Search Results

Search found 59975 results on 2399 pages for 'data comparison'.

Page 781/2399 | < Previous Page | 777 778 779 780 781 782 783 784 785 786 787 788  | Next Page >

  • Working with Reporting Services Filters – Part 2: The LIKE Operator

    - by smisner
    In the first post of this series, I introduced the use of filters within the report rather than in the query. I included a list of filter operators, and then focused on the use of the IN operator. As I mentioned in the previous post, the use of some of these operators is not obvious, so I'm going to spend some time explaining them as well as describing ways that you can use report filters in Reporting Services in this series of blog posts. Now let's look at the LIKE operator. If you write T-SQL queries, you've undoubtedly used the LIKE operator to produce a query using the % symbol as a wildcard for multiple characters like this: select * from DimProduct where EnglishProductName like '%Silver%' And you know that you can use the _ symbol as a wildcard for a single character like this: select * from DimProduct where EnglishProductName like '_L Mountain Frame - Black, 4_'   So when you encounter the LIKE operator in a Reporting Services filter, you probably expect it to work the same way. But it doesn't. You use the * symbol as a wildcard for multiple characters as shown here: Expression Data Type Operator Value [EnglishProductName] Text Like *Silver* Note that you don’t have to include quotes around the string that you use for comparison. Books Online has an example of using the % symbol as a wildcard for a single character, but I have not been able to successfully use this wildcard. If anyone has a working example, I’d love to see it!

    Read the article

  • Strategy for backwards compatibility of persistent storage

    - by Baqueta
    In my experience, trying to ensure that new versions of an application retain compatibility with data storage from previous versions can often be a painful process. What I currently do is to save a version number for each 'unit' of data (be it a file, database row/table, or whatever) and ensure that the version number gets updated each time the data changes in some way. I also create methods to convert from v1 to v2, v2 to v3, and so on. That way, if I'm at v7 and I encounter a v3 file, I can do v3-v4-v5-v6-v7. So far this approach seems to be working out well, but I haven't had to make use of it extensively yet so there may be unforseen problems. I'm also concerned that if the objects I'm loading change significantly, I'll either have to keep around old versions of the classes or face updating all my conversion methods to handle the new class definition. Is my approach sound? Are there other/better approaches I could be using? Are there any design patterns applicable to this problem?

    Read the article

  • Are the Ubuntu ISO images updated From release .ubuntu.com

    - by tijybba
    Just got idea from this(may not be related though) question however. Are the ISO images from the official site updated with updates in Core Ubuntu system , like Kernel Updates , desktop Environment Updates(unity), i mean Updates of BASE system including X-org, Office suite, Package Manager , Update manager or Gnome Base Modules, those released in update Branches like precise-Updates branch. The reason i am asking this is , if i download the ISO image of Ubuntu 12.04 Say after two or three months from release , i have to do an update of approximately 200~300 MB's size. So why are these ISO images not updated to recent updates, i am aware that all of the components are not updated at the same time , but let's say after One month from actual release ( Both LTS and normal releases), the updated components can be added to form a Updated ISO in regular intervals, which provides new users to use latest versions and features with improved stability and less bandwidth Consumption. I am not mentioning the idea of comparison to Rolling Release , or External PPA's added updates , and neither Netinstal but the ISO of updated Packages .This can be provided as optional download. Since my question is within the boundary of Official updates releases so stability could not be the reason. I guess there are custom packagers out there , but having an official option would be better. It helps in distributing Newest ISO OS which impresses a lot new users , since it makes availability of newer features and a faster system ofcourse. Another reason of asking this is here. Edit: Since almost all new (Desktop) users download the Default ISO's having one or few issue , which may have been corrected in following updates. But most of new laptop users i encountered gave up because of it , so should i suggest ,for laptop not listed on Certified H/W list , to try daily Builds , if needed.

    Read the article

  • MS Access MSChart.Graph.8 not printing

    - by Tanj
    Software: Microsoft Access 2007 SP2 Database File Version: Access 2000 I have an access program that I inherited from a previous employee. It uses forms for reports and since I don't have much experience in access I have continued to do this. I have created a copy of the program for another project and modified it to suit. I am having trouble getting more then one chart to print. All the charts display in form view, they all have the same properties (excepting data, position, etc.) For some reason they are not printing. They don't even show up in the print preview. I am thinking it must be something with the graphs themselves as they sometimes lose all information. I have to open the graphs in edit mode and change the data source from column to row and back again so that it gets redrawn. (Refresh doesn't fix it) So right now I don't even have a clue as to where to look so ideas are welcome. Edit #1 It seems to be a problem with linking to an unbound form. Subform Field Linker: Can't build a link between unbound forms. The query for the main form is SELECT tTest.ixTest, tMotorTypes.ixMotorType, tMotorTypes.asMotorType, tMotorTypes.fDeprecated, tTestType.asTest, tTest.asSerialNum, tTest.asOrderNum, tTest.asFrameNum, tTest.asRotorNum, tTest.asOperator, tTest.iStation, tTest.dtTestDate, tTest.ixTestType FROM tMotorTypes INNER JOIN (tTestType INNER JOIN tTest ON tTestType.ixTestType=tTest.ixTestType) ON tMotorTypes.ixMotorType=tTest.ixMotorType; The query for the chart is: SELECT qGraphRSTTemperatures.Frequency, qGraphRSTTemperatures.[Drive End], qGraphRSTTemperatures.[Non Drive End], qGraphRSTTemperatures.[Air In], qGraphRSTTemperatures.Core FROM qGraphRSTTemperatures ORDER BY qGraphRSTTemperatures.ixTemperature; Query qGraphRSTTemperatures: SELECT tElectricalData.dblFrequency AS Frequency, tTemperatures.dblDrvEnd AS [Drive End], tTemperatures.dblNonDrvEnd AS [Non Drive End], tTemperatures.dblAirIn AS [Air In], tTemperatures.dblCore AS Core, tSubTest.ixTest, tTemperatures.ixTemperature FROM (tSubTest INNER JOIN tElectricalData ON tSubTest.ixSubTest = tElectricalData.ixSubTest) LEFT JOIN tTemperatures ON tElectricalData.ixElectrical = tTemperatures.ixElectrical WHERE (((tSubTest.ixSubTestType)=5)) ORDER BY tSubTest.ixTest, tTemperatures.ixTemperature; So how come, in the form view it shows the graph with the correct data when linked thus: Child field: ixTest Master field: ixTest but won't print the graph. The graph will print if I remove the links, but then I have all the data from chart query as it is not limited by ixTest. edit #2 It seems to be a data retrieval/rendering issue in printing. Is there anything in printing that changes the context of records with respect to parent/child relationships?

    Read the article

  • How can I making Twitter, Facebook and Reddit share buttons load last?

    - by Daniel Bingham
    I have a website with a number of pages that sport twitter, facebook and reddit share buttons. They take forever to load and until they do the rest of the page doesn't load. So how I can make them load last? Currently, they are loaded something like this: <div class="item"><a href="http://twitter.com/share" class="twitter-share-button" data-count="vertical" data-via="FridgeToFood" data-related="danielBingham:Recipe and update tweets from Fridge to Food.">Tweet</a><script type="text/javascript" src="http://platform.twitter.com/widgets.js"></script></div> <div class="item"><script src="http://connect.facebook.net/en_US/all.js#xfbml=1"></script><fb:like layout="box_count" width="40"></fb:like></div> <div class="item"> <script type="text/javascript">reddit_target='recipes';</script> <script type="text/javascript" src="http://reddit.com/static/button/button2.js"></script> </div> They are in a div called "shareWrapper" and are loading to one side of the page. The buttons load where ever the script code is placed. As far as I know, I can't place the script code at the bottom of the page and move the resulting buttons after the fact. I want them to appear near the top, which right now means they are stopping everything below them from loading for several seconds. I tried loading them using javascript, but using JQuery's $(document).ready(), but that failed. It seems to leave the page in some sort of loading loop from which it never emerges. Are there other ways to get these to load last?

    Read the article

  • Simple Architecture Verification

    - by Jean Carlos Suárez Marranzini
    I just made an architecture for an application with the function of scoring, saving and loading tennis games. The architecture has 2 kinds of elements: components & layers. Components: Standalone elements that can be consumed by other components or by layers. They might also consume functionality from the model/bottom layer. Layers: Software components whose functionality rests on previous layers (except for the model layer). -Layers: -Models: Data and it's behavior. -Controllers: A layer that allows interaction between the views and the models. -Views: The presentation layer for interacting with the user. -Components: -Persistence: Makes sure the game data can be stored away for later retrieval. -Time Machine: Records changes in the game through time so it's possible to navigate the game back and forth. -Settings: Contains the settings that determine how some of the game logic will apply. -Game Engine: Contains all the game logic, which it applies to the game data to determine the path the game should take. This is an image of the architecture (I don't have enough rep to post images): http://i49.tinypic.com/35lt5a9.png The requierements which this architecture should satisfy are the following: Save & load games. Move through game history and see how the scoreboard changes as the game evolves. Tie-breaks must be properly managed. Games must be classified by hit-type. Every point can be modified. Match name and player names must be stored. Game logic must be configurable by the user. I would really appreciate any kind of advice or comments on this architecture. To see if it is well built and makes sense as a whole. I took the idea from this link. http://en.wikipedia.org/wiki/Model%E2%80%93view%E2%80%93controller

    Read the article

  • Set default owner/user

    - by Daniel Hollands
    I'm a web developer, and so have set-up an old machine in the office as an Ubuntu Server, for the purposes of testing websites. I've set-up LAMP and have created a /var/www folder, from which all my local sites are served. The issue is that of user permissions, i.e. any files that I copy into that folder (from my Windows machine via the network) automatically take on me (daniel) as their owner. The problem is that I want www-data to become the owner. I did some research and saw that it should be possible to use setuid (and setgid) to automatically set www-data as the owner of all files put into /var/www automatically, so far I've not had any luck making it work. Can someone help please? Thank you UPDATE: Would this do what I want it to do? Default file permissions for php user www-data UPDATE 2: I've kinda fixed my issue by changing my samba settings. Using Webmin, I was able to go in and change the default settings (as seen here: http://imageshack.us/photo/my-images/521/captureon.png/)

    Read the article

  • PHP/MySQL Database application development tool

    - by RCH
    I am an amateur PHP coder, and have built a couple of dozen projects from scratch (including fairly simple e-commerce systems with user authentication, PayPal integration etc - all coded by hand from a clean page. Have also done a price comparison engine that takes data from multiple sites etc.). But I am no expert with OO and other such advanced techniques - I just have a fairly decent grasp of the basics of data processing, logic, functions and trying to optimize code as much as possible. I just want to make this clear so you have some idea of where I'm coming from. I have a couple of fairly large new projects on my plate for corporate clients - both require bespoke database-driven applications with complex relationships, many tables and lots of different front-end functions to manipulate that data for the internal staff in these companies. I figured building these systems from scratch would probably be a huge waste of time. Instead, there must be tools out there that will allow me to construct MySQL databases and build the pages with things like pagination, action buttons, table construction etc. Some kind of database abstraction layer, or system generator, if you will. What tool do you recommend for such a purpose for someone at my level? Open source would be great, but I don't mind paying for something decent as well. Thanks for any advice.

    Read the article

  • links for 2010-12-17

    - by Bob Rhubart
    Overview of Oracle Enterprise Manager Management Packs How does Oracle Enterprise Manager Grid Control do so much across so many different systems? Porus Homi Havewala has the answers.  (tags: oracle otn grid soa entarch) How to overcome cloud computing hurdles - Computerworld What does it take to go from 'we should move to the cloud' to a successful cloud computing strategy? This excerpt from Silver Clouds, Dark Linings offers advice on crossing cloud chasms and developing a successful roadmap. (tags: ping.fm) Security in OBIEE 11g, Part 2 Guest blogger Pravin Janardanam continues the discussion about OBIEE 11g Authorization and other Security aspects. (tags: oracle otn security businessintelligence obiee) Oracle Fusion Middleware Security: A Quick Note about Oracle Access Manager 11g and WebLogic "OAM 11g integrates with WebLogic using the very same components used to integrate OAM 10.1.4.3. Under most circumstances, that means using the OAM Identity Asserter...which asserts the OAM_REMOTE_USER header as the user principal in the JAAS subject." - Brian Eidelman (tags: WebLogic oracle) Comparison Between Cluster Multicast Messaging and Unicast Messaging Mode Weblogic wonders!!! "When servers are in a cluster, these member servers communicate with each other by sending heartbeats and indicating that they are alive. For this communication between the servers, either unicast or multicast messaging is used." -- Divya Duryea (tags: weblogic oracle) Ron Batra: Cloud Computing Series: IV: Database.com, ExaData on Demand and connecting the dots Oracle ACE Direct Ron Batra offers his assessment of recent rumblings in the Cloud. (tags: oracle otn oracleace cloud database)

    Read the article

  • help merging perl code routines together for file processing

    - by jdamae
    I need some perl help in putting these (2) processes/code to work together. I was able to get them working individually to test, but I need help bringing them together especially with using the loop constructs. I'm not sure if I should go with foreach..anyways the code is below. Also, any best practices would be great too as I'm learning this language. Thanks for your help. Here's the process flow I am looking for: -read a directory -look for a particular file -use the file name to strip out some key information to create a newly processed file -process the input file -create the newly processed file for each input file read (if i read in 10, I create 10 new files) Sample Recs: col1,col2,col3,col4,col5 [email protected],[email protected],8,2009-09-24 21:00:46,1 [email protected],[email protected],16,2007-08-18 22:53:12,33 [email protected],[email protected],16,2007-08-18 23:41:23,33 Here's my test code: Target Filetype: `/backups/test/foo101.name.aue-foo_p002.20110124.csv` Part 1: my $target_dir = "/backups/test/"; opendir my $dh, $target_dir or die "can't opendir $target_dir: $!"; while (defined(my $file = readdir($dh))) { next if ($file =~ /^\.+$/); #Get filename attributes if ($file =~ /^foo(\d{3})\.name\.(\w{3})-foo_p(\d{1,4})\.\d+.csv$/) { print "$1\n"; print "$2\n"; print "$3\n"; } print "$file\n"; } Part 2: use strict; use Digest::MD5 qw(md5_hex); #Create new file open (NEWFILE, ">/backups/processed/foo$1.name.$2-foo_p$3.out") || die "cannot create file"; my $data = ''; my $line1 = <>; chomp $line1; my @heading = split /,/, $line1; my ($sep1, $sep2, $eorec) = ( "^A", "^E", "^D"); while (<>) { my $digest = md5_hex($data); chomp; my (@values) = split /,/; my $extra = "__mykey__$sep1$digest$sep2" ; $extra .= "$heading[$_]$sep1$values[$_]$sep2" for (0..scalar(@values)); $data .= "$extra$eorec"; print NEWFILE "$data"; } #print $data; close (NEWFILE);

    Read the article

  • Basic procedural generated content works, but how could I do the same in reverse?

    - by andrew
    My 2D world is made up of blocks. At the moment, I create a block and assign it a number between 1 and 4. The number assigned to the nth block is always the same (i.e if the player walks backwards or restarts the game.) and is generated in the function below. As shown here by this animation, the colours represent the number. function generate_data(n) math.randomseed(n) -- resets the random so that the 'random' number for n is always the same math.random() -- fixes lua random bug local no = math.random(4) --print(no, n) return no end Now I want to limit the next block's number - a block of 1 will always have a block 2 after it, while block 2 will either have a block 1,2 or 3 after it, etc. Before, all the blocks data was randomly generated, initially, and then saved. This data was then loaded and used instead of being randomly called. While working this way, I could specify what the next block would be easily and it would be saved for consistency. I have now removed this saving/loading in favour of procedural generation as I realised that save whiles would get very big after travelling. Back to the present. While travelling forward (to the right), it is easy to limit what the next blocks number will be. I can generate it at the same time as the other data. The problem is when travelling backwards (to the left) I can not think of a way to load the previous block so that it is always the same. Does anyone have any ideas on how I could sort this out?

    Read the article

  • Another good free utility - Campwood Software Source Monitor

    - by TATWORTH
    The Campwoood Source Monitor at http://www.campwoodsw.com/sourcemonitor.html  says in its introduction "The freeware program SourceMonitor lets you see inside your software source code to find out how much code you have and to identify the relative complexity of your modules. For example, you can use SourceMonitor to identify the code that is most likely to contain defects and thus warrants formal review. SourceMonitor, written in C++, runs through your code at high speed, typically at least 10,000 lines of code per second." It is indeed very high-speed and is useful as it: Collects metrics in a fast, single pass through source files. Measures metrics for source code written in C++, C, C#, VB.NET, Java, Delphi, Visual Basic (VB6) or HTML. Includes method and function level metrics for C++, C, C#, VB.NET, Java, and Delphi. Offers Modified Complexity metric option. Saves metrics in checkpoints for comparison during software development projects. Displays and prints metrics in tables and charts, including Kiviat diagrams. Operates within a standard Windows GUI or inside your scripts using XML command files. Exports metrics to XML or CSV (comma-separated-value) files for further processing with other tools.

    Read the article

  • Of transactions and Mongo

    - by Nuri Halperin
    Originally posted on: http://geekswithblogs.net/nuri/archive/2014/05/20/of-transactions-and-mongo-again.aspxWhat's the first thing you hear about NoSQL databases? That they lose your data? That there's no transactions? No joins? No hope for "real" applications? Well, you *should* be wondering whether a certain of database is the right one for your job. But if you do so, you should be wondering that about "traditional" databases as well! In the spirit of exploration let's take a look at a common challenge: You are a bank. You have customers with accounts. Customer A wants to pay B. You want to allow that only if A can cover the amount being transferred. Let's looks at the problem without any context of any database engine in mind. What would you do? How would you ensure that the amount transfer is done "properly"? Would you prevent a "transaction" from taking place unless A can cover the amount? There are several options: Prevent any change to A's account while the transfer is taking place. That boils down to locking. Apply the change, and allow A's balance to go below zero. Charge person A some interest on the negative balance. Not friendly, but certainly a choice. Don't do either. Options 1 and 2 are difficult to attain in the NoSQL world. Mongo won't save you headaches here either. Option 3 looks a bit harsh. But here's where this can go: ledger. See, and account doesn't need to be represented by a single row in a table of all accounts with only the current balance on it. More often than not, accounting systems use ledgers. And entries in ledgers - as it turns out – don't actually get updated. Once a ledger entry is written, it is not removed or altered. A transaction is represented by an entry in the ledger stating and amount withdrawn from A's account and an entry in the ledger stating an addition of said amount to B's account. For sake of space-saving, that entry in the ledger can happen using one entry. Think {Timestamp, FromAccountId, ToAccountId, Amount}. The implication of the original question – "how do you enforce non-negative balance rule" then boils down to: Insert entry in ledger Run validation of recent entries Insert reverse entry to roll back transaction if validation failed. What is validation? Sum up the transactions that A's account has (all deposits and debits), and ensure the balance is positive. For sake of efficiency, one can roll up transactions and "close the book" on transactions with a pseudo entry stating balance as of midnight or something. This lets you avoid doing math on the fly on too many transactions. You simply run from the latest "approved balance" marker to date. But that's an optimization, and premature optimizations are the root of (some? most?) evil.. Back to some nagging questions though: "But mongo is only eventually consistent!" Well, yes, kind of. It's not actually true that Mongo has not transactions. It would be more descriptive to say that Mongo's transaction scope is a single document in a single collection. A write to a Mongo document happens completely or not at all. So although it is true that you can't update more than one documents "at the same time" under a "transaction" umbrella as an atomic update, it is NOT true that there' is no isolation. So a competition between two concurrent updates is completely coherent and the writes will be serialized. They will not scribble on the same document at the same time. In our case - in choosing a ledger approach - we're not even trying to "update" a document, we're simply adding a document to a collection. So there goes the "no transaction" issue. Now let's turn our attention to consistency. What you should know about mongo is that at any given moment, only on member of a replica set is writable. This means that the writable instance in a set of replicated instances always has "the truth". There could be a replication lag such that a reader going to one of the replicas still sees "old" state of a collection or document. But in our ledger case, things fall nicely into place: Run your validation against the writable instance. It is guaranteed to have a ledger either with (after) or without (before) the ledger entry got written. No funky states. Again, the ledger writing *adds* a document, so there's no inconsistent document state to be had either way. Next, we might worry about data loss. Here, mongo offers several write-concerns. Write-concern in Mongo is a mode that marshals how uptight you want the db engine to be about actually persisting a document write to disk before it reports to the application that it is "done". The most volatile, is to say you don't care. In that case, mongo would just accept your write command and say back "thanks" with no guarantee of persistence. If the server loses power at the wrong moment, it may have said "ok" but actually no written the data to disk. That's kind of bad. Don't do that with data you care about. It may be good for votes on a pole regarding how cute a furry animal is, but not so good for business. There are several other write-concerns varying from flushing the write to the disk of the writable instance, flushing to disk on several members of the replica set, a majority of the replica set or all of the members of a replica set. The former choice is the quickest, as no network coordination is required besides the main writable instance. The others impose extra network and time cost. Depending on your tolerance for latency and read-lag, you will face a choice of what works for you. It's really important to understand that no data loss occurs once a document is flushed to an instance. The record is on disk at that point. From that point on, backup strategies and disaster recovery are your worry, not loss of power to the writable machine. This scenario is not different from a relational database at that point. Where does this leave us? Oh, yes. Eventual consistency. By now, we ensured that the "source of truth" instance has the correct data, persisted and coherent. But because of lag, the app may have gone to the writable instance, performed the update and then gone to a replica and looked at the ledger there before the transaction replicated. Here are 2 options to deal with this. Similar to write concerns, mongo support read preferences. An app may choose to read only from the writable instance. This is not an awesome choice to make for every ready, because it just burdens the one instance, and doesn't make use of the other read-only servers. But this choice can be made on a query by query basis. So for the app that our person A is using, we can have person A issue the transfer command to B, and then if that same app is going to immediately as "are we there yet?" we'll query that same writable instance. But B and anyone else in the world can just chill and read from the read-only instance. They have no basis to expect that the ledger has just been written to. So as far as they know, the transaction hasn't happened until they see it appear later. We can further relax the demand by creating application UI that reacts to a write command with "thank you, we will post it shortly" instead of "thank you, we just did everything and here's the new balance". This is a very powerful thing. UI design for highly scalable systems can't insist that the all databases be locked just to paint an "all done" on screen. People understand. They were trained by many online businesses already that your placing of an order does not mean that your product is already outside your door waiting (yes, I know, large retailers are working on it... but were' not there yet). The second thing we can do, is add some artificial delay to a transaction's visibility on the ledger. The way that works is simply adding some logic such that the query against the ledger never nets a transaction for customers newer than say 15 minutes and who's validation flag is not set. This buys us time 2 ways: Replication can catch up to all instances by then, and validation rules can run and determine if this transaction should be "negated" with a compensating transaction. In case we do need to "roll back" the transaction, the backend system can place the timestamp of the compensating transaction at the exact same time or 1ms after the original one. Effectively, once A or B visits their ledger, both transactions would be visible and the overall balance "as of now" would reflect no change.  The 2 transactions (attempted/ reverted) would be visible , since we do actually account for the attempt. Hold on a second. There's a hole in the story: what if several transfers from A to some accounts are registered, and 2 independent validators attempt to compute the balance concurrently? Is there a chance that both would conclude non-sufficient-funds even though rolling back transaction 100 would free up enough for transaction 117 (some random later transaction)? Yes. there is that chance. But the integrity of the business rule is not compromised, since the prime rule is don't dispense money you don't have. To minimize or eliminate this scenario, we can also assign a single validation process per origin account. This may seem non-scalable, but it can easily be done as a "sharded" distribution. Say we have 11 validation threads (or processing nodes etc.). We divide the account number space such that each validator is exclusively responsible for a certain range of account numbers. Sounds cunningly similar to Mongo's sharding strategy, doesn't it? Each validator then works in isolation. More capacity needed? Chop the account space into more chunks. So where  are we now with the nagging questions? "No joins": Huh? What are those for? "No transactions": You mean no cross-collection and no cross-document transactions? Granted - but don't always need them either. "No hope for real applications": well... There are more issues and edge cases to slog through, I'm sure. But hopefully this gives you some ideas of how to solve common problems without distributed locking and relational databases. But then again, you can choose relational databases if they suit your problem.

    Read the article

  • SSISDB Analysis Script on Gist

    - by Davide Mauri
    I've created two simple, yet very useful, script to extract some useful data to quickly monitor SSIS packages execution in SQL Server 2012 and after.get-ssis-execution-status  get-ssis-data-pumped-rows  I've started to use gist since it comes very handy, for this "quick'n'dirty" scripts and snippets, and you can find the above scripts and others (hopefully the number will increase over time...I plan to use gist to store all the code snippet I used to store in a dedicated folder on my machine) there.Now, back to the aforementioned scripts. The first one ("get-ssis-execution-status") returns a list of all executed and executing packages along with latest successful and running executions (so that on can have an idea of the expected run time)error messageswarning messages related to duplicate rows found in lookupsthe second one ("get-ssis-data-pumped-rows") returns information on DataFlows status. Here there's something interesting, IMHO. Nothing exceptional, let it be clear, but nonetheless useful: the script extract information on destinations and row sent to destinations right from the messages produced by the DataFlow component. This helps to quickly understand how many rows as been sent and where...without having to increase the logging level.Enjoy! PSI haven't tested it with SQL Server 2014, but AFAIK they should work without problems. Of course any feedback on this is welcome. 

    Read the article

  • Making simple forms in web applications

    - by levalex
    How do you work with forms in your web applications? I am not talking about RESTful applications, I don't want to build heavy front-end using frameworks like Backbone. For example, I need to add "contact us" form. I need to check data which was filled by user and tell him that his data was sent. Requirements: I want to use AJAX. I want to validate form on back-end side and don't want to duplicate the same code on front-end side. I have my own solution, but it doesn't satisfy me. I make an AJAX request with serialized data on form submit and get response. The next is checking "Content-type" header. html - It means that errors with filling form are exists and response html is form with error labels. - I will replace my form with response html. json and response.error_code == 0 - It means that form was successfully submited. - I will show user notification about success. json and response.error_code != 0 - Something was broken on back-end (like connection with database). other - I display the following message : We have been notified and have started to work with that problem. Please, try it later. The problem of that way is that I can't use it with forms that upload file. What is your practise? What libraries and principles do you use?

    Read the article

  • ODI 11g - Faster Files

    - by David Allan
    Deep in the trenches of ODI development I raised my head above the parapet to read a few odds and ends and then think why don’t they know this? Such as this article here – in the past customers (see forum) were told to use a staging route which has a big overhead for large files. This KM is an example of the great extensibility capabilities of ODI, its quite simple, just a new KM that; improves the out of the box experience – just build the mapping and the appropriate KM is used improves out of the box performance for file to file data movement. This improvement for out of the box handling for File to File data integration cases (from the 11.1.1.5.2 companion CD and on) dramatically speeds up the file integration handling. In the past I had seem some consultants write perl versions of the file to file integration case, now Oracle ships this KM to fill the gap. You can find the documentation for the IKM here. The KM uses pure java to perform the integration, using java.io classes to read and write the file in a pipe – it uses java threading in order to super-charge the file processing, and can process several source files at once when the datastore's resource name contains a wildcard. This is a big step for regular file processing on the way to super-charging big data files using Hadoop – the KM works with the lightweight agent and regular filesystems. So in my design below transforming a bunch of files, by default the IKM File to File (Java) knowledge module was assigned. I pointed the KM at my JDK (since the KM generates and compiles java), and I also increased the thread count to 2, to take advantage of my 2 processors. For my illustration I transformed (can also filter if desired) and moved about 1.3Gb with 2 threads in 140 seconds (with a single thread it took 220 seconds) - by no means was this on any super computer by the way. The great thing here is that it worked well out of the box from the design to the execution without any funky configuration, plus, and a big plus it was much faster than before, So if you are doing any file to file transformations, check it out!

    Read the article

  • Team Foundation Service Preview now open for all!

    - by Tarun Arora
    The concept of TFS in the cloud was first presented back in early 2010, the product team worked hard to preview a constantly evolving solution at the BUILD conference last year and after having completed 31 Sprints today the preview service has been opened for all. No more invitation codes required, TfsPreview has been made public! “Since we announced the Team Foundation Service Preview at the BUILD conference last year, we’ve limited the on boarding of new customers by requiring invitation codes to create accounts.  The main reason for this has been to control the growth of the service to make sure it didn’t run away from us and end up with a bad user experience.  In this time period, we’ve continued to work on our infrastructure, performance, scale, monitoring, management and, of course, some cool new features like cloud build. ”   - Brian Harry Since the service is still in preview, it is free for all… If you haven’t, now is the best time to try out the offering. There is no fixed time line on how long before service becomes chargeable but the terms of service support production use, the service is reliable and the product team committed to carry all of your data forward into production. “The service will remain in “preview” for a while longer while we work through additional features like data portability, commercial terms, etc but the terms of service support production use, the service is reliable and we expect to carry all of your data forward into production. ”  - Brian Harry As of today it’s possible to use TFS Preview with VS 2012 RC, VS 2010 SP1, VS 2008 SP1, the service currently does not work with VS 2005, this is something the product team is actively working on. You can refer to Brian’s announcement blog post here, http://blogs.msdn.com/b/bharry/archive/2012/06/11/team-foundation-service-preview-is-public.aspx

    Read the article

  • How do I convince my boss that it's OK to use an application to access an outside website?

    - by Cyberherbalist
    That is, if you agree that it's OK. We have a need to maintain an accurate internal record of bank routing numbers, and my boss wants me to set up a process where once a week someone goes to the Federal Reserve's website, clicks on the link to get the list of routing numbers (or the link giving the updates since a particular date), and then manually uploads the resultant text file to an application that will make the update to our data. I told him that a manual process was not at all necessary, and that I could write a routine that would access the FED's routing numbers in the application that keeps our data updated, and put it on whatever schedule was appropriate. But he is greatly opposed to doing this, and calls it "hacking the Federal Reserve website." I think he's afraid that the FED is going to get after us. I showed him the FED's robot.txt file, and the only thing it forbids is an automated indexing of pages with extension .cf*: User-agent: * # applies to all robots Disallow: CF # disallow indexing of all CF* directories and pages This says nothing about accessing the same data automatically that you could access manually. Anyone have a good counterargument to the idea that we'd be "hacking" the FED?

    Read the article

  • Fleunt NHibernate not working outside of nunit test fixtures

    - by thorkia
    Okay, here is my problem... I created a Data Layer using the RTM Fluent Nhibernate. My create session code looks like this: _session = Fluently.Configure(). Database(SQLiteConfiguration.Standard.UsingFile("Data.s3db")) .Mappings( m => { m.FluentMappings.AddFromAssemblyOf<ProductMap>(); m.FluentMappings.AddFromAssemblyOf<ProductLogMap>(); }) .ExposeConfiguration(BuildSchema) .BuildSessionFactory(); When I reference the module in a test project, then create a test fixture that looks something like this: [Test] public void CanAddProduct() { var product = new Product {Code = "9", Name = "Test 9"}; IProductRepository repository = new ProductRepository(); repository.AddProduct(product); using (ISession session = OrmHelper.OpenSession()) { var fromDb = session.Get<Product>(product.Id); Assert.IsNotNull(fromDb); Assert.AreNotSame(fromDb, product); Assert.AreEqual(fromDb.Id, product.Id); } My tests pass. When I open up the created SQLite DB, the new Product with Code 9 is in it. the tables for Product and ProductLog are there. Now, when I create a new console application, and reference the same library, do something like this: Product product = new Product() {Code = "10", Name = "Hello"}; IProductRepository repository = new ProductRepository(); repository.AddProduct(product); Console.WriteLine(product.Id); Console.ReadLine(); It doesn't work. I actually get pretty nasty exception chain. To save you lots of head aches, here is the summary: Top Level exception: An invalid or incomplete configuration was used while creating a SessionFactory. Check PotentialReasons collection, and InnerException for more detail.\r\n\r\n The PotentialReasons collection is empty The Inner exception: The IDbCommand and IDbConnection implementation in the assembly System.Data.SQLite could not be found. Ensure that the assembly System.Data.SQLite is located in the application directory or in the Global Assembly Cache. If the assembly is in the GAC, use element in the application configuration file to specify the full name of the assembly. Both the unit test library and the console application reference the exact same version of System.Data.SQLite. Both projects have the exact same DLLs in the debug folder. I even tried copying SQLite DB the unit test library created into the debug directory of the console app, and removed the build schema lines and it still fails If anyone can help me figure out why this won't work outside of my unit tests it would be greatly appreciated. This crazy bug has me at a stand still.

    Read the article

  • C++ Iterator lifetime and detecting invalidation

    - by DK.
    Based on what's considered idiomatic in C++11: should an iterator into a custom container survive the container itself being destroyed? should it be possible to detect when an iterator becomes invalidated? are the above conditional on "debug builds" in practice? Details: I've recently been brushing up on my C++ and learning my way around C++11. As part of that, I've been writing an idiomatic wrapper around the uriparser library. Part of this is wrapping the linked list representation of parsed path components. I'm looking for advice on what's idiomatic for containers. One thing that worries me, coming most recently from garbage-collected languages, is ensuring that random objects don't just go disappearing on users if they make a mistake regarding lifetimes. To account for this, both the PathList container and its iterators keep a shared_ptr to the actual internal state object. This ensures that as long as anything pointing into that data exists, so does the data. However, looking at the STL (and lots of searching), it doesn't look like C++ containers guarantee this. I have this horrible suspicion that the expectation is to just let containers be destroyed, invalidating any iterators along with it. std::vector certainly seems to let iterators get invalidated and still (incorrectly) function. What I want to know is: what is expected from "good"/idiomatic C++11 code? Given the shiny new smart pointers, it seems kind of strange that STL allows you to easily blow your legs off by accidentally leaking an iterator. Is using shared_ptr to the backing data an unnecessary inefficiency, a good idea for debugging or something expected that STL just doesn't do? (I'm hoping that grounding this to "idiomatic C++11" avoids charges of subjectivity...)

    Read the article

  • OpenGL sprites and point size limitation

    - by Srdan
    I'm developing a simple particle system that should be able to perform on mobile devices (iOS, Andorid). My plan was to use GL_POINT_SPRITE/GL_PROGRAM_POINT_SIZE method because of it's efficiency (GL_POINTS are enough), but after some experimenting, I found myself in a trouble. Sprite size is limited (to usually 64 pixels). I'm calculating size using this formula gl_PointSize = in_point_size * some_factor / distance_to_camera to make particle sizes proportional to distance to camera. But at some point, when camera is close enough, problem with size limitation emerges and whole system starts looking unrealistic. Is there a way to avoid this problem? If no, what's alternative? I was thinking of manually generating billboard quad for each particle. Now, I have some questions about that approach. I guess minimum geometry data would be four vertices per particle and index array to make quads from these vertices (with GL_TRIANGLE_STRIP). Additionally, for each vertex I need a color and texture coordinate. I would put all that in an interleaved vertex array. But as you can see, there is much redundancy. All vertices of same particle share same color value, and four texture coordinates are same for all particles. Because of how glDrawArrays/Elements works, I see no way to optimise this. Do you know of a better approach on how to organise per-particle data? Should I use buffers or vertex arrays, or there is no difference because each time I have to update all particles' data. About particles simulation... Where to do it? On CPU or on a vertex processors? Something tells me that mobile's CPU would do it faster than it's vertex unit (at least today in 2012 :). So, any advice on how to make a simple and efficient particle system without particle size limitation, for mobile device, would be appreciated. (animation of camera passing through particles should be realistic)

    Read the article

  • Flex AS3: ProgressBar doesn't move

    - by jolierouge
    Hey All, I am a little stuck and need some advice/help. I have a progress bar: <mx:ProgressBar id="appProgress" mode="manual" width="300" label="{appProgressMsg}" minimum="0" maximum="100"/> I have two listener functions, one sets the progress, and one sets the appProgressMsg: public function incProgress(e:TEvent):void { var p:uint = Math.floor(e.data.number / e.data.total * 100); trace("Setting Perc." + p); appProgress.setProgress(p, 100); } public function setApplicationProgressStep(e:TEvent):void { trace("Setting step:" + e.data); appProgressMsg = e.data; } I want to reuse this progress bar alot. And not necessarily for ProgressEvents, but when going through steps. For instance, I loop over a bunch of database inserts, and want to undate the progress etc. Here is a sample: public function updateDatabase(result:Object):void { var total:int = 0; var i:int = 0; var r:SQLResult; trace("updateDatabase called."); for each (var table:XML in this.queries.elements("table")) { var key:String = table.attribute("name"); if (result[key]) { send(TEvent.UpdateApplicationProgressStep, "Updating " + key); i = 1; total = result[key].length; for each (var row:Object in result[key]) { //now, we need to see if we already have this record. send(TEvent.UpdateApplicationProgress, { number:i, total: total } ); r = this.query("select * from " + key + " where server_id = '" + row.id + "'"); if (r.data == null) { //there is no entry with this id, make one. this.query(table.insert, row); } else { //it exists, so let's update. this.update(key, row); } i++; } } } } Everything works fine. That is, the listener functions are called and I get trace output like: updateDatabase called. Setting step:Updating project Setting Perc 25 Setting Perc 50 Setting Perc 75 Setting Perc 100 The issue is, only the very last percent and step is shown. that is, when it's all done, the progress bar jumps to 100% and shows the last step label. Does anyone know why this is? Thanks in advance for any help, Jason

    Read the article

  • CodePlex Daily Summary for Saturday, November 10, 2012

    CodePlex Daily Summary for Saturday, November 10, 2012Popular ReleasesImageGlass: Version 1.5: http://i1214.photobucket.com/albums/cc483/phapsuxeko/ImageGlass/1.png v1.5.4401.3015 Thumbnail bar: Increase loading speed Thumbnail image with ratio Support personal customization: mouse up, mouse down, mouse hover, selected item... Scroll to show all items Image viewer Zoom by scroll, or selected rectangle Speed up loading Zoom to cursor point New background design and customization and others... v1.5.4430.483 Thumbnail bar: Auto move scroll bar to selected image Show / Hi...Building Windows 8 Apps with C# and XAML: Full Source Chapters 1 - 10 for Windows 8 Fix 002: This is the full source from all chapters of the book, compiled and tested on Windows 8 RTM. Includes: A fix for the Netflix example from Chapter 6 that was missing a service reference A fix for the ImageHelper issue (images were not being saved) - this was due to the buffer being inadequate and required streaming the writeable bitmap to a buffer first before encoding and savingmyCollections: Version 2.3.2.0: New in this version : Added TheGamesDB.net API for Games and NDS Added Support for Windows Media Center Added Support for myMovies Added Support for XBMC Added Support for Dune HD Added Support for Mede8er Added Support for WD HDTV Added Fast search options Added order by Artist/Album for music You can now create covers and background for games You can now update your ID3 tag with the info of myCollections Fixed several provider Performance improvement New Splash ...Draw: Draw 1.0: Drawing PadPdfReport: PdfReport 1.4: - Added Html Footer Template. - Added a new HeaderTemplate (HtmlHeader) to simplify creating the headers of pages and groups by using HTML. See HtmlHeader/HtmlHeaderPdfReport.cs sample for more info. - Added a new sample (HtmlCellTemplate/HtmlCellTemplatePdfReport.cs) to show how to use the HtmlField template for creating the custom cell templates. - Added transparency setting to DiagonalWatermark. See ProgressReportPdfReport for more info. - Added DynamicCompile/DynamicCompilePdfReport.cs sa...Player Framework by Microsoft: Player Framework for Windows 8 (v1.0): IMPORTANT: List of breaking changes from preview 7 Ability to move control panel or individual elements outside media player. more info... New Entertainment app theme for out of the box support for Windows 8 Entertainment app guidelines. more info... VSIX reference names shortened. Allows seeing plugin name from "Add Reference" dialog without resizing. FreeWheel SmartXML now supports new "Standard" event callback type. Other minor misc fixes and improvements ADDITIONAL DOWNLOADSSmo...WebSearch.Net: WebSearch.Net 3.1: WebSearch.Net is an open-source research platform that provides uniform data source access, data modeling, feature calculation, data mining, etc. It facilitates the experiments of web search researchers due to its high flexibility and extensibility. The platform can be used or extended by any language compatible for .Net 2 framework, from C# (recommended), VB.Net to C++ and Java. Thanks to the large coverage of knowledge in web search research, it is necessary to model the techniques and main...Umbraco CMS: Umbraco 4.10.0: NugetNuGet BlogRead the release blog post for 4.10.0. Whats newMVC support New request pipeline Many, many bugfixes (see the issue tracker for a complete list) Read the documentation for the MVC bits. Breaking changesWe have done all we can not to break backwards compatibility, but we had to do some minor breaking changes: Removed graphicHeadlineFormat config setting from umbracoSettings.config (an old relic from the 3.x days) U4-690 DynamicNode ChildrenAsList was fixed, altering it'...MySQL Tuner for Windows: 0.3: Welcome to the third beta of MySQL Tuner for Windows! This release fixes bugs in the displaying of numbers, and a crash that occurred due to the program incorrectly closing and disposing of resources, Be warned that there will be bugs in this release, so please do not use on production or critical systems. Do post details of issues found to the issue tracker, and I will endeavour to fix them, when I can. I would love to have your feedback, and if possible your support! Requirements Microso...SharePoint Manager 2013: SharePoint Manager 2013 Release ver 1.0.12.1106: SharePoint Manager 2013 Release (ver: 1.0.12.1106) is now ready for SharePoint 2013. The new version has an expanded view of the SharePoint object model and has been tested on SharePoint 2013 RTM. As a bonus, the new version is also available for SharePoint 2010 as a separate download.D3D9Client: D3D9Client R7: New release for Orbiter 2010-P1 - Added horizon/sun angle for night-lights into the configuration file (default 10deg) - Some runway lights related bugs are fixed - Added more configuration options for runway lightsFiskalizacija za developere: FiskalizacijaDev 1.2: Verzija 1.2. je, prije svega, odgovor na novu verziju Tehnicke specifikacije (v1.1.) koja je objavljena prije nekoliko dana. Pored novosti vezanih uz (sitne) izmjene u spomenutoj novoj verziji Tehnicke dokumentacije, projekt smo prošili sa nekim dodatnim feature-ima od kojih je vecina proizašla iz vaših prijedloga - hvala :) Novosti u v1.2. su: - Neusuglašenost zahtjeva (http://fiskalizacija.codeplex.com/workitem/645) - Sample projekt - iznosi se množe sa 100 (http://fiskalizacija.codeplex.c...MFCMAPI: October 2012 Release: Build: 15.0.0.1036 Full release notes at SGriffin's blog. If you just want to run the MFCMAPI or MrMAPI, get the executables. If you want to debug them, get the symbol files and the source. The 64 bit builds will only work on a machine with Outlook 2010 64 bit installed. All other machines should use the 32 bit builds, regardless of the operating system. Facebook BadgeJayData - The cross-platform HTML5 data-management library for JavaScript: JayData 1.2.3: JayData is a unified data access library for JavaScript to CRUD + Query data from different sources like OData, MongoDB, WebSQL, SqLite, HTML5 localStorage, Facebook or YQL. The library can be integrated with Knockout.js or Sencha Touch 2 and can be used on Node.js as well. See it in action in this 6 minutes video Sencha Touch 2 example app using JayData: Netflix browser. What's new in JayData 1.2.3 For detailed release notes check the release notes. TypeScript supportWrite your code in a ...MCEBuddy 2.x: MCEBuddy 2.3.7: Changelog for 2.3.7 (32bit and 64bit) 1. Improved performance of MP4 Fast and M4V Fast Profiles (no deinterlacing, removed --decomb) 2. Improved priority handling 3. Added support for Pausing and Resume conversions 4. Added support for fallback to source directory if network destination directory is unavailable 5. MCEBuddy now installs ShowAnalyzer during installation 6. Added support for long description atom in iTunesFoxyXLS: FoxyXLS Releases: Source code and samplesWindow Manager: Window Manager 1.0: First releaseProDinner - ASP.NET MVC Sample (EF4.4, N-Tier, jQuery): 8: update to ASP.net MVC Awesome 3.0 udpate to EntityFramework 4.4 update to MVC 4 added dinners grid on homepageASP.net MVC Awesome - jQuery Ajax Helpers: 3.0: added Grid helper added XML Documentation added textbox helper added Client Side API for AjaxList removed .SearchButton from AjaxList AjaxForm and Confirm helpers have been merged into the Form helper optimized html output for AjaxDropdown, AjaxList, Autocomplete works on MVC 3 and 4BlogEngine.NET: BlogEngine.NET 2.7: Cheap ASP.NET Hosting - $4.95/Month - Click Here!! Click Here for More Info Cheap ASP.NET Hosting - $4.95/Month - Click Here! If you want to set up and start using BlogEngine.NET right away, you should download the Web project. If you want to extend or modify BlogEngine.NET, you should download the source code. If you are upgrading from a previous version of BlogEngine.NET, please take a look at the Upgrading to BlogEngine.NET 2.7 instructions. If you looking for Web Application Project, ...New ProjectsAuthor-it Variables Report Plug-in: The Variable Information Plug-in is an Author-it plug-in that displays information about all the variables in a library.bobthebuilder: Bob the builderCalendarSite: This webapplicaton is a calendar and appoinement manager app.ControlIP: This project helps users make the control task in a projectDaily Math Training: Daily Math Training app is Windows 8 app for children.Diablo III: This Project will contain several Diablo III related Implementations (a Windows 8 App another API Implementation, maybe some Calculators...)gelform: Ejercicio UTNJHashWin: Calcular hash MD5 SHA1 SHA256 SHA512 de arquivos MultiTask .NET 4.0 C# Marron TFS: Marron TFSMiddle Tennessee State University Robotics Team: Home of the MTSU Raider Robotics team.myexam: examN2F Router: The Zibings, N2Framework, url mapping/router project.NZBMatrix Advanced Feed Reader: This project is simple filter for NZBMatrix website feeds to update user with new movies which falls into the predefined customizable search templates.OpenTrading: Projeto em VB .net 3.5 que utiliza Active Record para mapeamento do banco de dados diretamente na classe das entidades e NHibernate para persistência dos dadosProjeto Parque - PI: Este projeto é resultado de um trabalho que realizamos todos os semestres na faculdade. 4º BSI - Centro Universitário Senac. QuesSys: this project is about examinationsample of inheritance: this project a sample of inheritance between classes at C#SharePoint 2013 Search Query Tool: Tool to test and debug the Search REST API in SharePoint 2013. Issue GET and POST search and suggestions queries. Use also against SharePoint Online. SharePoint Metro Grid: SharePoint 2010/2013 web part that displays links in Metro Tile Grid format. Just like Windows 8. Includes 215 images in 7 different sizes. Or use your own.SQL Server Partitioned Table Framework: The SQL Server Partitioned Table Framework (PTF) consists of a set of T-SQL Procedures that ease the maintenance work associated with partitioned tables.String -By GigaPlux Inc.: String OS is a operating system that was developed using C# and is based on the cosmos project .It has a beautiful File System by GRUNT.svu cms: first cms with chat and message between user vConsole for Windows 8: Just a very simple little app to help IT folks stuck in vSphere 5.0 with Win 8 console issues until you get to the next version of vSphereWikiprep#: The Wikiprep program is completely written in C# and based on .Net 4.0 . It processes Wikipedia dumps.Windows 8 Lock Screen: Windows 8 Lock Screen is a app that works like the Windows 8 Lock Screenwsccproject: updates will be made in due time during development

    Read the article

  • ZFS Basics

    - by user12614620
    Stage 1 basics: creating a pool # zpool create $NAME $REDUNDANCY $DISK1_0..N [$REDUNDANCY $DISK2_0..N]... $NAME = name of the pool you're creating. This will also be the name of the first filesystem and, by default, be placed at the mountpoint "/$NAME" $REDUNDANCY = either mirror or raidzN, and N can be 1, 2, or 3. If you leave N off, then it defaults to 1. $DISK1_0..N = the disks assigned to the pool. Example 1: zpool create tank mirror c4t1d0 c4t2d0 name of pool: tank redundancy: mirroring disks being mirrored: c4t1d0 and c4t2d0 Capacity: size of a single disk Example 2: zpool create tank raidz c4t1d0 c4t2d0 c4t3d0 c4t4d0 c4t5d0 Here the redundancy is raidz, and there are five disks, in a 4+1 (4 data, 1 parity) config. This means that the capacity is 4 times the disk size. If the command used "raidz2" instead, then the config would be 3+2. Likewise, "raidz3" would be a 2+3 config. Example 3: zpool create tank mirror c4t1d0 c4t2d0 mirror c4t3d0 c4t4d0 This is the same as the first mirror example, except there are two mirrors now. ZFS will stripe data across both mirrors, which means that writing data will go a bit faster. Note: you cannot create a mirror of two raidzs. You can create a raidz of mirrors, but to do that requires trickery.

    Read the article

  • Speeding up templates in GAE-Py by aggregating RPC calls

    - by Sudhir Jonathan
    Here's my problem: class City(Model): name = StringProperty() class Author(Model): name = StringProperty() city = ReferenceProperty(City) class Post(Model): author = ReferenceProperty(Author) content = StringProperty() The code isn't important... its this django template: {% for post in posts %} <div>{{post.content}}</div> <div>by {{post.author.name}} from {{post.author.city.name}}</div> {% endfor %} Now lets say I get the first 100 posts using Post.all().fetch(limit=100), and pass this list to the template - what happens? It makes 200 more datastore gets - 100 to get each author, 100 to get each author's city. This is perfectly understandable, actually, since the post only has a reference to the author, and the author only has a reference to the city. The __get__ accessor on the post.author and author.city objects transparently do a get and pull the data back (See this question). Some ways around this are Use Post.author.get_value_for_datastore(post) to collect the author keys (see the link above), and then do a batch get to get them all - the trouble here is that we need to re-construct a template data object... something which needs extra code and maintenance for each model and handler. Write an accessor, say cached_author, that checks memcache for the author first and returns that - the problem here is that post.cached_author is going to be called 100 times, which could probably mean 100 memcache calls. Hold a static key to object map (and refresh it maybe once in five minutes) if the data doesn't have to be very up to date. The cached_author accessor can then just refer to this map. All these ideas need extra code and maintenance, and they're not very transparent. What if we could do @prefetch def render_template(path, data) template.render(path, data) Turns out we can... hooks and Guido's instrumentation module both prove it. If the @prefetch method wraps a template render by capturing which keys are requested we can (atleast to one level of depth) capture which keys are being requested, return mock objects, and do a batch get on them. This could be repeated for all depth levels, till no new keys are being requested. The final render could intercept the gets and return the objects from a map. This would change a total of 200 gets into 3, transparently and without any extra code. Not to mention greatly cut down the need for memcache and help in situations where memcache can't be used. Trouble is I don't know how to do it (yet). Before I start trying, has anyone else done this? Or does anyone want to help? Or do you see a massive flaw in the plan?

    Read the article

< Previous Page | 777 778 779 780 781 782 783 784 785 786 787 788  | Next Page >