and add multiple CQWPs to the home page filtering each one on a different category I know some may find this topic old and dusty but I didn’t see a lot out there specifically on filtering the Yes/No fields and the whole 1/0 trick was a little wonky, so I figured a few pictures would help walk through overcoming yet another SharePoint weirdness. With a little work and some creative juices you can easily us the power of aggregation and the CQWP to build a news site from content on your team sites.

    Read the article

  • Oracle CRM On Demand Release 24 is Generally Available

    - by Richard Lefebvre
    Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 We are pleased to announce that Oracle CRM On Demand Release 24 is Generally Available as of October 25, 2013 Get smarter, more productive and the best value with Oracle CRM On Demand Release 24. Oracle CRM On Demand continues to be the most complete Software-as-a-Service (SaaS) CRM solution available. Now, with Release 24, organizations of all types and sizes benefit from actionable insight anywhere, anytime, as well as key enhancements in mobility, embedded social, analytics, integration and extensibility, and ease of use.Next Generation Mobile and Desktop Solutions : Oracle CRM On Demand Release 24 offers a complete set of mobile and desktop solutions that improve productivity by enabling reps to access and update information anywhere, anytime. Capabilities include: Oracle CRM On Demand Disconnected Mobile Sales (DMS) – A disconnected native iPad solution, DMS has been further streamlined mobile sales process by adding Structured Product Messaging to record brand specific call objectives, enhancements in HTML5 eDetailing including message response tracking and improvements in administration and configuration such as more field management options for read only fields, role management and enhanced logging. Oracle CRM On Demand Connected Mobile Sales. This add-on mobile service provides a configurable mobile solution on iOS, BlackBerry and now Android devices. You can access data from CRM On Demand in real time with a rich, native user experience, that is comfortable and familiar to current iOS, BlackBerry and Android users. New features also include Single Sign On to enhance security for mobile users.  Oracle CRM On Demand Desktop: This application centralizes essential CRM information in the familiar Microsoft Outlook environment,increasing user adoption and decreasing training costs. Users can manage CRM data while disconnected, then synchronize bi-directionally when they are back on the network. New in Oracle CRM On Demand Desktop Version 3 is the ability to synchronize by Books of Business, and improved Online Lookup. Mobile Browser Support: The following mobile device browsers are now supported: Apple iPhone, Apple iPad, Windows 8 Tablets, and Google Android. Leverage the Social Enterprise Engaging customers via social channels is rapidly becoming a significant key to enhanced customer experience as it provides proactive customer service, targeted messaging and greater intimacy throughout the entire customer lifecycle. Listening to customers on the social channels can identify a customers’ sphere of influence and the real value they bring to their organization, or the impact they can have on the opportunity. Servicing the customer’s need is the first step towards loyalty to a brand, integrating with social channels allows us to maximize brand affinity and virally expand customer engagements thus increasing revenue. Oracle CRM On Demand is leveraging the Social Enterprise through its integration with Oracle’s Social Relationship Management (SRM) product suite by providing out-of-the-box integration with Social Engagement and Monitoring (SEM), Social Marketing (SM) and Oracle Social Network (OSN). With Oracle CRM On Demand Release 24, users are able to create a service request from a social post via SEM and have leads entered on a SM lead form automatically entered into Oracle CRM On Demand along with the campaign, streamlining the lead qualification process. Get Smarter with Actionable Insight The difference between making good decisions and great decisions depends heavily upon the quality, structure, and availability of information at hand. Oracle CRM On Demand Release 24 expands upon its industry-leading analytics capabilities to provide greater business insight than ever before. New capabilities include flexible permissions on analytics reports folders, allowing for read only access to reports, and additional field and object coverage. Get More Productive with Powerful Tools Oracle CRM On Demand Release 24 introduces a new set of powerful capabilities designed to maximize productivity. A significant new feature for customizing Oracle CRM On Demand is a JavaScript API. The JS API allows customers to add new buttons, suppress existing buttons and even change what happens when a user clicks an existing button. Other usability enhancements, such as personalized related information applets, extended case insensitive search provide users with better, more intuitive, experience. Additional privileges for viewing private activities and notes allow administrators to reassign records as needed, and Custom Object management. Workflow has been added to the Order Item object; and now tasks can be assigned to a relative user, such as an Account Owner, allowing more complex business processes to be automated and adhered to. Get the Best Value Oracle CRM On Demand delivers unprecedented value with the broadest set of capabilities from a single-provider solution, the industry’s lowest total cost of ownership, the most on-demand deployment options, the deepest CRM expertise and experience of any CRM provider, and the most secure CRM in the cloud. With Release 24, Oracle CRM On Demand now includes even more enterprise-grade security, integration, and extensibility features, along with enhanced industry editions to save you time and money. New features include: Business Process Administration: A new privilege has been added that allows administrators to override a Business Process Administration rule.This privilege permits users to edit a locked record, or unlock a record, in the event of a material change that needs to be reflected per corporatepolicy. Additionally, the Products Detailed object has been added to Business Process Administration, enabling record locking and logic to be applied. Expanded Integration: Oracle continues to improve Web Services each release, by adding more object coverage enabling customers and partners to easily integrate with CRM On Demand. Bottom Line Oracle CRM On Demand Release 24 enables organizations to get smarter, get more productive, and get the best value, period. For more information on Oracle CRM On Demand Release 24, please visit oracle.com/crmondemand

    Read the article

  • Platformer Starter Kit - Collision Issues

    - by Cyral
    I'm having trouble with my game that is based off the XNA Platformer starter kit. My game uses smaller tiles (16x16) then the original (32x40) which I'm thinking may be having an effect on collision (Being it needs to be more precise). Standing on the edge of a tile and jumping causes the player to move off the the tile when he lands. And 80% of the time, when the player lands, he goes flying though SOLID tiles in a diagonal fashion. This is very annoying as it is almost impossible to test other features, when spawning and jumping will result in the player landing in another part of the level or falling off the edge completely. The code is as follows: /// <summary> /// Updates the player's velocity and position based on input, gravity, etc. /// </summary> public void ApplyPhysics(GameTime gameTime) { float elapsed = (float)gameTime.ElapsedGameTime.TotalSeconds; Vector2 previousPosition = Position; // Base velocity is a combination of horizontal movement control and // acceleration downward due to gravity. velocity.X += movement * MoveAcceleration * elapsed; velocity.Y = MathHelper.Clamp(velocity.Y + GravityAcceleration * elapsed, -MaxFallSpeed, MaxFallSpeed); velocity.Y = DoJump(velocity.Y, gameTime); // Apply pseudo-drag horizontally. if (IsOnGround) velocity.X *= GroundDragFactor; else velocity.X *= AirDragFactor; // Prevent the player from running faster than his top speed. velocity.X = MathHelper.Clamp(velocity.X, -MaxMoveSpeed, MaxMoveSpeed); // Apply velocity. Position += velocity * elapsed; Position = new Vector2((float)Math.Round(Position.X), (float)Math.Round(Position.Y)); // If the player is now colliding with the level, separate them. HandleCollisions(); // If the collision stopped us from moving, reset the velocity to zero. if (Position.X == previousPosition.X) velocity.X = 0; if (Position.Y == previousPosition.Y) velocity.Y = 0; } /// <summary> /// Detects and resolves all collisions between the player and his neighboring /// tiles. When a collision is detected, the player is pushed away along one /// axis to prevent overlapping. There is some special logic for the Y axis to /// handle platforms which behave differently depending on direction of movement. /// </summary> private void HandleCollisions() { // Get the player's bounding rectangle and find neighboring tiles. Rectangle bounds = BoundingRectangle; int leftTile = (int)Math.Floor((float)bounds.Left / Tile.Width); int rightTile = (int)Math.Ceiling(((float)bounds.Right / Tile.Width)) - 1; int topTile = (int)Math.Floor((float)bounds.Top / Tile.Height); int bottomTile = (int)Math.Ceiling(((float)bounds.Bottom / Tile.Height)) - 1; // Reset flag to search for ground collision. isOnGround = false; // For each potentially colliding tile, for (int y = topTile; y <= bottomTile; ++y) { for (int x = leftTile; x <= rightTile; ++x) { // If this tile is collidable, ItemCollision collision = Level.GetCollision(x, y); if (collision != ItemCollision.Passable) { // Determine collision depth (with direction) and magnitude. Rectangle tileBounds = Level.GetBounds(x, y); Vector2 depth = RectangleExtensions.GetIntersectionDepth(bounds, tileBounds); if (depth != Vector2.Zero) { float absDepthX = Math.Abs(depth.X); float absDepthY = Math.Abs(depth.Y); // Resolve the collision along the shallow axis. if (absDepthY < absDepthX || collision == ItemCollision.Platform) { // If we crossed the top of a tile, we are on the ground. if (previousBottom <= tileBounds.Top) isOnGround = true; // Ignore platforms, unless we are on the ground. if (collision == ItemCollision.Impassable || IsOnGround) { // Resolve the collision along the Y axis. Position = new Vector2(Position.X, Position.Y + depth.Y); // Perform further collisions with the new bounds. bounds = BoundingRectangle; } } else if (collision == ItemCollision.Impassable) // Ignore platforms. { // Resolve the collision along the X axis. Position = new Vector2(Position.X + depth.X, Position.Y); // Perform further collisions with the new bounds. bounds = BoundingRectangle; } } } } } // Save the new bounds bottom. previousBottom = bounds.Bottom; } It also tends to jitter a little bit sometimes, I'm solved some of this with some fixes I found here on stackexchange, But Ive only seen one other case of the flying though blocks problem. This question seems to have a similar problem in the video, but mine is more crazy. Again this is a very annoying bug! Any help would be greatly appreciated! EDIT: Speed stuff // Constants for controling horizontal movement private const float MoveAcceleration = 13000.0f; private const float MaxMoveSpeed = 1750.0f; private const float GroundDragFactor = 0.48f; private const float AirDragFactor = 0.58f; // Constants for controlling vertical movement private const float MaxJumpTime = 0.35f; private const float JumpLaunchVelocity = -3500.0f; private const float GravityAcceleration = 3400.0f; private const float MaxFallSpeed = 550.0f; private const float JumpControlPower = 0.14f;

    Read the article

  • Memento with optional state?

    - by Korey Hinton
    EDIT: As pointed out by Steve Evers and pdr, I am not correctly implementing the Memento pattern, my design is actually State pattern. Menu Program I built a console-based menu program with multiple levels that selects a particular test to run. Each level more precisely describes the operation. At any level you can type back to go back one level (memento). Level 1: Server Type? [1] Server A [2] Server B Level 2: Server environment? [1] test [2] production Level 3: Test type? [1] load [2] unit Level 4: Data Collection? [1] Legal docs [2] Corporate docs Level 4.5 (optional): Load Test Type [2] Multi TIF [2] Single PDF Level 5: Command Type? [1] Move [2] Copy [3] Remove [4] Custom Level 6: Enter a keyword [setup, cleanup, run] Design States PROBLEM: Right now the STATES enum is the determining factor as to what state is BACK and what state is NEXT yet it knows nothing about what the current memento state is. Has anyone experienced a similar issue and found an effective way to handle mementos with optional state? static enum STATES { SERVER, ENVIRONMENT, TEST_TYPE, COLLECTION, COMMAND_TYPE, KEYWORD, FINISHED } Possible Solution (Not-flexible) In reference to my code below, every case statement in the Menu class could check the state of currentMemo and then set the STATE (enum) accordingly to pass to the Builder. However, this doesn't seem flexible very flexible to change and I'm struggling to see an effective way refactor the design. class Menu extends StateConscious { private State state; private Scanner reader; private ServerUtils utility; Menu() { state = new State(); reader = new Scanner(System.in); utility = new ServerUtils(); } // Recurring menu logic public void startPromptingLoop() { List<State> states = new ArrayList<>(); states.add(new State()); boolean redoInput = false; boolean userIsDone = false; while (true) { // get Memento from last loop Memento currentMemento = states.get(states.size() - 1) .saveMemento(); if (currentMemento == null) currentMemento = new Memento.Builder(0).build(); if (!redoInput) System.out.println(currentMemento.prompt); redoInput = false; // prepare Memento for next loop Memento nextMemento = null; STATES state = STATES.values()[states.size() - 1]; // get user input String selection = reader.nextLine(); switch (selection) { case "exit": reader.close(); return; // only escape case "quit": nextMemento = new Memento.Builder(first(), currentMemento, selection).build(); states.clear(); break; case "back": nextMemento = new Memento.Builder(previous(state), currentMemento, selection).build(); if (states.size() <= 1) { states.remove(0); } else { states.remove(states.size() - 1); states.remove(states.size() - 1); } break; case "1": nextMemento = new Memento.Builder(next(state), currentMemento, selection).build(); break; case "2": nextMemento = new Memento.Builder(next(state), currentMemento, selection).build(); break; case "3": nextMemento = new Memento.Builder(next(state), currentMemento, selection).build(); break; case "4": nextMemento = new Memento.Builder(next(state), currentMemento, selection).build(); break; default: if (state.equals(STATES.CATEGORY)) { String command = selection; System.out.println("Executing " + command + " command on: " + currentMemento.type + " " + currentMemento.environment); utility.executeCommand(currentMemento.nickname, command); userIsDone = true; states.clear(); nextMemento = new Memento.Builder(first(), currentMemento, selection).build(); } else if (state.equals(STATES.KEYWORD)) { nextMemento = new Memento.Builder(next(state), currentMemento, selection).build(); states.clear(); nextMemento = new Memento.Builder(first(), currentMemento, selection).build(); } else { redoInput = true; System.out.println("give it another try"); continue; } break; } if (userIsDone) { // start the recurring menu over from the beginning for (int i = 0; i < states.size(); i++) { if (i != 0) { states.remove(i); // remove all except first } } reader = new Scanner(System.in); this.state = new State(); userIsDone = false; } if (!redoInput) { this.state.restoreMemento(nextMemento); states.add(this.state); } } } }

    Read the article

  • About Solaris 11 and UltraSPARC II/III/IV/IV+

    - by nospam(at)example.com (Joerg Moellenkamp)
    I know that I will get the usual amount of comments like "Oh, Jörg ? you can't be negative about Oracle" for this article. However as usual I want to explain the logic behind my reasoning. Yes ? I know that there is a lot of UltraSPARC III, IV and IV+ gear out there. But there are some very basic questions: Does your application you are currently running on this gear stops running just because you can't run Solaris 11 on it? What is the need to upgrade a system already in production to Solaris 11? I have the impression, that some people think that the systems get useless in the moment Oracle releases Solaris 11. I know that Sun sold UltraSPARC IV+ systems until 2009. The Sun SF490 introduced 2004 for example, that was a Sun SF480 with UltraSPARC IV and later with UltraSPARC IV+. And yes, Sun made some speedbumps. At that time the systems of the UltraSPARC III to IV+ generations were supported on Solaris 8, on Solaris 9 and on Solaris 10. However from my perspective we sold them to customers, which weren't able to migrate to Solaris 10 because they used applications not supported on Solaris 9 or who just didn't wanted to migrate to Solaris 10. Believe it or not ? I personally know two customers that migrated core systems to Solaris 10 in ? well 2008/9. This was especially true when the M3000 was announced in 2008 when it closed the darned single socket gap. It may be different at you site, however that's what I remember about that time when talking with customers. At first: Just because there is no Solaris 11 for UltraSPARC III, IV and IV+, it doesn't mean that Solaris 10 will go away anytime soon. I just want to point you to "Expect Lifetime Support - Hardware and Operating Systems". It states about Premier Support:Maintenance and software upgrades are included for Oracle operating systems and Oracle VM for a minimum of eight years from the general availability date.GA for Solaris 10 was in 2005. Plus 8 years ? 2013 ? at minimum. Then you can still opt for 3 years of "Extended Support" ? 2016 ? at minimum. 2016 your systems purchased in 2009 are 7 years old. Even on systems purchased at the very end of the lifetime of that system generation. That are the rules as written in the linked document. I said minimum The actual dates are even further in the future: Premier Support for Solaris 10 ends in 2015, Extended support ends 2018. Sustaining support ? indefinite. You will find this in the document "Oracle Lifetime Support Policy: Oracle Hardware and Operating Systems".So I don't understand when some people write, that Oracle is less protective about hardware investments than Sun. And for hardware it's the same as with Sun: Service 5 years after EOL as part of Premier Support. I would like to write about a different perspective as well: I have to be a little cautious here, because this is going in the roadmap area, so I will mention the public sources here: John Fowler told last year that we have to expect at at least 3x the single thread performance of T3 for T4. We have 8 cores in T4, as stated by Rick Hetherington. Let's assume for a moment that a T4 core will have the performance of a UltraSPARC core (just to simplify math and not to disclosing anything about the performance, all existing SPARC cores are considered equal). So given this pieces of information, you could consolidate 8 V215, 4 or 8 V245, 2 full blown V445,2 full blown 490, 2 full blown M3000 on a single T4 SPARC processor. The Fowler roadmap prezo talked about 4-socket systems with T4. So 32 V215, 16 to 8 V245, 8 fullblown V445, 8 full blown V490, 8 full blown M3000 in a system image. I think you get the idea. That said, most of the systems we are talking about have already amortized and perhaps it's just time to invest in new systems to yield other advantages like reduced space consumptions, like reduced power consumption, like some of the neat features sun4v gives you, and yes ? reduced number of processor licenses for Oracle and less money for Oracle HW/SW support. As much as I dislike it myself that my own UltraSPARC III and UltraSPARC II based systems won't run on Solaris 11 (and I have quite a few of them in my personal lab), I really think that the impact on production environments will be much less than most people think now. By the way: The reason for this move is a quite significant new feature. I will tell you that it was this feature, when it's out. I assume, telling just a word more could lead to much more time to blog.

    Read the article

  • .NET development on a Retina MacBook Pro with Windows 8

    - by Jeff
    I remember sitting in Building 5 at Microsoft with some of my coworkers, when one of them came in with a shiny new 11” MacBook Air. It was nearly two years ago, and we found it pretty odd that the OEM’s building Windows machines sucked at industrial design in a way that defied logic. While Dell and HP were in a race to the bottom building commodity crap, Apple was staying out of the low-end market completely, and focusing on better design. In the process, they managed to build machines people actually wanted, and maintain an insanely high margin in the process. I stopped buying the commodity crap and custom builds in 2006, when Apple went Intel. As a .NET guy, I was still in it for Microsoft’s stack of development tools, which I found awesome, but had back to back crappy laptops from HP and Dell. After that original 15” MacBook Pro, I also had a Mac Pro tower (that I sold after three years for $1,500!), a 27” iMac, and my favorite, a 17” MacBook Pro (the unibody style) with an SSD added from OWC. The 17” was a little much to carry around because it was heavy, but it sure was nice getting as much as eight hours of battery life, and the screen was amazing. When the rumors started about a 15” model with a “retina” screen inspired by the Air, I made up my mind I wanted one, and ordered it the day it came out. I sold my 17”, after three years, for $750 to a friend who is really enjoying it. I got the base model with the upgrade to 16 gigs of RAM. It feels solid for being so thin, and if you’ve used the third generation iPad or the newer iPhone, you’ll be just as thrilled with the screen resolution. I’m typically getting just over six hours of battery life while running a VM, but Parallels 8 allegedly makes some power improvements, so we’ll see what happens. (It was just released today.) The nice thing about VM’s are that you can run more than one at a time. Primarily I run the Windows 8 VM with four cores (the laptop is quad-core, but has 8 logical cores due to hyperthreading or whatever Intel calls it) and 8 gigs of RAM. I also have a Windows Server 2008 R2 VM I spin up when I need to test stuff in a “real” server environment, and I give it two cores and 4 gigs of RAM. The Windows 8 VM spins up in about 8 seconds. Visual Studio 2012 takes a few more seconds, but count part of that as the “ReSharper tax” as it does its startup magic. The real beauty, the thing I looked most forward to, is that beautifully crisp C# text. Consolas has never looked as good as it does at 10pt. as it does on this display. You know how it looks great at 80pt. when conference speakers demo stuff on a projector? Think that sharpness, only tiny. It’s just gorgeous. Beyond that, everything is just so responsive and fast. Builds of large projects happen in seconds, hundreds of unit tests run in seconds… you just don’t spend a lot of time waiting for stuff. It’s kind of painful to go back to my 27” iMac (which would be better if I put an SSD in it before its third birthday). Are there negatives? A few minor issues, yes. As is the case with OS X, not everything scales right. You’ll see some weirdness at times with splash screens and icons and such. Chrome’s text rendering (in Windows) is apparently not aware of how to deal with higher DPI’s, so text is fuzzy (the OS X version is super sharp, however). You’ll also have to do some fiddling with keyboard settings to use the Windows 8 keyboard shortcuts. Overall, it’s as close to a no-compromise development experience as I’ve ever had. I’m not even going to bother with Boot Camp because the VM route already exceeds my expectations. You definitely get what you pay for. If this one also lasts three years and I can turn around and sell it, it’s worth it for something I use every day.

    Read the article

  • How customers view and interact with a company

    The Harvard Business Review article written by Rayport and Jaworski is aptly titled “Best Face Forward” because it sheds light on how customers view and interact with a company. In the past most business interaction between customers was performed in a face to face meeting where one party would present an item for sale and then the other would decide whether to purchase the item. In addition, if there was a problem with a purchased item then they would bring the item back to the person who sold the item for resolution. One of my earliest examples of witnessing this was when I was around 6 or 7 years old and I was allowed to spend the summer in Tennessee with my Grandparents. My Grandfather had just written a book about the local history of his town and was selling them to his friends and local bookstores. I still remember he offered to pay me a small commission for every book I helped him sell because I was carrying the books around for him. Every sale he made was face to face with his customers which allowed him to share his excitement for the book with everyone. In today’s modern world there is less and less human interaction as the use of computers and other technologies allow us to communicate within seconds even though both parties may be across the globe or just next door. That being said, customers view a company through multiple access points called faces that represent the ability to interact without actually seeing a human face. As a software engineer this is a good and a bad thing because direct human interaction and technology based interaction have both good and bad attributes based on the customer. How organizations coordinate business and IT functions, to provide quality service varies based on each individual business and the goals and directives put in place by its management. According to Rayport and Jaworski, the type of interaction used through a particular access point may lend itself to be people-dominate, machine-dominate, or a combination of both. The method by which a company communicates information through an access point is a strategic choice that relates costs and customer outcomes. To simplify this, the choice is based on what can give the customer the best experience interacting with the company when the cost of the interaction is also a factor. I personally see examples of this every day at work. The company website is machine-dominate with people updating and maintaining information, our groups department is people dominate because most of the customer interaction is done at the customers location and is backed up by machine based data sources, and our sales/member service department is a hybrid because employees work in tandem with machines in order for them to assist customers with signing up or any other issue they may have. The positive and negative aspects of human and machine interfaces are a key aspect in deciding which interface to use when allowing customers to access a company or a combination of the two. Rayport and Jaworski also used MIT professor Erik Brynjolfsson preliminary catalog of human and machine strengths. He stated that humans outperform machines in judgment, pattern recognition, exception processing, insight, and creativity. I have found this to be true based on the example of how sales and member service reps at my company handle a multitude of questions and various situations with a lot of unknown variables. A machine interface could never effectively be able to handle these scenarios because there are too many variables to consider and would not have the built-in logic to process each customer’s claims and needs. In addition, he also stated that machines outperform humans in collecting, storing, transmitting and routine processing. An example of this would be my employer’s website. Customers can simply go online and purchase a product without even talking to a sales or member services representative. The information is then stored in a database so that the customer can always go back and review there order, and access their selected services. A human, no matter how smart they are would never be able to keep track of hundreds of thousands of customers let alone know what they purchased or how much they paid. In today’s technology driven economy every company must offer their customers multiple methods of accessibly in order to survive. The more of an opportunity a company has to create a positive experience for their customers, in my opinion, they more likely the customer will return to that company again. I have noticed this with my personal shopping habits and experiences. References Rayport, J., & Jaworski, B. (2004). Best Face Forward. Harvard Business Review, 82(12), 47-58. Retrieved from Business Source Complete database.

    Read the article

  • A Basic Thread

    - by Joe Mayo
    Most of the programs written are single-threaded, meaning that they run on the main execution thread. For various reasons such as performance, scalability, and/or responsiveness additional threads can be useful. .NET has extensive threading support, from the basic threads introduced in v1.0 to the Task Parallel Library (TPL) introduced in v4.0. To get started with threads, it's helpful to begin with the basics; starting a Thread. Why Do I Care? The scenario I'll use for needing to use a thread is writing to a file.  Sometimes, writing to a file takes a while and you don't want your user interface to lock up until the file write is done. In other words, you want the application to be responsive to the user. How Would I Go About It? The solution is to launch a new thread that performs the file write, allowing the main thread to return to the user right away.  Whenever the file writing thread completes, it will let the user know.  In the meantime, the user is free to interact with the program for other tasks. The following examples demonstrate how to do this. Show Me the Code? The code we'll use to work with threads is in the System.Threading namespace, so you'll need the following using directive at the top of the file: using System.Threading; When you run code on a thread, the code is specified via a method.  Here's the code that will execute on the thread: private static void WriteFile() { Thread.Sleep(1000); Console.WriteLine("File Written."); } The call to Thread.Sleep(1000) delays thread execution. The parameter is specified in milliseconds, and 1000 means that this will cause the program to sleep for approximately 1 second.  This method happens to be static, but that's just part of this example, which you'll see is launched from the static Main method.  A thread could be instance or static.  Notice that the method does not have parameters and does not have a return type. As you know, the way to refer to a method is via a delegate.  There is a delegate named ThreadStart in System.Threading that refers to a method without parameters or return type, shown below: ThreadStart fileWriterHandlerDelegate = new ThreadStart(WriteFile); I'll show you the whole program below, but the ThreadStart instance above goes in the Main method. The thread uses the ThreadStart instance, fileWriterHandlerDelegate, to specify the method to execute on the thread: Thread fileWriter = new Thread(fileWriterHandlerDelegate); As shown above, the argument type for the Thread constructor is the ThreadStart delegate type. The fileWriterHandlerDelegate argument is an instance of the ThreadStart delegate type. This creates an instance of a thread and what code will execute, but the new thread instance, fileWriter, isn't running yet. You have to explicitly start it, like this: fileWriter.Start(); Now, the code in the WriteFile method is executing on a separate thread. Meanwhile, the main thread that started the fileWriter thread continues on it's own.  You have two threads running at the same time. Okay, I'm Starting to Get Glassy Eyed. How Does it All Fit Together? The example below is the whole program, pulling all the previous bits together. It's followed by its output and an explanation. using System; using System.Threading; namespace BasicThread { class Program { static void Main() { ThreadStart fileWriterHandlerDelegate = new ThreadStart(WriteFile); Thread fileWriter = new Thread(fileWriterHandlerDelegate); Console.WriteLine("Starting FileWriter"); fileWriter.Start(); Console.WriteLine("Called FileWriter"); Console.ReadKey(); } private static void WriteFile() { Thread.Sleep(1000); Console.WriteLine("File Written"); } } } And here's the output: Starting FileWriter Called FileWriter File Written So, Why are the Printouts Backwards? The output above corresponds to Console.Writeline statements in the program, with the second and third seemingly reversed. In a single-threaded program, "File Written" would print before "Called FileWriter". However, this is a multi-threaded (2 or more threads) program.  In multi-threading, you can't make any assumptions about when a given thread will run.  In this case, I added the Sleep statement to the WriteFile method to greatly increase the chances that the message from the main thread will print first. Without the Thread.Sleep, you could run this on a system with multiple cores and/or multiple processors and potentially get different results each time. Interesting Tangent but What Should I Get Out of All This? Going back to the main point, launching the WriteFile method on a separate thread made the program more responsive.  The file writing logic ran for a while, but the main thread returned to the user, as demonstrated by the print out of "Called FileWriter".  When the file write finished, it let the user know via another print statement. This was a very efficient use of CPU resources that made for a more pleasant user experience. Joe

    Read the article

  • Accelerating 2d object collision with other objects [on hold]

    - by Silent Cave
    Making my very first attempt at game programming with SDL/OpenGL. So I made an object Actor witch can move in all four sides with acceleration. And there are bunch of other rectangles to collide to. the image Movement and collision detection alghorythms work just fine by itself, but when combined to prevent the green rectangle from crossing black rectangles, it gives me a kind of funny resault. Let me show you the code first: from Actor.h class Actor{ public: SDL_Rect * dim; alphaColor * col; float speed; float xlGrav, xrGrav, yuGrav, ydGrav; float acceleration; bool left,right,up,down; Actor(SDL_Rect * dim,alphaColor * col, float speed, float acceleration); bool colides(const SDL_Rect & rect); bool check_for_collisions(const std::vector<SDL_Rect*> & gameObjects ); }; from actor.cpp bool Actor::colides(const SDL_Rect & rect){ if (dim->x + dim->w < rect.x) return false; if (dim->x > rect.x + rect.w) return false; if (dim->y + dim->h < rect.y) return false; if (dim->y > rect.y + rect.h) return false; return true; } movement logic from main.cpp if (actor->left){ if(actor->xlGrav < actor->speed){ actor->xlGrav += actor->speed*actor->acceleration; }else actor->xlGrav = actor->speed; actor->dim->x -= actor->xlGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->x += actor->xlGrav; actor->xlGrav = 0; } } if (!actor->left){ if(actor->xlGrav - actor->speed*actor->acceleration > 0){ actor->xlGrav -= actor->speed*actor->acceleration; }else actor->xlGrav = 0; actor->dim->x -= actor->xlGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->x += actor->xlGrav; actor->xlGrav = 0; } } if (actor->right){ if(actor->xrGrav < actor->speed){ actor->xrGrav += actor->speed*actor->acceleration; }else actor->xrGrav = actor->speed; actor->dim->x += actor->xrGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->x -= actor->xrGrav; actor->xrGrav = 0; } } if (!actor->right){ if(actor->xrGrav - actor->speed*actor->acceleration > 0){ actor->xrGrav -= actor->speed*actor->acceleration; }else actor->xrGrav = 0; actor->dim->x += actor->xrGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->x -= actor->xrGrav; actor->xrGrav = 0; } } if (actor->up){ if(actor->yuGrav < actor->speed){ actor->yuGrav += actor->speed*actor->acceleration; }else actor->yuGrav = actor->speed; actor->dim->y -= actor->yuGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->y += actor->yuGrav; actor->yuGrav = 0; } } if (!actor->up){ if(actor->yuGrav - actor->speed*actor->acceleration > 0){ actor->yuGrav -= actor->speed*actor->acceleration; }else actor->yuGrav = 0; actor->dim->y -= actor->yuGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->y += actor->yuGrav; actor->yuGrav = 0; } } if (actor->down){ if(actor->ydGrav < actor->speed){ actor->ydGrav += actor->speed*actor->acceleration; }else actor->ydGrav = actor->speed; actor->dim->y += actor->ydGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->y -= actor->ydGrav; actor->ydGrav = 0; } } if (!actor->down){ if(actor->ydGrav - actor->speed*actor->acceleration > 0){ actor->ydGrav -= actor->speed*actor->acceleration; }else actor->ydGrav = 0; actor->dim->y += actor->ydGrav; if(actor->check_for_collisions(gameObjects)){ actor->dim->y -= actor->ydGrav; actor->ydGrav = 0; } } So, if the green box approaches an obstacle from up or left, everything goes as planned - object stops, and it's acceleration drops to zero. But if it comes from bottom or right, it enters into obstacles inner space and starts strangely dance, I'd rather say move in inverted controls. What do I fail to see?

    Read the article

  • Think before you animate

    - by David Paquette
    Animations are becoming more and more common in our applications.  With technologies like WPF, Silverlight and jQuery, animations are becoming easier for developers to use (and abuse).  When used properly, animation can augment the user experience.  When used improperly, animation can degrade the user experience.  Sometimes, the differences can be very subtle. I have recently made use of animations in a few projects and I very quickly realized how easy it is to abuse animation techniques.  Here are a few things I have learned along the way. 1) Don’t animate for the sake of animating We’ve all seen the PowerPoint slides with annoying slide transitions that animate 20 different ways.  It’s distracting and tacky.  The same holds true for your application.  While animations are fun and becoming easy to implement, resist the urge to use the technology just because you think the technology is amazing.   2) Animations should (and do) have meaning I recently built a simple Windows Phone 7 (WP7) application, Steeped (download it here).  The application has 2 pages.  The first page lists a number of tea types.  When the user taps on one of the tea types, the application navigates to the second page with information about that tea type and some options for the user to choose from.       One of the last things I did before submitting Steeped to the marketplace was add a page transition between the 2 pages.  I choose the Slide / Fade Out transition.  When the user selects a tea type, the main page slides to the left and fades out.  At the same time, the details page slides in from the right and fades in.  I tested it and thought it looked great so I submitted the app.  A few days later, I asked a friend to try the app.  He selected a tea type, and I was a little surprised by how he used the app.  When he wanted to navigate back to the main page, instead of pressing the back button on the phone, he tried to use a swiping gesture.  Of course, the swiping gesture did nothing because I had not implemented that feature.  After thinking about it for a while, I realized that the page transition I had chosen implied a particular behaviour.  As a user, if an action I perform causes an item (in this case the page) to move, then my expectation is that I should be able to move it back.  I have since added logic to handle the swipe gesture and I think the app flows much better now. When using animation, it pays to ask yourself:  What story does this animation tell my users?   3) Watch the replay Some animations might seem great initially but can get annoying over time.  When you use an animation in your application, make sure you try using it over and over again to make sure it doesn’t get annoying.  When I add an animation, I try watch it at least 25 times in a row.  After watching the animation repeatedly, I can make a more informed decision whether or not I should keep the animation.  Often, I end up shortening the length of the animations.   4) Don’t get in the users way An animation should never slow the user down.  When implemented properly, an animation can give a perceived bump in performance.  A good example of this is a the page transitions in most of the built in apps on WP7.  Obviously, these page animations don’t make the phone any faster, but they do provide a more responsive user experience.  Why?  Because most of the animations begin as soon as the user has performed some action.  The destination page might not be fully loaded yet, but the system responded immediately to user action, giving the impression that the system is more responsive.  If the user did not see anything happen until after the destination page was fully loaded, the application would feel clumsy and slow.  Also, it is important to make sure the animation does not degrade the performance (or perceived performance) of the application.   Jut a few things to consider when using animations.  As is the case with many technologies, we often learn how to misuse it before we learn how to use it effectively.

    Read the article

  • Breaking 1NF to model subset constraints. Does this sound sane?

    - by Chris Travers
    My first question here. Appologize if it is in the wrong forum but this seems pretty conceptual. I am looking at doing something that goes against conventional wisdom and want to get some feedback as to whether this is totally insane or will result in problems, so critique away! I am on PostgreSQL 9.1 but may be moving to 9.2 for this part of this project. To re-iterate: Does it seem sane to break 1NF in this way? I am not looking for debugging code so much as where people see problems that this might lead. The Problem In double entry accounting, financial transactions are journal entries with an arbitrary number of lines. Each line has either a left value (debit) or a right value (credit) which can be modelled as a single value with negatives as debits and positives as credits or vice versa. The sum of all debits and credits must equal zero (so if we go with a single amount field, sum(amount) must equal zero for each financial journal entry). SQL-based databases, pretty much required for this sort of work, have no way to express this sort of constraint natively and so any approach to enforcing it in the database seems rather complex. The Write Model The journal entries are append only. There is a possibility we will add a delete model but it will be subject to a different set of restrictions and so is not applicable here. If and when we allow deletes, we will probably do them using a simple ON DELETE CASCADE designation on the foreign key, and require that deletes go through a dedicated stored procedure which can enforce the other constraints. So inserts and selects have to be accommodated but updates and deletes do not for this task. My Proposed Solution My proposed solution is to break first normal form and model constraints on arrays of tuples, with a trigger that breaks the rows out into another table. CREATE TABLE journal_line ( entry_id bigserial primary key, account_id int not null references account(id), journal_entry_id bigint not null, -- adding references later amount numeric not null ); I would then add "table methods" to extract debits and credits for reporting purposes: CREATE OR REPLACE FUNCTION debits(journal_line) RETURNS numeric LANGUAGE sql IMMUTABLE AS $$ SELECT CASE WHEN $1.amount < 0 THEN $1.amount * -1 ELSE NULL END; $$; CREATE OR REPLACE FUNCTION credits(journal_line) RETURNS numeric LANGUAGE sql IMMUTABLE AS $$ SELECT CASE WHEN $1.amount > 0 THEN $1.amount ELSE NULL END; $$; Then the journal entry table (simplified for this example): CREATE TABLE journal_entry ( entry_id bigserial primary key, -- no natural keys :-( journal_id int not null references journal(id), date_posted date not null, reference text not null, description text not null, journal_lines journal_line[] not null ); Then a table method and and check constraints: CREATE OR REPLACE FUNCTION running_total(journal_entry) returns numeric language sql immutable as $$ SELECT sum(amount) FROM unnest($1.journal_lines); $$; ALTER TABLE journal_entry ADD CONSTRAINT CHECK (((journal_entry.running_total) = 0)); ALTER TABLE journal_line ADD FOREIGN KEY journal_entry_id REFERENCES journal_entry(entry_id); And finally we'd have a breakout trigger: CREATE OR REPLACE FUNCTION je_breakout() RETURNS TRIGGER LANGUAGE PLPGSQL AS $$ BEGIN IF TG_OP = 'INSERT' THEN INSERT INTO journal_line (journal_entry_id, account_id, amount) SELECT NEW.id, account_id, amount FROM unnest(NEW.journal_lines); RETURN NEW; ELSE RAISE EXCEPTION 'Operation Not Allowed'; END IF; END; $$; And finally CREATE TRIGGER AFTER INSERT OR UPDATE OR DELETE ON journal_entry FOR EACH ROW EXECUTE_PROCEDURE je_breaout(); Of course the example above is simplified. There will be a status table that will track approval status allowing for separation of duties, etc. However the goal here is to prevent unbalanced transactions. Any feedback? Does this sound entirely insane? Standard Solutions? In getting to this point I have to say I have looked at four different current ERP solutions to this problems: Represent every line item as a debit and a credit against different accounts. Use of foreign keys against the line item table to enforce an eventual running total of 0 Use of constraint triggers in PostgreSQL Forcing all validation here solely through the app logic. My concerns are that #1 is pretty limiting and very hard to audit internally. It's not programmer transparent and so it strikes me as being difficult to work with in the future. The second strikes me as being very complex and required a series of contraints and foreign keys against self to make work, and therefore it strikes me as complex, hard to sort out at least in my mind, and thus hard to work with. The fourth could be done as we force all access through stored procedures anyway and this is the most common solution (have the app total things up and throw an error otherwise). However, I think proof that a constraint is followed is superior to test cases, and so the question becomes whether this in fact generates insert anomilies rather than solving them. If this is a solved problem it isn't the case that everyone agrees on the solution....

    Read the article

  • Using Resources the Right Way

    - by BuckWoody
    It’s an interesting time in computing technology. At one point there was a dearth of information available for solving a given problem, or educating ourselves on broader topics so that we can solve problems in the future. With dozens, perhaps hundreds or thousands of web sites and content available (for free, in many cases) from vendors, peers, even colleges and universities, it seems like there is actually too much information. Who has the time to absorb all this information and training? Even if you had the inclination, where to start? In fact, it seems so overwhelming that I often hear people saying that they can’t find the training they need, or that vendor X or Y “doesn’t help their users”. On questioning these folks, however, I often find that they – and sometimes I - haven’t put in the effort to learn what resources we have. That’s where blogs, like this one, can help. If you follow a blog, either by checking it often or perhaps subscribing to the Really Simple Syndication (RSS) feed, you’ll be able to spread out the search or create a mental filter for the information you need. But it’s not enough just read a blog or a web page. The creators need real feedback – what doesn’t work, and what does. Yes, you’re allowed to tell a vendor or writer “This helped me because…” so that you reinforce the positives. To be sure, bring up what doesn’t work as well –  that’s fine. But be specific, and be constructive. You’d be surprised at how much it matters. I know for a fact at Microsoft we listen – there is a real live person that reads your comments. I’m sure this is true of other vendors, and I also know that most blog authors – yours truly most especially – wants to know what you think.   In this blog entry I’d to call your attention to three resources you have at your disposal, and how you can use them to help. I’ll try to bring up things like this from time to time that I find useful, and cover in them in more depth like this. Think of this as a synopsis of a longer set of resources that you can use to filter whether you want to research further, bookmark, or forward on to a circle of friends where you think it might help them.   Data Driven Design Concepts http://msdn.microsoft.com/en-us/library/windowsazure/jj156154 I’ll start with a great site that walks you through the process of designing a solution from a data-first perspective. As you know, I believe all computing is merely re-arranging data. If you follow that logic as well, you’ll realize that whenever you create a solution, you should start at the data-end of the application. This resource helps you do that. Even if you don’t use the specific technologies the instructions use, the concepts hold for almost any other technology that deals with data. This should be a definite bookmark for a developer, DBA, or Data Architect. When I mentioned my admiration for this resource here at Microsoft, the team that created it contacted me and asked if I’d share an e-mail address to my readers so that you can comment on it. You’re guaranteed to be heard – you can suggest changes, talk about how useful – or not – it is, and so on. Here’s that address:  [email protected]   End-to-End Example of a complete Hybrid Application – with Live Demo https://azurestocktrader.cloudapp.net/Default.aspx I learn by example. I also like having ready-made, live, functional demos that show the completed solution at work. If you’ve ever wanted to learn how a complex, complete, hybrid application that bridges on-premises systems with cloud-based databases, code, functions and more, this is it. It’s a stock-trading simulator, and you can get everything from the design to the code itself, or you can just play with the application. It’s running on Windows Azure, the actual production servers we use for everything else. Using a Cloud-Based Service https://azureconfigweb.cloudapp.net/Default.aspx Along with that stock-trading application, you have a full demonstration and usable code sample of a web-based service available. If you’re a developer, this is a style of code you need to understand for everything from iPhone development to a full Service-Oriented Architecture (SOA) environment. So check out these resources. I’ll post more from time to time as I run across them. Hopefully they’ll be as useful to you as they are to me. Oh, and if you have a comment on any of the resources, let them know. And if you have any comments about these or any of my entries, feel free to post away. To quote a famous TV Show: “Hello Seattle – I’m listening…”

    Read the article

  • Inside BackgroundWorker

    - by João Angelo
    The BackgroundWorker is a reusable component that can be used in different contexts, but sometimes with unexpected results. If you are like me, you have mostly used background workers while doing Windows Forms development due to the flexibility they offer for running a background task. They support cancellation and give events that signal progress updates and task completion. When used in Windows Forms, these events (ProgressChanged and RunWorkerCompleted) get executed back on the UI thread where you can freely access your form controls. However, the logic of the progress changed and worker completed events being invoked in the thread that started the background worker is not something you get directly from the BackgroundWorker, but instead from the fact that you are running in the context of Windows Forms. Take the following example that illustrates the use of a worker in three different scenarios: – Console Application or Windows Service; – Windows Forms; – WPF. using System; using System.ComponentModel; using System.Threading; using System.Windows.Forms; using System.Windows.Threading; class Program { static AutoResetEvent Synch = new AutoResetEvent(false); static void Main() { var bw1 = new BackgroundWorker(); var bw2 = new BackgroundWorker(); var bw3 = new BackgroundWorker(); Console.WriteLine("DEFAULT"); var unspecializedThread = new Thread(() => { OutputCaller(1); SynchronizationContext.SetSynchronizationContext( new SynchronizationContext()); bw1.DoWork += (sender, e) => OutputWork(1); bw1.RunWorkerCompleted += (sender, e) => OutputCompleted(1); // Uses default SynchronizationContext bw1.RunWorkerAsync(); }); unspecializedThread.IsBackground = true; unspecializedThread.Start(); Synch.WaitOne(); Console.WriteLine(); Console.WriteLine("WINDOWS FORMS"); var windowsFormsThread = new Thread(() => { OutputCaller(2); SynchronizationContext.SetSynchronizationContext( new WindowsFormsSynchronizationContext()); bw2.DoWork += (sender, e) => OutputWork(2); bw2.RunWorkerCompleted += (sender, e) => OutputCompleted(2); // Uses WindowsFormsSynchronizationContext bw2.RunWorkerAsync(); Application.Run(); }); windowsFormsThread.IsBackground = true; windowsFormsThread.SetApartmentState(ApartmentState.STA); windowsFormsThread.Start(); Synch.WaitOne(); Console.WriteLine(); Console.WriteLine("WPF"); var wpfThread = new Thread(() => { OutputCaller(3); SynchronizationContext.SetSynchronizationContext( new DispatcherSynchronizationContext()); bw3.DoWork += (sender, e) => OutputWork(3); bw3.RunWorkerCompleted += (sender, e) => OutputCompleted(3); // Uses DispatcherSynchronizationContext bw3.RunWorkerAsync(); Dispatcher.Run(); }); wpfThread.IsBackground = true; wpfThread.SetApartmentState(ApartmentState.STA); wpfThread.Start(); Synch.WaitOne(); } static void OutputCaller(int workerId) { Console.WriteLine( "bw{0}.{1} | Thread: {2} | IsThreadPool: {3}", workerId, "RunWorkerAsync".PadRight(18), Thread.CurrentThread.ManagedThreadId, Thread.CurrentThread.IsThreadPoolThread); } static void OutputWork(int workerId) { Console.WriteLine( "bw{0}.{1} | Thread: {2} | IsThreadPool: {3}", workerId, "DoWork".PadRight(18), Thread.CurrentThread.ManagedThreadId, Thread.CurrentThread.IsThreadPoolThread); } static void OutputCompleted(int workerId) { Console.WriteLine( "bw{0}.{1} | Thread: {2} | IsThreadPool: {3}", workerId, "RunWorkerCompleted".PadRight(18), Thread.CurrentThread.ManagedThreadId, Thread.CurrentThread.IsThreadPoolThread); Synch.Set(); } } Output: //DEFAULT //bw1.RunWorkerAsync | Thread: 3 | IsThreadPool: False //bw1.DoWork | Thread: 4 | IsThreadPool: True //bw1.RunWorkerCompleted | Thread: 5 | IsThreadPool: True //WINDOWS FORMS //bw2.RunWorkerAsync | Thread: 6 | IsThreadPool: False //bw2.DoWork | Thread: 5 | IsThreadPool: True //bw2.RunWorkerCompleted | Thread: 6 | IsThreadPool: False //WPF //bw3.RunWorkerAsync | Thread: 7 | IsThreadPool: False //bw3.DoWork | Thread: 5 | IsThreadPool: True //bw3.RunWorkerCompleted | Thread: 7 | IsThreadPool: False As you can see the output between the first and remaining scenarios is somewhat different. While in Windows Forms and WPF the worker completed event runs on the thread that called RunWorkerAsync, in the first scenario the same event runs on any thread available in the thread pool. Another scenario where you can get the first behavior, even when on Windows Forms or WPF, is if you chain the creation of background workers, that is, you create a second worker in the DoWork event handler of an already running worker. Since the DoWork executes in a thread from the pool the second worker will use the default synchronization context and the completed event will not run in the UI thread.

    Read the article

  • Avoiding coupling

    - by Seralize
    It is also true that a system may become so coupled, where each class is dependent on other classes that depend on other classes, that it is no longer possible to make a change in one place without having a ripple effect and having to make subsequent changes in many places.[1] This is why using an interface or an abstract class can be valuable in any object-oriented software project. Quote from Wikipedia Starting from scratch I'm starting from scratch with a project that I recently finished because I found the code to be too tightly coupled and hard to refactor, even when using MVC. I will be using MVC on my new project aswell but want to try and avoid the pitfalls this time, hopefully with your help. Project summary My issue is that I really wish to keep the Controller as clean as possible, but it seems like I can't do this. The basic idea of the program is that the user picks wordlists which is sent to the game engine. It will pick random words from the lists until there are none left. Problem at hand My main problem is that the game will have 'modes', and need to check the input in different ways through a method called checkWord(), but exactly where to put this and how to abstract it properly is a challenge to me. I'm new to design patterns, so not sure whether there exist any might fit my problem. My own attempt at abstraction Here is what I've gotten so far after hours of 'refactoring' the design plans, and I know it's long, but it's the best I could do to try and give you an overview (Note: As this is the sketch, anything is subject to change, all help and advice is very welcome. Also note the marked coupling points): Wordlist class Wordlist { // Basic CRUD etc. here! // Other sample methods: public function wordlistCount($user_id) {} // Returns count of how many wordlists a user has public function getAll($user_id) {} // Returns all wordlists of a user } Word class Word { // Basic CRUD etc. here! // Other sample methods: public function wordCount($wordlist_id) {} // Returns count of words in a wordlist public function getAll($wordlist_id) {} // Returns all words from a wordlist public function getWordInfo($word_id) {} // Returns information about a word } Wordpicker class Wordpicker { // The class needs to know which words and wordlists to exclude protected $_used_words = array(); protected $_used_wordlists = array(); // Wordlists to pick words from protected $_wordlists = array(); /* Public Methods */ public function setWordlists($wordlists = array()) {} public function setUsedWords($used_words = array()) {} public function setUsedWordlists($used_wordlists = array()) {} public function getRandomWord() {} // COUPLING POINT! Will most likely need to communicate with both the Wordlist and Word classes /* Protected Methods */ protected function _checkAvailableWordlists() {} // COUPLING POINT! Might need to check if wordlists are deleted etc. protected function _checkAvailableWords() {} // COUPLING POINT! Method needs to get all words in a wordlist from the Word class } Game class Game { protected $_session_id; // The ID of a game session which gets stored in the database along with game details protected $_game_info = array(); // Game instantiation public function __construct($user_id) { if (! $this->_session_id = $this->_gameExists($user_id)) { // New game } else { // Resume game } } // This is the method I tried to make flexible by using abstract classes etc. // Does it even belong in this class at all? public function checkWord($answer, $native_word, $translation) {} // This method checks the answer against the native word / translation word, depending on game mode public function getGameInfo() {} // Returns information about a game session, or creates it if it does not exist public function deleteSession($session_id) {} // Deletes a game session from the database // Methods dealing with game session information protected function _gameExists($user_id) {} protected function _getProgress($session_id) {} protected function _updateProgress($game_info = array()) {} } The Game /* CONTROLLER */ /* "Guess the word" page */ // User input $game_type = $_POST['game_type']; // Chosen with radio buttons etc. $wordlists = $_POST['wordlists']; // Chosen with checkboxes etc. // Starts a new game or resumes one from the database $game = new Game($_SESSION['user_id']); $game_info = $game->getGameInfo(); // Instantiates a new Wordpicker $wordpicker = new Wordpicker(); $wordpicker->setWordlists((isset($game_info['wordlists'])) ? $game_info['wordlists'] : $wordlists); $wordpicker->setUsedWordlists((isset($game_info['used_wordlists'])) ? $game_info['used_wordlists'] : NULL); $wordpicker->setUsedWords((isset($game_info['used_words'])) ? $game_info['used_words'] : NULL); // Fetches an available word if (! $word_id = $wordpicker->getRandomWord()) { // No more words left - game over! $game->deleteSession($game_info['id']); redirect(); } else { // Presents word details to the user $word = new Word(); $word_info = $word->getWordInfo($word_id); } The Bit to Finish /* CONTROLLER */ /* "Check the answer" page */ // ?????????????????? ( http://pastebin.com/cc6MtLTR ) Make sure you toggle the 'Layout Width' to the right for a better view. Thanks in advance. Questions To which extent should objects be loosely coupled? If object A needs info from object B, how is it supposed to get this without losing too much cohesion? As suggested in the comments, models should hold all business logic. However, as objects should be independent, where to glue them together? Should the model contain some sort of "index" or "client" area which connects the dots? Edit: So basically what I should do for a start is to make a new model which I can more easily call with oneliners such as $model->doAction(); // Lots of code in here which uses classes! How about the method for checking words? Should it be it's own object? I'm not sure where I should put it as it's pretty much part of the 'game'. But on another hand, I could just leave out the 'abstraction and OOPness' and make it a method of the 'client model' which will be encapsulated from the controller anyway. Very unsure about this.

    Read the article

  • Pirates, Treasure Chests and Architectural Mapping

    Pirate 1: Why do pirates create treasure maps? Pirate 2: I do not know.Pirate 1: So they can find their gold. Yes, that was a bad joke, but it does illustrate a point. Pirates are known for drawing treasure maps to their most prized possession. These documents detail the decisions pirates made in order to hide and find their chests of gold. The map allows them to trace the steps they took originally to hide their treasure so that they may return. As software engineers, programmers, and architects we need to treat software implementations much like our treasure chest. Why is software like a treasure chest? It cost money, time,  and resources to develop (Usually) It can make or save money, time, and resources (Hopefully) If we operate under the assumption that software is like a treasure chest then wouldn’t make sense to document the steps, rationale, concerns, and decisions about how it was designed? Pirates are notorious for documenting where they hide their treasure.  Shouldn’t we as creators of software do the same? By documenting our design decisions and rationale behind them will help others be able to understand and maintain implemented systems. This can only be done if the design decisions are correctly mapped to its corresponding implementation. This allows for architectural decisions to be traced from the conceptual model, architectural design and finally to the implementation. Mapping gives software professional a method to trace the reason why specific areas of code were developed verses other options. Just like the pirates we need to able to trace our steps from the start of a project to its implementation,  so that we will understand why specific choices were chosen. The traceability of a software implementation that actually maps back to its originating design decisions is invaluable for ensuring that architectural drifting and erosion does not take place. The drifting and erosion is prevented by allowing others to understand the rational of why an implementation was created in a specific manor or methodology The process of mapping distinct design concerns/decisions to the location of its implemented is called traceability. In this context traceability is defined as method for connecting distinctive software artifacts. This process allows architectural design models and decisions to be directly connected with its physical implementation. The process of mapping architectural design concerns to a software implementation can be very complex. However, most design decision can be placed in  a few generalized categories. Commonly Mapped Design Decisions Design Rationale Components and Connectors Interfaces Behaviors/Properties Design rational is one of the hardest categories to map directly to an implementation. Typically this rational is mapped or document in code via comments. These comments consist of general design decisions and reasoning because they do not directly refer to a specific part of an application. They typically focus more on the higher level concerns. Components and connectors can directly be mapped to architectural concerns. Typically concerns subdivide an application in to distinct functional areas. These functional areas then can map directly back to their originating concerns.Interfaces can be mapped back to design concerns in one of two ways. Interfaces that pertain to specific function definitions can be directly mapped back to its originating concern(s). However, more complicated interfaces require additional analysis to ensure that the proper mappings are created. Depending on the complexity some Behaviors\Properties can be translated directly into a generic implementation structure that is ready for business logic. In addition, some behaviors can be translated directly in to an actual implementation depending on the complexity and architectural tools used. Mapping design concerns to an implementation is a lot of work to maintain, but is doable. In order to ensure that concerns are mapped correctly and that an implementation correctly reflects its design concerns then one of two standard approaches are usually used. All Changes Come From ArchitectureBy forcing all application changes to come through the architectural model prior to implementation then the existing mappings will be used to locate where in the implementation changes need to occur. Allow Changes From Implementation Or Architecture By allowing changes to come from the implementation and/or the architecture then the other area must be kept in sync. This methodology is more complex compared to the previous approach.  One reason to justify the added complexity for an application is due to the fact that this approach tends to detect and prevent architectural drift and erosion. Additionally, this approach is usually maintained via software because of the complexity. Reference:Taylor, R. N., Medvidovic, N., & Dashofy, E. M. (2009). Software architecture: Foundations, theory, and practice Hoboken, NJ: John Wiley & Sons  

    Read the article

  • Space invaders clone not moving properly

    - by ThePlan
    I'm trying to make a basic space invaders clone in allegro 5, I've got my game set up, basic events and such, here is the code: #include <allegro5/allegro.h> #include <allegro5/allegro_image.h> #include <allegro5/allegro_primitives.h> #include <allegro5/allegro_font.h> #include <allegro5/allegro_ttf.h> #include "Entity.h" // GLOBALS ========================================== const int width = 500; const int height = 500; const int imgsize = 3; bool key[5] = {false, false, false, false, false}; bool running = true; bool draw = true; // FUNCTIONS ======================================== void initSpaceship(Spaceship &ship); void moveSpaceshipRight(Spaceship &ship); void moveSpaceshipLeft(Spaceship &ship); void initInvader(Invader &invader); void moveInvaderRight(Invader &invader); void moveInvaderLeft(Invader &invader); void initBullet(Bullet &bullet); void fireBullet(); void doCollision(); void updateInvaders(); void drawText(); enum key_t { UP, DOWN, LEFT, RIGHT, SPACE }; enum source_t { INVADER, DEFENDER }; int main(void) { if(!al_init()) { return -1; } Spaceship ship; Invader invader; Bullet bullet; al_init_image_addon(); al_install_keyboard(); al_init_font_addon(); al_init_ttf_addon(); ALLEGRO_DISPLAY *display = al_create_display(width, height); ALLEGRO_EVENT_QUEUE *event_queue = al_create_event_queue(); ALLEGRO_TIMER *timer = al_create_timer(1.0 / 60); ALLEGRO_BITMAP *images[imgsize]; ALLEGRO_FONT *font1 = al_load_font("arial.ttf", 20, 0); al_register_event_source(event_queue, al_get_keyboard_event_source()); al_register_event_source(event_queue, al_get_display_event_source(display)); al_register_event_source(event_queue, al_get_timer_event_source(timer)); images[0] = al_load_bitmap("defender.bmp"); images[1] = al_load_bitmap("invader.bmp"); images[2] = al_load_bitmap("explosion.bmp"); al_convert_mask_to_alpha(images[0], al_map_rgb(0, 0, 0)); al_convert_mask_to_alpha(images[1], al_map_rgb(0, 0, 0)); al_convert_mask_to_alpha(images[2], al_map_rgb(0, 0, 0)); initSpaceship(ship); initBullet(bullet); initInvader(invader); al_start_timer(timer); while(running) { ALLEGRO_EVENT ev; al_wait_for_event(event_queue, &ev); if(ev.type == ALLEGRO_EVENT_TIMER) { draw = true; if(key[RIGHT] == true) moveSpaceshipRight(ship); if(key[LEFT] == true) moveSpaceshipLeft(ship); } else if(ev.type == ALLEGRO_EVENT_DISPLAY_CLOSE) running = false; else if(ev.type == ALLEGRO_EVENT_KEY_DOWN) { switch(ev.keyboard.keycode) { case ALLEGRO_KEY_ESCAPE: running = false; break; case ALLEGRO_KEY_LEFT: key[LEFT] = true; break; case ALLEGRO_KEY_RIGHT: key[RIGHT] = true; break; case ALLEGRO_KEY_SPACE: key[SPACE] = true; break; } } else if(ev.type == ALLEGRO_KEY_UP) { switch(ev.keyboard.keycode) { case ALLEGRO_KEY_LEFT: key[LEFT] = false; break; case ALLEGRO_KEY_RIGHT: key[RIGHT] = false; break; case ALLEGRO_KEY_SPACE: key[SPACE] = false; break; } } if(draw && al_is_event_queue_empty(event_queue)) { draw = false; al_draw_bitmap(images[0], ship.pos_x, ship.pos_y, 0); al_flip_display(); al_clear_to_color(al_map_rgb(0, 0, 0)); } } al_destroy_font(font1); al_destroy_event_queue(event_queue); al_destroy_timer(timer); for(int i = 0; i < imgsize; i++) al_destroy_bitmap(images[i]); al_destroy_display(display); } // FUNCTION LOGIC ====================================== void initSpaceship(Spaceship &ship) { ship.lives = 3; ship.speed = 2; ship.pos_x = width / 2; ship.pos_y = height - 20; } void initInvader(Invader &invader) { invader.health = 100; invader.count = 40; invader.speed = 0.5; invader.pos_x = 300; invader.pos_y = 300; } void initBullet(Bullet &bullet) { bullet.speed = 10; } void moveSpaceshipRight(Spaceship &ship) { ship.pos_x += ship.speed; if(ship.pos_x >= width) ship.pos_x = width-30; } void moveSpaceshipLeft(Spaceship &ship) { ship.pos_x -= ship.speed; if(ship.pos_x <= 0) ship.pos_x = 0+30; } However it's not behaving the way I want it to behave, in fact the behavior for the ship movement is un-normal. Basically I specified that the ship only moves when the right/left key is down, however the ship is moving constantly to the direction of the key pressed, it never stops although it should only move while my key is down. Even more weird behavior, when I press the opposite key the ship completely stops no matter what else I press. What's wrong with the code? Why does the ship move constantly even after I specified it only moves when a key is down?

    Read the article

  • Implementing a Custom Coherence PartitionAssignmentStrategy

    - by jpurdy
    A recent A-Team engagement required the development of a custom PartitionAssignmentStrategy (PAS). By way of background, a PAS is an implementation of a Java interface that controls how a Coherence partitioned cache service assigns partitions (primary and backup copies) across the available set of storage-enabled members. While seemingly straightforward, this is actually a very difficult problem to solve. Traditionally, Coherence used a distributed algorithm spread across the cache servers (and as of Coherence 3.7, this is still the default implementation). With the introduction of the PAS interface, the model of operation was changed so that the logic would run solely in the cache service senior member. Obviously, this makes the development of a custom PAS vastly less complex, and in practice does not introduce a significant single point of failure/bottleneck. Note that Coherence ships with a default PAS implementation but it is not used by default. Further, custom PAS implementations are uncommon (this engagement was the first custom implementation that we know of). The particular implementation mentioned above also faced challenges related to managing multiple backup copies but that won't be discussed here. There were a few challenges that arose during design and implementation: Naive algorithms had an unreasonable upper bound of computational cost. There was significant complexity associated with configurations where the member count varied significantly between physical machines. Most of the complexity of a PAS is related to rebalancing, not initial assignment (which is usually fairly simple). A custom PAS may need to solve several problems simultaneously, such as: Ensuring that each member has a similar number of primary and backup partitions (e.g. each member has the same number of primary and backup partitions) Ensuring that each member carries similar responsibility (e.g. the most heavily loaded member has no more than one partition more than the least loaded). Ensuring that each partition is on the same member as a corresponding local resource (e.g. for applications that use partitioning across message queues, to ensure that each partition is collocated with its corresponding message queue). Ensuring that a given member holds no more than a given number of partitions (e.g. no member has more than 10 partitions) Ensuring that backups are placed far enough away from the primaries (e.g. on a different physical machine or a different blade enclosure) Achieving the above goals while ensuring that partition movement is minimized. These objectives can be even more complicated when the topology of the cluster is irregular. For example, if multiple cluster members may exist on each physical machine, then clearly the possibility exists that at certain points (e.g. following a member failure), the number of members on each machine may vary, in certain cases significantly so. Consider the case where there are three physical machines, with 3, 3 and 9 members each (respectively). This introduces complexity since the backups for the 9 members on the the largest machine must be spread across the other 6 members (to ensure placement on different physical machines), preventing an even distribution. For any given problem like this, there are usually reasonable compromises available, but the key point is that objectives may conflict under extreme (but not at all unlikely) circumstances. The most obvious general purpose partition assignment algorithm (possibly the only general purpose one) is to define a scoring function for a given mapping of partitions to members, and then apply that function to each possible permutation, selecting the most optimal permutation. This would result in N! (factorial) evaluations of the scoring function. This is clearly impractical for all but the smallest values of N (e.g. a partition count in the single digits). It's difficult to prove that more efficient general purpose algorithms don't exist, but the key take away from this is that algorithms will tend to either have exorbitant worst case performance or may fail to find optimal solutions (or both) -- it is very important to be able to show that worst case performance is acceptable. This quickly leads to the conclusion that the problem must be further constrained, perhaps by limiting functionality or by using domain-specific optimizations. Unfortunately, it can be very difficult to design these more focused algorithms. In the specific case mentioned, we constrained the solution space to very small clusters (in terms of machine count) with small partition counts and supported exactly two backup copies, and accepted the fact that partition movement could potentially be significant (preferring to solve that issue through brute force). We then used the out-of-the-box PAS implementation as a fallback, delegating to it for configurations that were not supported by our algorithm. Our experience was that the PAS interface is quite usable, but there are intrinsic challenges to designing PAS implementations that should be very carefully evaluated before committing to that approach.

    Read the article

  • Using XA Transactions in Coherence-based Applications

    - by jpurdy
    While the costs of XA transactions are well known (e.g. increased data contention, higher latency, significant disk I/O for logging, availability challenges, etc.), in many cases they are the most attractive option for coordinating logical transactions across multiple resources. There are a few common approaches when integrating Coherence into applications via the use of an application server's transaction manager: Use of Coherence as a read-only cache, applying transactions to the underlying database (or any system of record) instead of the cache. Use of TransactionMap interface via the included resource adapter. Use of the new ACID transaction framework, introduced in Coherence 3.6.   Each of these may have significant drawbacks for certain workloads. Using Coherence as a read-only cache is the simplest option. In this approach, the application is responsible for managing both the database and the cache (either within the business logic or via application server hooks). This approach also tends to provide limited benefit for many workloads, particularly those workloads that either have queries (given the complexity of maintaining a fully cached data set in Coherence) or are not read-heavy (where the cost of managing the cache may outweigh the benefits of reading from it). All updates are made synchronously to the database, leaving it as both a source of latency as well as a potential bottleneck. This approach also prevents addressing "hot data" problems (when certain objects are updated by many concurrent transactions) since most database servers offer no facilities for explicitly controlling concurrent updates. Finally, this option tends to be a better fit for key-based access (rather than filter-based access such as queries) since this makes it easier to aggressively invalidate cache entries without worrying about when they will be reloaded. The advantage of this approach is that it allows strong data consistency as long as optimistic concurrency control is used to ensure that database updates are applied correctly regardless of whether the cache contains stale (or even dirty) data. Another benefit of this approach is that it avoids the limitations of Coherence's write-through caching implementation. TransactionMap is generally used when Coherence acts as system of record. TransactionMap is not generally compatible with write-through caching, so it will usually be either used to manage a standalone cache or when the cache is backed by a database via write-behind caching. TransactionMap has some restrictions that may limit its utility, the most significant being: The lock-based concurrency model is relatively inefficient and may introduce significant latency and contention. As an example, in a typical configuration, a transaction that updates 20 cache entries will require roughly 40ms just for lock management (assuming all locks are granted immediately, and excluding validation and writing which will require a similar amount of time). This may be partially mitigated by denormalizing (e.g. combining a parent object and its set of child objects into a single cache entry), at the cost of increasing false contention (e.g. transactions will conflict even when updating different child objects). If the client (application server JVM) fails during the commit phase, locks will be released immediately, and the transaction may be partially committed. In practice, this is usually not as bad as it may sound since the commit phase is usually very short (all locks having been previously acquired). Note that this vulnerability does not exist when a single NamedCache is used and all updates are confined to a single partition (generally implying the use of partition affinity). The unconventional TransactionMap API is cumbersome but manageable. Only a few methods are transactional, primarily get(), put() and remove(). The ACID transactions framework (accessed via the Connection class) provides atomicity guarantees by implementing the NamedCache interface, maintaining its own cache data and transaction logs inside a set of private partitioned caches. This feature may be used as either a local transactional resource or as logging XA resource. However, a lack of database integration precludes the use of this functionality for most applications. A side effect of this is that this feature has not seen significant adoption, meaning that any use of this is subject to the usual headaches associated with being an early adopter (greater chance of bugs and greater risk of hitting an unoptimized code path). As a result, for the moment, we generally recommend against using this feature. In summary, it is possible to use Coherence in XA-oriented applications, and several customers are doing this successfully, but it is not a core usage model for the product, so care should be taken before committing to this path. For most applications, the most robust solution is normally to use Coherence as a read-only cache of the underlying data resources, even if this prevents taking advantage of certain product features.

    Read the article

< Previous Page | 354 355 356 357 358 359 360 361 362 363 364 365  | Next Page >