Search Results

Search found 18331 results on 734 pages for 'blog engine'.

Page 271/734 | < Previous Page | 267 268 269 270 271 272 273 274 275 276 277 278  | Next Page >

  • PowerPivot Workshop in Frankfurt (and London early-bird expiring soon) #ppws

    - by Marco Russo (SQLBI)
    One week ago I described the PowerPivot Workshop Roadshow that we are planning in several European countries. The news today is that the Workshop will be in Frankfurt (Germany) on February 21-22, 2011 ! The registrations are open on www.powerpivotworkshop.com web site. The early-bird price for Frankfurt will expire on February 4, 2011. And if you are willing to attend the London date on Febrary 7-8, remember that early-bird price for London is going to expire on Monday (January 17) ! Save your money...(read more)

    Read the article

  • Cloud Computing Pricing - It's like a Hotel

    - by BuckWoody
    I normally don't go into the economics or pricing side of Distributed Computing, but I've had a few friends that have been surprised by a bill lately and I wanted to quickly address at least one aspect of it. Most folks are used to buying software and owning it outright - like buying a car. We pay a lot for the car, and then we use it whenever we want. We think of the "cloud" services as a taxi - we'll just pay for the ride we take an no more. But it's not quite like that. It's actually more like a hotel. When you subscribe to Azure using a free offering like the MSDN subscription, you don't have to pay anything for the service. But when you create an instance of a Web or Compute Role, Storage, that sort of thing, you can think of the idea of checking into a hotel room. You get the key, you pay for the room. For Azure, using bandwidth, CPU and so on is billed just like it states in the Azure Portal. so in effect there is a cost for the service and then a cost to use it, like water or power or any other utility. Where this bit some folks is that they created an instance, played around with it, and then left it running. No one was using it, no one was on - so they thought they wouldn't be charged. But they were. It wasn't much, but it was a surprise.They had the hotel room key, but they weren't in the room, so to speak. To add to their frustration, they had to talk to someone on the phone to cancel the account. I understand the frustration. Although we have all this spelled out in the sign up area, not everyone has the time to read through all that. I get that. So why not make this easier? As an explanation, we bill for that time because the instance is still running, and we have to tie up resources to be available the second you want them, and that costs money. As far as being able to cancel from the portal, that's also something that needs to be clearer. You may not be aware that you can spin up instances using code - and so cancelling from the Portal would allow you to do the same thing. Since a mistake in code could erase all of your instances and the account, we make you call to make sure you're you and you really want to take it down. Not a perfect system by any means, but we'll evolve this as time goes on. For now, I wanted to make sure you're aware of what you should do. By the way, you don't have to cancel your whole account not to be billed. Just delete the instance from the portal and you won't be charged. You don't have to call anyone for that. And just FYI - you can download the SDK for Azure and never even hit the online version at all for learning and playing around. No sign-up, no credit card, PO, nothing like that. In fact, that's how I demo Azure all the time. Everything runs right on your laptop in an emulated environment.  

    Read the article

  • Parameterize Charts using Excel Slicers in PowerPivot

    - by Marco Russo (SQLBI)
    One new nice feature of Excel 2010 is the Slicer. Usually, slicers are used to filter data in a PivotTable. But they might be also useful to parameterize an algorithm or a chart! We discussed this technique in our book , but Alberto Ferrari wrote a post that shows how to use this technique to allow the user to select two stocks that should be compared in an Excel Chart – as you might imagine, this will work also when you will publish the workbook on SharePoint! This is the result: Nice to see that...(read more)

    Read the article

  • OT: Improbable use for an iPad?

    - by merrillaldrich
    Here's an interesting tidbit: I have noticed an even more pronounced trend toward centralized or virtual workstations lately. Both my wife and I can sit at home, as we are now, at the dining room table and work on our laptops (exciting life, I know!) but both of us are not actually working locally on these machines. We are both remoting into machines at our respective workplaces. Hers is a desktop machine physically located at her desk, while mine is a virtual workstation in my company's data center...(read more)

    Read the article

  • SQL Saturday #146 : Nashua, NH

    - by AaronBertrand
    Today was SQL Saturday #146, put on by Mike Walsh, Jack Corbett, and a host of other volunteers and organizers. Scott and I missed the speaker dinner last night, but we headed up from Rhode Island at 6:00 AM and made a good day of it. We had lots of great conversations with both existing friends and potential customers. After lunch I participated in a panel discussion with Joey D'Antoni and Andrew Kelly, led my Mike. We basically talked about various things DBAs are responsible for - and ultimately...(read more)

    Read the article

  • Working with Reporting Services Filters–Part 5: OR Logic

    - by smisner
    When you combine multiple filters, Reporting Services uses AND logic. Once upon a time, there was actually a drop-down list for selecting AND or OR between filters which was very confusing to people because often it was grayed out. Now that selection is gone, but no matter. It wouldn’t help us solve the problem that I want to describe today. As with many problems, Reporting Services gives us more than one way to apply OR logic in a filter. If I want a filter to include this value OR that value for the same field, one approach is to set up the filter is to use the IN operator as I explained in Part 1 of this series. But what if I want to base the filter on two different fields? I  need a different solution. Using the AdventureWorksDW2008R2 database, I have a report that lists product sales: Let’s say that I want to filter this report to show only products that are Bikes (a category) OR products for which sales were greater than $1,000 in a year. If I set up the filter like this: Expression Data Type Operator Value [Category] Text = Bikes [SalesAmount]   > 1000 Then AND logic is used which means that both conditions must be true. That’s not the result I want. Instead, I need to set up the filter like this: Expression Data Type Operator Value =Fields!EnglishProductCategoryName.Value = "Bikes" OR Fields!SalesAmount.Value > 1000 Boolean = =True The OR logic needs to be part of the expression so that it can return a Boolean value that we test against the Value. Notice that I have used =True rather than True for the value. The filtered report appears below. Any non-bike product appears only if the total sales exceed $1,000, whereas Bikes appear regardless of sales. (You can’t see it in this screenshot, but Mountain-400-W Silver, 38 has sales of $923 in 2007 but gets included because it is in the Bikes category.)

    Read the article

  • Process Improvement and the Data Professional

    - by BuckWoody
    Don’t be afraid of that title – I’m not talking about Six Sigma or anything super-formal here. In many organizations, there are more folks in other IT roles than in the Data Professional area. In other words, there are more developers, system administrators and so on than there are the “DBA” role. That means we often have more to do than the time we need to do it. And, oddly enough, the first thing that is sacrificed is process improvement – the little things we need to do to make the day go faster in the first place. Then we get even more behind, the work piles up and…well, you know all about that. Earlier I challenged you to find 10-30 minutes a day to study. Some folks wrote back and asked “where do I start”? Well, why not be super-efficient and combine that time with learning how to make yourself more efficient? Try out a new scripting language, learn a new tool that automates things or find out ways others have automated their systems. In general, find out what you’re doing and how, and then see if that can be improved. It’s kind of like doing a performance tuning gig on yourself! If you’re pressed for time, look for bite-sized articles (like the ones I’ve done here for PowerShell and SQL Server) that you can follow in a “serial” fashion. In a short time you’ll have a new set of knowledge you can use to make your day faster. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • How do you deal with information overload?

    - by talonx
    There are so many (good) programming blogs out there. Some of them are consistent in what they post - as in they stick to programming topics. Some of them occasionally post on other unrelated topics. Also, not every programming post might be relevant to me. I might have read one good post once, and not wishing to miss any future good ones - subscribed to the blog. Subscribing to too many blog feeds usually leads to just skimming through all of them (which takes time as well). Another option might be to subscribe to aggregators, like Hacker News - but that too has a huge rate of link accumulation. How do you manage if you wish to keep up with the programming blogosphere and still maintain a good signal to noise ratio?

    Read the article

  • PASS Summit 2010 BI Workshop Feedbacks

    - by Davide Mauri
    As many other speakers already did, I’d like to share with the SQL Community the feedback of my PASS Summit 2010 Workshop. For those who were not there, my workshop was the “BI From A-Z” and the main objective of that workshop was to introduce people in the BI world not only from a technical point of view but insist a lot on the methodological and “engineered” approach. The will to put more engineering in the IT (and specially in the BI field) is something that has been growing stronger and stronger in me every day for of this last 5 years since is simply envy the fact that Airbus, Fincatieri, BMW (just to name a few) can create very complex machine “just” using putting people together and giving them some rules to follow (Of course this is an oversimplification but I think you get what I mean). The key point of engineering is that, after having defined the project blueprint, you have the possibility to give to a huge number of people, the rules to follow, the correct tools in order to implement the rules easily and semi-automatically and a way to measure the quality of the results. Could this be done in IT? Very big question, so my scope is now limited to BI. So that’s the main point of my workshop: and entry-level approach to BI (level was 200) in order to allow attendees to know the basics, to understand what tools they should use for which purpose and, above all, a set of rules and tools in order to make a BI solution scalable in terms of people working on it, while still maintaining a very good quality. All done not focusing only on the practice but explaining the theory behind to see how it can help *a lot* to build a correct solution despite the technology used to implement it. The idea is to reach a point where more then 70% of the work done to create a BI solution can be reused even if technologies changes. This is a very demanding challenge nowadays with the coming of Denali and its column-aligned storage and the shiny-new DAX language. As you may understand I was looking forward to get the feedback since you may have noticed that there’s a lot of “architectural” stuff in IT but really nothing on “engineering”. So how the session could be perceived by the attendees was really unknown to me. The feedback could also give a good indication if the need of more “engineering” is something I feel only by myself or if is something more broad. I’m very happy to be able to say that the overall score of 4.75 put my workshop in the TOP 20 session (on near 200 sessions)! Here’s the detailed evaluations: How would you rate the usefulness of the information presented in your day-to-day environment? 4.75 Answer:    # of Responses 3    1         4    12        5    42               How would you rate the Speaker's presentation skills? 4.80 Answer:    # of Responses 3 : 1         4 : 9         5 : 45               How would you rate the Speaker's knowledge of the subject? 4.95 Answer:    # of Responses 4 :  3         5 : 52               How would you rate the accuracy of the session title, description and experience level to the actual session? 4.75 Answer:    # of Responses 3 : 2         4 : 10         5 : 43               How would you rate the amount of time allocated to cover the topic/session? 4.44 Answer:    # of Responses 3 : 7         4 : 17        5 : 31               How would you rate the quality of the presentation materials? 4.62 Answer:    # of Responses 4 : 21        5 : 34 The comments where all very positive. Many of them asked for more time on the subject (or to shorten the very last topics). I’ll make treasure of these comments and will review the content accordingly. We’ll organize a two-day classes on this topic, where also more examples will be shown and some arguments will be explained more deeply. I’d just like to answer a comment that asks how much of what I shown is “universally applicable”. I can tell you that all of our BI project follow these rules and they’ve been applied to different markets (Insurance, Fashion, GDO) with different people and different teams and they allowed us to be “Adaptive” against the customer. The more the rules are well defined and the more there are tools that supports their implementations, the easier is to add new people to the project and to add or change solution features. Think of a car. How come that almost any mechanic can help you to fix a problem? Because they know what to expect. Because there a rules that allow them to identify the problem without having to discover each time how the car has been implemented build. And this is of course also true for car upgrades/improvements. Last but not least: thanks a lot to everyone for coming!

    Read the article

  • Printing PowerPoint slides in black and white

    - by John Paul Cook
    When I do SQL Server training, sometimes students want to print all of the PowerPoint slides and use them for note taking during class. For such purposes, the background is usually better off being suppressed. This is most efficiently done by changing Print Settings as shown below: Personally I recommend that people take notes directly in the slides instead of printing them. PowerPoint has a notes area. If you do want to print slides and notes, once again use the Print Settings to specify this:...(read more)

    Read the article

  • C++ and SDL resource management for 2D game

    - by KuruptedMagi
    My first question is about stateManagers. I do not use the singleton pattern (read many random posts with various reasons not to use it), I have gameStateManager which runs the pointer cCurrentGameState-render(), etc. I want to make a transitioning game, this engine should ideally cover both a platformer and a bird's eye RPG (with some recoding, I just mean the base engine), both of which will load different levels and events, such as world map, dungeon, shops, etc. So I then thought, rather then having to store all this data within all the states, I would break the engine into gameStates, and playStates... when gameState reaches gameStatePlay(), gameStatePlay simply runs the usual handleInput, logic, and render for the playStates, just as the low level gameStateManager does. This lets me store all the player data within the base playstate class without storing useless data in the gameStates. Now I have added a seperate mapEditor, which uses editorStates from gameStateEditor. Is this too much usage of the gameState concept? It seems to work pretty well for me, so I was wondering if I am too far off a common implementation of this. My second question is on image resources. I have my sprite class with nothing but static members, mainly loadImage, applySurface, and my screen pointer. I also have a map pairing imageName enums with actual SDL_Surface pointers, and one pairing clipNumber enums with a wrapper class for a vector of clips, so that each reference in the map can have different amounts of clips with different sizes. I thought it would be better to store all these images, and screen within one static body, since 20 different goblins all use the same sprite sheet, and all need to print to the same screen, and of course, this way I do not need to pass my screen reference to every little entity. The imageMap seems to work very well, I can even add the ability to search through the map at creation of entity type to see if a particular image at creation, creating if it doesnt exist, and destroying the image if the last entity that needs it was just destroyed. The vectored clip map however, seems to take too long to initialize, so if i run past the state that initializes them to fast, the game crashes <. Plus, the clip map call is half of this line =P SPRITE::applySurface( cEditorMap.cTiles[x][y].iX, cEditorMap.cTiles[x][y].iY, SPRITE::mImages[ IMAGE_TILEMAP ], SPRITE::screen, SPRITE::mImageClips[IMAGE_TILEMAP]->clips.at( cEditorMap.cTiles[x][y].iTileType ) ); Again, do I have the right idea? I like the imageMap, but am I better off with each entity storing its own clips? My last question is about collision detection. I only grasp the basics, will look at per-pixel and circular soon, but how can I determine which side the collision comes from with just the basic square collision detection, I tried breaking each entity into 4 collision zones, but that just gave me problems with walking through walls and the like <. Also, is per-pixel color collision a good way to decide what collision just occured, or is checking multiple colors for multiple entities too taxing each cycle?

    Read the article

  • Did You Know? I'm doing 3 more online seminars with SSWUG!

    - by Kalen Delaney
    As I told you in April , I recorded two more seminars with Stephen Wynkoop, on aspects of Query Processing. The first one will be broadcast on June 30 and the second on August 27. In between, we'll broadcast my Index Internals seminar, on July 23. Workshops can be replayed for up to a week after the broadcast, and you can even buy a DVD of the workshop. You can get more details by clicking on the workshop name, below, or check out the announcement on the SSWUG site at http://www.sswug.org/editorials/default.aspx?id=1948...(read more)

    Read the article

  • TechEd 2010 Day Four: Learning how to help others learn

    - by BuckWoody
    I do quite a few presentations, and teach at the University of Washington, and also teach other classes. But I'm always learning from others how to help others learn. At events like TechEd I have access to some of the best speakers around, so I try to find out what they do that works. I attended a great session by allen White, in which he demonstrated a set of PowerShell scripts. He said that Dan Jones of the Microsoft Manageability team told him while he demonstrated a script he needed to provide some visual way to represent the process. Allen used one of the oldest visualizations around - a flowchart. It was the first time I'd seen one used to illustrate a PowerShell script, and it was very effective. I'm totally stealing the idea. All of us are teachers - we help others on our team understand what we're up to. Make sure you make notes for what you find effective in dealing with you, and then meld that into your own way of teaching. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Taming Hopping Windows

    - by Roman Schindlauer
    At first glance, hopping windows seem fairly innocuous and obvious. They organize events into windows with a simple periodic definition: the windows have some duration d (e.g. a window covers 5 second time intervals), an interval or period p (e.g. a new window starts every 2 seconds) and an alignment a (e.g. one of those windows starts at 12:00 PM on March 15, 2012 UTC). var wins = xs     .HoppingWindow(TimeSpan.FromSeconds(5),                    TimeSpan.FromSeconds(2),                    new DateTime(2012, 3, 15, 12, 0, 0, DateTimeKind.Utc)); Logically, there is a window with start time a + np and end time a + np + d for every integer n. That’s a lot of windows. So why doesn’t the following query (always) blow up? var query = wins.Select(win => win.Count()); A few users have asked why StreamInsight doesn’t produce output for empty windows. Primarily it’s because there is an infinite number of empty windows! (Actually, StreamInsight uses DateTimeOffset.MaxValue to approximate “the end of time” and DateTimeOffset.MinValue to approximate “the beginning of time”, so the number of windows is lower in practice.) That was the good news. Now the bad news. Events also have duration. Consider the following simple input: var xs = this.Application                 .DefineEnumerable(() => new[]                     { EdgeEvent.CreateStart(DateTimeOffset.UtcNow, 0) })                 .ToStreamable(AdvanceTimeSettings.IncreasingStartTime); Because the event has no explicit end edge, it lasts until the end of time. So there are lots of non-empty windows if we apply a hopping window to that single event! For this reason, we need to be careful with hopping window queries in StreamInsight. Or we can switch to a custom implementation of hopping windows that doesn’t suffer from this shortcoming. The alternate window implementation produces output only when the input changes. We start by breaking up the timeline into non-overlapping intervals assigned to each window. In figure 1, six hopping windows (“Windows”) are assigned to six intervals (“Assignments”) in the timeline. Next we take input events (“Events”) and alter their lifetimes (“Altered Events”) so that they cover the intervals of the windows they intersect. In figure 1, you can see that the first event e1 intersects windows w1 and w2 so it is adjusted to cover assignments a1 and a2. Finally, we can use snapshot windows (“Snapshots”) to produce output for the hopping windows. Notice however that instead of having six windows generating output, we have only four. The first and second snapshots correspond to the first and second hopping windows. The remaining snapshots however cover two hopping windows each! While in this example we saved only two events, the savings can be more significant when the ratio of event duration to window duration is higher. Figure 1: Timeline The implementation of this strategy is straightforward. We need to set the start times of events to the start time of the interval assigned to the earliest window including the start time. Similarly, we need to modify the end times of events to the end time of the interval assigned to the latest window including the end time. The following snap-to-boundary function that rounds a timestamp value t down to the nearest value t' <= t such that t' is a + np for some integer n will be useful. For convenience, we will represent both DateTime and TimeSpan values using long ticks: static long SnapToBoundary(long t, long a, long p) {     return t - ((t - a) % p) - (t > a ? 0L : p); } How do we find the earliest window including the start time for an event? It’s the window following the last window that does not include the start time assuming that there are no gaps in the windows (i.e. duration < interval), and limitation of this solution. To find the end time of that antecedent window, we need to know the alignment of window ends: long e = a + (d % p); Using the window end alignment, we are finally ready to describe the start time selector: static long AdjustStartTime(long t, long e, long p) {     return SnapToBoundary(t, e, p) + p; } To find the latest window including the end time for an event, we look for the last window start time (non-inclusive): public static long AdjustEndTime(long t, long a, long d, long p) {     return SnapToBoundary(t - 1, a, p) + p + d; } Bringing it together, we can define the translation from events to ‘altered events’ as in Figure 1: public static IQStreamable<T> SnapToWindowIntervals<T>(IQStreamable<T> source, TimeSpan duration, TimeSpan interval, DateTime alignment) {     if (source == null) throw new ArgumentNullException("source");     // reason about DateTime and TimeSpan in ticks     long d = Math.Min(DateTime.MaxValue.Ticks, duration.Ticks);     long p = Math.Min(DateTime.MaxValue.Ticks, Math.Abs(interval.Ticks));     // set alignment to earliest possible window     var a = alignment.ToUniversalTime().Ticks % p;     // verify constraints of this solution     if (d <= 0L) { throw new ArgumentOutOfRangeException("duration"); }     if (p == 0L || p > d) { throw new ArgumentOutOfRangeException("interval"); }     // find the alignment of window ends     long e = a + (d % p);     return source.AlterEventLifetime(         evt => ToDateTime(AdjustStartTime(evt.StartTime.ToUniversalTime().Ticks, e, p)),         evt => ToDateTime(AdjustEndTime(evt.EndTime.ToUniversalTime().Ticks, a, d, p)) -             ToDateTime(AdjustStartTime(evt.StartTime.ToUniversalTime().Ticks, e, p))); } public static DateTime ToDateTime(long ticks) {     // just snap to min or max value rather than under/overflowing     return ticks < DateTime.MinValue.Ticks         ? new DateTime(DateTime.MinValue.Ticks, DateTimeKind.Utc)         : ticks > DateTime.MaxValue.Ticks         ? new DateTime(DateTime.MaxValue.Ticks, DateTimeKind.Utc)         : new DateTime(ticks, DateTimeKind.Utc); } Finally, we can describe our custom hopping window operator: public static IQWindowedStreamable<T> HoppingWindow2<T>(     IQStreamable<T> source,     TimeSpan duration,     TimeSpan interval,     DateTime alignment) {     if (source == null) { throw new ArgumentNullException("source"); }     return SnapToWindowIntervals(source, duration, interval, alignment).SnapshotWindow(); } By switching from HoppingWindow to HoppingWindow2 in the following example, the query returns quickly rather than gobbling resources and ultimately failing! public void Main() {     var start = new DateTimeOffset(new DateTime(2012, 6, 28), TimeSpan.Zero);     var duration = TimeSpan.FromSeconds(5);     var interval = TimeSpan.FromSeconds(2);     var alignment = new DateTime(2012, 3, 15, 12, 0, 0, DateTimeKind.Utc);     var events = this.Application.DefineEnumerable(() => new[]     {         EdgeEvent.CreateStart(start.AddSeconds(0), "e0"),         EdgeEvent.CreateStart(start.AddSeconds(1), "e1"),         EdgeEvent.CreateEnd(start.AddSeconds(1), start.AddSeconds(2), "e1"),         EdgeEvent.CreateStart(start.AddSeconds(3), "e2"),         EdgeEvent.CreateStart(start.AddSeconds(9), "e3"),         EdgeEvent.CreateEnd(start.AddSeconds(3), start.AddSeconds(10), "e2"),         EdgeEvent.CreateEnd(start.AddSeconds(9), start.AddSeconds(10), "e3"),     }).ToStreamable(AdvanceTimeSettings.IncreasingStartTime);     var adjustedEvents = SnapToWindowIntervals(events, duration, interval, alignment);     var query = from win in HoppingWindow2(events, duration, interval, alignment)                 select win.Count();     DisplayResults(adjustedEvents, "Adjusted Events");     DisplayResults(query, "Query"); } As you can see, instead of producing a massive number of windows for the open start edge e0, a single window is emitted from 12:00:15 AM until the end of time: Adjusted Events StartTime EndTime Payload 6/28/2012 12:00:01 AM 12/31/9999 11:59:59 PM e0 6/28/2012 12:00:03 AM 6/28/2012 12:00:07 AM e1 6/28/2012 12:00:05 AM 6/28/2012 12:00:15 AM e2 6/28/2012 12:00:11 AM 6/28/2012 12:00:15 AM e3 Query StartTime EndTime Payload 6/28/2012 12:00:01 AM 6/28/2012 12:00:03 AM 1 6/28/2012 12:00:03 AM 6/28/2012 12:00:05 AM 2 6/28/2012 12:00:05 AM 6/28/2012 12:00:07 AM 3 6/28/2012 12:00:07 AM 6/28/2012 12:00:11 AM 2 6/28/2012 12:00:11 AM 6/28/2012 12:00:15 AM 3 6/28/2012 12:00:15 AM 12/31/9999 11:59:59 PM 1 Regards, The StreamInsight Team

    Read the article

  • Upgraded Ubuntu 12.04 -> 12.10 and Drupal 7 site now get error

    - by Paul B
    I do all my Drupal 7 webdev and today I took advantage to upgrade my local WebDev box O/S to Ubuntu 12.10 from 12.04 and now I get the following errors for all my D7 projects on my localhost WebDev box (Ubuntu 12.10) It was all fine pre Ubuntu 12.04: Error The website encountered an unexpected error. Please try again later. Error message PDOException: SQLSTATE[42000]: Syntax error or access violation: 1286 Unknown storage engine 'InnoDB': SELECT expire, value FROM {semaphore} WHERE name = :name; Array ( [:name] => variable_init ) in lock_may_be_available() (line 167 of /var/www/jobsdaily/includes/lock.inc). A quick research and look into the phpmyadmin (3.4.11.1) and it seems InnoDB is an issue and when I click on a table to see data I get #1286 - Unknown storage engine 'InnoDB'. I have all my D7 sql backed up, but don't really want to go down the whole 'import' route, since it's 10 months work! Anyone had this issues and can anyone suggest fix ideas? Thanks

    Read the article

  • Rules of Holes #4 -Do You Have the BIG Picture?

    - by ArnieRowland
    Some folks decry the concept of being in a 'Hole'. For them, there is no such thing as 'Technical Debt', no such thing as maintaining weak and wobbly legacy code, no such thing as bad designs, no such thing as under-skilled or poorly performing co-workers, no such thing as 'fighting fires', or no such thing as management that doesn't share the corporate vision. They just go to work and do their job, keep their head down, and do whatever is required. Mostly. Until the day they are swallowed by the...(read more)

    Read the article

  • PASS Summit Survey

    - by andyleonard
    One item mentioned in the PASS Board Q & A was the PASS Summit survey, which should be in your Inbox today if you attended the PASS Summit 2013. Charlotte? Did you enjoy having the Summit in Charlotte? If so, the PASS Summit Survey is your primary and most effective means of communicating this fact to the PASS Board and PASS HQ. The same holds if you didn’t like the Summit in Charlotte. Would you rather have the Summit remain forever in Seattle? Would you like to see the Summit in Seattle two...(read more)

    Read the article

  • What are the most effective learning methods?

    - by BuckWoody
    After I got done speaking at the SQL Server 2008 R2 Launch Event yesterday I came back to the hotel room for a web-meeting with some of the other teachers at the University of Washington. As teachers we are always looking to improve the knowledge transfer to our students – and the Program Director found an interesting study that I thought I might share here. Below is an un-labeled chart showing the effectiveness of learning methods according to a recent study. At the top are the labels. (“Teaching” here means students teaching each other). Try the experiment we did: place the labels where you think they’ll go. I’ll post the completed chart tomorrow. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • SQL Saturday #156 : Providence, RI

    - by AaronBertrand
    Well, East Greenwich, RI. Another successful event, this one put on by John Miner, Brandon Leach, Steve Simon, Scott Abrants and a host of other folks. Several #SQLFamily friends in attendance as well: Grant Fritchey, Mike Walsh, Jack Corbett, Wayne Sheffield and others. I gave a session in the morning and then a session to cap off the day. Thanks to everyone who attended! The downloads are here: T-SQL : Bad Habits & Best Practices The Ins & Outs of Contained Databases...(read more)

    Read the article

  • DTLoggedExec 1.1.2008.4 Released!

    - by Davide Mauri
    Today I've relased the latest version of my DTExec replacement tool, DTLoggedExec. The main changes are the following: Used a new strategy for version numbers. Now it will follow the following pattern Major.Minor.TargetSQLServerVersion.Revision Added support for Auto Configurations Fixed a bug that reported incorrect number of errors and warnings to Log Providers Fixed a buf that prevented correct casting of values when using /Set and /Param options Errors and Warnings are now counted more precisely. Updated database and log import scripts to categorize logs by projects and sections. E.g.: Project: MyBIProject; Sections: Staging, Datawarehouse Removed unused report stored procedures from database Updated Samples: 12 samples are now available to show ALL DTLoggedExec features From this version only SSIS 2008 will be supported http://dtloggedexec.codeplex.com/releases/view/62218  It useful to say something more on a couple of specific points: From this version only SSIS 2008 will be supportedYes, Integration Services 2005 are not supported anymore. The latest version capable of running SSIS 2005 Packages is the 1.0.0.2. Updated database and log import scripts to categorize logs by projects and sectionsWhen you import a log file, you can now assign it to a Project and to a Section of that project. In this way it's easier to gather statistical information for an entire project or a subsection of it. This also allows to store logged data of package belonging to different projects in the same database. For example:  Updated SamplesA complete set of samples that shows how to use all DTLoggedExec features are now shipped with the product. Enjoy! Added support for Auto ConfigurationsThis point will have a post on its own, since it's quite important and is by far the biggest new feature introduced in this release. To explain it in a few words, I can just say that you don't need to waste time with complex DTS configuration files or options, since a package will configure itself automatically. You just need to write a single statement as a parameter for DTLoggedExec. This feature can simplify deployment *a lot* :)   I the next days I'll write the mentioned post on Auto-Configurations and i'll update the documentation available on theDTLoggedExec website:   http://dtloggedexec.davidemauri.it/MainPage.ashx

    Read the article

  • Oracle Social Network Developer Challenge: HarQen Nodal

    - by Kellsey Ruppel
    Originally posted by Jake Kuramoto on The Apps Lab blog. We wrapped the Oracle Social Network Developer Challenge last week at OpenWorld, and this week, I’ll be sharing all the entries. All the teams that entered our challenge did a ton of work and built really interesting integrations with Oracle Social Network, and I want to showcase their hard work and innovative ideas. Today, I give you Nodal from the HarQen (@harqen) team, Kris Gösser (@krisgosser), Jesse Vogt (@jesse_vogt) and Matt Stockton (@mstockton). The guys from HarQen built Nodal to provide a visual way to navigate your connections and conversations in Oracle Social Network and view relationships. Using Nodal, you can: Search through names and profiles in Oracle Social Network. Choose people and view their social graphs in a visually useful way. Expand nodes in the social graph and add that person’s social graph to the Nodal view for comparison. Move nodes around and lock them in place for easier viewing, using a physics engine for movement. Adjust the physics engine properties according to your viewing preferences. Select nodes in the social graph and create a conversation directly based on the selection. Here are some shots of Nodal. They really don’t do the physics engine justice, but maybe the guys at Harqen will post a video of what they did for your viewing pleasure. #gallery-1 { margin: auto; } #gallery-1 .gallery-item { float: left; margin-top: 10px; text-align: center; width: 33%; } #gallery-1 img { border: 2px solid #cfcfcf; } #gallery-1 .gallery-caption { margin-left: 0; }   Nodal’s visuals wowed the judges and the audience, and anyone with a decent-sized social network presence understands the need for good network visualization. Tools like Nodal allow you to discover hidden connections in your network and maximize the value of your weak ties and find mavens, a very important key to getting work done. Thanks to the HarQen team for participating in our challenge. We hope they had a good experience. Look for the details of the other entries this week.

    Read the article

  • Introducing the Hardware Sales Consultant (Presales) Team in Greece

    - by fboufis
    Hello World and welcome to the blog of the Oracle Hardware Presales Team in Athens.The team is responsible for a cluster of six (6) countries which includes Greece, Cyprus, Malta, The Former Yugoslav Republic of Macedonia, Albania and Kosovo.We handle the complete hardware & systems software portfolio, namely: Engineered Systems: Purpose-build and General-purpose solutions Servers: SPARC (M & T-Series) & x86 (X-Series) servers Operating Systems: Oracle Solaris & Oracle Linux Virtualization Technologies: Oracle VM, Solaris Zones & Dynamic Domains Storage: NAS (ZFSSA), SAN (Axiom) & Tape (StorageTek) Systems Software: High Availability (Solaris Cluster) & Systems Management (Ops Center) and a multitude of other products, all of which will be the main topic of our blog. We design and propose solutions based on these products and assist both customers and partners in integrating those solutions in existing datacenters.We will be happy to support you in your projects, provide information and discuss your business issues, so do not hesitate to contact us.Filippos Boufis – Oracle Hardware Principal Sales Consultant

    Read the article

  • Lightning whip particle effects

    - by Fibericon
    I'm currently using Mercury Particle Engine for the particle effects in my game, and I'm trying to create a sort of lightning whip - basically a lightning effect bound to a line that curves when the player moves. I know how to use the editor, and I have particle effects working in game. However, I'm completely lost as to where I should start for this specific particle effect. Perhaps if I could find the code for it in a different particle engine, I could convert it, but I can't seem to find that either. What I did find was a lot of tutorials for creating the lines associated with lightning programmatically, which doesn't help in this case because I don't want it to be rigid. Perhaps it would be more like some sort of laser beam with crackling effects around it? I'm running into a wall as far as even beginning to implement this goes.

    Read the article

  • SQL Server 2012 Service Pack 1 is available!

    - by AaronBertrand
    Microsoft has released Service Pack 1 for SQL Server 2012. Though so far it is only being made available for the x64 platform. You can read about the fixes in the following KB article: http://support.microsoft.com/kb/20120401 The build # appears to be the same as RTM: 11.0.2100. Which may mean it's just a ploy to bring on board all those folks who still think it's necessary to wait for SP1. I haven't tried installing it yet, as I'm still on travel, but please let me know if you have any issues....(read more)

    Read the article

  • Many-to-many relationships in pharmacology

    - by John Paul Cook
    When I was in my pharmacology class this morning, I realized that the instructor was presenting a classic relational database management system problem: the many-to-many relationship. He said that all of us in nursing school must know our drugs backwards and forwards. I know how to model that! There are so many things in both healthcare and higher education that could benefit from an appropriate application of technology. As a student, I'd like to be able to start with a drug, a disease, a name of...(read more)

    Read the article

< Previous Page | 267 268 269 270 271 272 273 274 275 276 277 278  | Next Page >