Search Results

Search found 12211 results on 489 pages for 'industry standard'.

Page 437/489 | < Previous Page | 433 434 435 436 437 438 439 440 441 442 443 444  | Next Page >

  • Managing Operational Risk of Financial Services Processes – part 2/2

    - by Sanjeev Sharma
    Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Calibri","sans-serif"; mso-bidi-font-family:"Times New Roman";} In my earlier blog post, I had described the factors that lead to compliance complexity of financial services processes. In this post, I will outline the business implications of the increasing process compliance complexity and the specific role of BPM in addressing the operational risk reduction objectives of regulatory compliance. First, let’s look at the business implications of increasing complexity of process compliance for financial institutions: · Increased time and cost of compliance due to duplication of effort in conforming to regulatory requirements due to process changes driven by evolving regulatory mandates, shifting business priorities or internal/external audit requirements · Delays in audit reporting due to quality issues in reconciling non-standard process KPIs and integrity concerns arising from the need to rely on multiple data sources for a given process Next, let’s consider some approaches to managing the operational risk of business processes. Financial institutions considering reducing operational risk of their processes, generally speaking, have two choices: · Rip-and-replace existing applications with new off-the shelf applications. · Extend capabilities of existing applications by modeling their data and process interactions, with other applications or user-channels, outside of the application boundary using BPM. The benefit of the first approach is that compliance with new regulatory requirements would be embedded within the boundaries of these applications. However pre-built compliance of any packaged application or custom-built application should not be mistaken as a one-shot fix for future compliance needs. The reason is that business needs and regulatory requirements inevitably out grow end-to-end capabilities of even the most comprehensive packaged or custom-built business application. Thus, processes that originally resided within the application will eventually spill outside the application boundary. It is precisely at such hand-offs between applications or between overlaying processes where vulnerabilities arise to unknown and accidental faults that potentially result in errors and lead to partial or total failure. The gist of the above argument is that processes which reside outside application boundaries, in other words, span multiple applications constitute a latent operational risk that spans the end-to-end value chain. For instance, distortion of data flowing from an account-opening application to a credit-rating system if left un-checked renders compliance with “KYC” policies void even when the “KYC” checklist was enforced at the time of data capture by the account-opening application. Oracle Business Process Management is enabling financial institutions to lower operational risk of such process ”gaps” for Financial Services processes including “Customer On-boarding”, “Quote-to-Contract”, “Deposit/Loan Origination”, “Trade Exceptions”, “Interest Claim Tracking” etc.. If you are faced with a similar challenge and need any guidance on the same feel free to drop me a note.

    Read the article

  • We have our standards, and we need them

    - by Tony Davis
    The presenter suddenly broke off. He was midway through his section on how to apply to the relational database the Continuous Delivery techniques that allowed for rapid-fire rounds of development and refactoring, while always retaining a “production-ready” state. He sighed deeply and then launched into an astonishing diatribe against Database Administrators, much of his frustration directed toward Oracle DBAs, in particular. In broad strokes, he painted the picture of a brave new deployment philosophy being frustratingly shackled by the relational database, and by especially by the attitudes of the guardians of these databases. DBAs, he said, shunned change and “still favored tools I’d have been embarrassed to use in the ’80′s“. DBAs, Oracle DBAs especially, were more attached to their vendor than to their employer, since the former was the primary source of their career longevity and spectacular remuneration. He contended that someone could produce the best IDE or tool in the world for Oracle DBAs and yet none of them would give a stuff, unless it happened to come from the “mother ship”. I sat blinking in astonishment at the speaker’s vehemence, and glanced around nervously. Nobody in the audience disagreed, and a few nodded in assent. Although the primary target of the outburst was the Oracle DBA, it made me wonder. Are we who work with SQL Server, database professionals or merely SQL Server fanbois? Do DBAs, in general, have an image problem? Is it a good career-move to be seen to be holding onto a particular product by the whites of our knuckles, to the exclusion of all else? If we seek a broad, open-minded, knowledge of our chosen technology, the database, and are blessed with merely mortal powers of learning, then we like standards. Vendors of RDBMSs generally don’t conform to standards by instinct, but by customer demand. Microsoft has made great strides to adopt the international SQL Standards, where possible, thanks to considerable lobbying by the community. The implementation of Window functions is a great example. There is still work to do, though. SQL Server, for example, has an unusable version of the Information Schema. One cast-iron rule of any RDBMS is that we must be able to query the metadata using the same language that we use to query the data, i.e. SQL, and we do this by running queries against the INFORMATION_SCHEMA views. Developers who’ve attempted to apply a standard query that works on MySQL, or some other database, but doesn’t produce the expected results on SQL Server are advised to shun the Standards-based approach in favor of the vendor-specific one, using the catalog views. The argument behind this is sound and well-documented, and of course we all use those catalog views, out of necessity. And yet, as database professionals, committed to supporting the best databases for the business, whatever they are now and in the future, surely our heart should sink somewhat when we advocate a vendor specific approach, to a developer struggling with something as simple as writing a guard clause. And when we read messages on the Microsoft documentation informing us that we shouldn’t rely on INFORMATION_SCHEMA to identify reliably the schema of an object, in SQL Server!

    Read the article

  • About the K computer

    - by nospam(at)example.com (Joerg Moellenkamp)
    Okay ? after getting yet another mail because of the new #1 on the Top500 list, I want to add some comments from my side: Yes, the system is using SPARC processor. And that is great news for a SPARC fan like me. It is using the SPARC VIIIfx processor from Fujitsu clocked at 2 GHz. No, it isn't the only one. Most people are saying there are two in the Top500 list using SPARC (#77 JAXA and #1 K) but in fact there are three. The Tianhe-1 (#2 on the Top500 list) super computer contains 2048 Galaxy "FT-1000" 1 GHz 8-core processors. Don't know it? The FeiTeng-1000 ? this proc is a 8 core, 8 threads per core, 1 ghz processor made in China. And it's SPARC based. By the way ? this sounds really familiar to me ? perhaps the people just took the opensourced UltraSPARC-T2 design, because some of the parameters sound just to similar. However it looks like that Tianhe-1 is using the SPARCs as input nodes and not as compute notes. No, I don't see it as the next M-series processor. Simple reason: You can't create SMP systems out of them ? it simply hasn't the functionality to do so. Even when there are multiple CPUs on a single board, they are not connected like an SMP/NUMA machine to a shared memory machine ? they are connected with the cluster interconnect (in this case the Tofu interconnect) and work like a large cluster. Yes, it has a lot of oomph in Linpack ? however I assume a lot came from the extensions to the SPARCv9 standard. No, Linpack has no relevance for any commercial workload ? Linpack is such a special load, that even some HPC people are arguing that it isn't really a good benchmark for HPC. It's embarrassingly parallel, it can work with relatively small interconnects compared to the interconnects in SMP systems (however we get in spheres SMP interconnects where a few years ago). Amdahl isn't hitting that hard when running Linpack. Yes, it's a good move to use SPARC. At some time in the last 10 years, there was an interesting twist in perception: SPARC was considered as proprietary architecture and x86 was the open architecture. However it's vice versa ? try to create a x86 clone and you have a lot of intellectual property problems, create a SPARC clone and you have to spend 100 bucks or so to get the specification from the SPARC Foundation and develop your own SPARC processor. Fujitsu is doing this for a long time now. So they had their own processor, their own know-how. So why was SPARC a good choice? Well ? essentially Fujitsu can do what they want with their core as it is their core, for example adding the extensions to the SPARCv9 chipset ? getting Intel to create extensions to x86 to help you with your product is a little bit harder. So Fujitsu could do they needed to do with their processor in order to create such a supercomputer. No, the K is really using no FPGA or GPU as accelerators. The K is really using the CPU at doing this job. Yes, it has a significantly enhanced FPU capable to execute 8 instructions in parallel. No, it doesn't run Solaris. Yes, it uses Linux. No, it doesn't hurt me ... as my colleague Roland Rambau (he knows a lot about HPC) said once to me ... it doesn't matter which OS is staying out of the way of the workload in HPC.

    Read the article

  • Free hosting solution for a very low-traffic website [duplicate]

    - by user966939
    This question already has an answer here: How to find web hosting that meets my requirements? 4 answers I run a very low-traffic website (about 40 users, basically all of which are daily active on the site). I don't see it changing anytime soon either, as there is no way to sign up on the site right now. Until now I have just been using a sub-directory on a friend's host (shared), to host the web site. But in only a few weeks from now, his subscription will end, and he has no plans on renewing it. So of course this means I'll have to move on to something else. But I don't think I'll find someone who'd be willing to share a... shared host with me again. And besides, the software used on that server is ancient (PHP 4.4.9 + MySQL 4.1.22). There's one obvious solution that comes to mind, I guess: choose a better host and pay for it myself. The problem here is that I have no real fixed income, as I'm only a student. So even if the pricing is dirt cheap, I just can't be certain I will be able to afford it, every single month, for... at least 2 years maybe? So I've looked at free hosting solutions instead. The least requirement I had was that it was completely free of ads. But no matter where I look, I always find something in a corner or two ("what can you expect from a free host?" - yeah I know, but I guess it was worth a shot). For example, on Byethost (one of the free hosts I tried), if you trigger a PHP error while error reporting is set to E_ALL, you will spawn some hidden ad... Besides Byethost, I've tried 000Webhost, x10Hosting, 2Freehosting/1Freehosting, Wink.ws, and they are only worse. Okay, I'm running low on ideas. But! What if I just hosted the site myself, on my own computer? That could work. I actually do have my computer on practically 24/7. But not really. Sometimes I need to reboot it, and sometimes we even have power outages. And what if the hardware needs an upgrade? It's not such a big deal for me if the site went down, because I know what's going on; but what about the users? If I do decide to host it myself, is there some way to show users an alternate page instead of them just seeing a generic "server not found" page in the browser when the site is not accessible? Or is there something I have been missing out on? Is there a different kind of "web hosting" solution out there that I haven't heard of? Here is what I'm really looking for: Free (as in, no costs) NO ads Bandwidth enough for a low-traffic forum with roughly 40 users (Semi-)Up-to-date PHP and MySQL (at least not older than a year) No standard (non-extension) PHP functions turned off - such as sleep() The mbstring extension is enabled Disk space: at least 5 MB At least one MySQL database Some bonus points would be: Max execution time of PHP scripts can be set Remote access to MySQL database What would be the best solution for me? Is there one?

    Read the article

  • Bios Memory settings and Virtualization + Ubuntu (Unofficial Answers Welcome) [closed]

    - by TardisGuy
    Attempting to optimize my (Main Windowless) Ubuntu system for my uses I will detail questions below, I understand this might be the wrong place to ask these questions. If so, my apologies and I thank you so much for your patience. Thanks to all the volenteers that have helped me learn ubuntu over the years (Since 5.10) This is a "short" list of questions I have been trying to figure out for some time. If you feel you can answer one but not another, that's already more than I could ask for. I have wrote this up in a format for easy navigation to important points Hopefully to less annoy your eyes. You're welcome :) or i'm sorry i annoy you. :( If you would be so kind, Please format answers as follows: question 1: _ _ _ _ _ or question 1-a: _ _ _ _ _ If you want to simply link me to relevant information, rather than type up something really detailed; that would be more than awesome! Memory Specific Questions Goal: Maximizing memory bandwith to better perform in Virtualization, and Large file compression. (Possible conflict?) Ganged vs Unganged "which is better?"** is relative, i know. But what about ganged vs unganged - With or without Bank/channel interleaving? a: Speculation - If i understand correctly, "channel interleaving has something to do with using both channels to read or write in a kind of "striping" pattern, as opposed to a standard half duplex operation.(probably wrong) but wouldn't ganged channels make this irrelevant? Memory Interleaving(bank). Does it have a down side? Does it require a ratio of clocks? (If I run 4x4gig ddr3) a. If im reading correctly(trying to learn), this is designed to spread operations between latency cycles to work around the higher latency of "normal" operation. b. However it seems to me that it has to be: divisible by fractions of a master clock? So if i run memory at 1333mhz, then the mean between 2 (physical) banks would operate every (roughly) 600Mhz? Warning! Possibly utter nonsense: (1333/2 interleaving to act like 1 memory module per 2 sticks of a total of 4 sticks, meaning 2x channels@4) c. which makes me wonder if there would be left over clock cycles the system would have to... "truncate/balance" or something? But I'm certain theres a feature somewhere i don't understand. Virtualization Questions AMD-V - Option of IOMMU Turned it on, why do i have extra option of "64MB"? If IOMMU is on, but "64MB" is "disabled", Is it on? (have scoured google, I still dont know) a. I think i understand that its supposed to (kind of) "set aside" a part of ram to act as a faster interactive zone for "stuff" (usb, Graphics, and... what?) b. I am using Nvidia graphics on AMD (Used kernel option "iommu=pt iommu=1, pt "passthrough"? No idea what they do, found it on google to solve boot up issue) c. Will this option help me use low latency sound hardware, like my midi keyboard? Can you recommend any additional tweaks? a. sysctl settings? b. swap settings? Grats, youve reached the end. Thanks for Reading.

    Read the article

  • Inside Red Gate - Exercising Externally

    - by simonc
    Over the next few weeks, we'll be performing experiments on SmartAssembly to confirm or refute various hypotheses we have about how people use the product, what is stopping them from using it to its full extent, and what we can change to make it more useful and easier to use. Some of these experiments can be done within the team, some within Red Gate, and some need to be done on external users. External testing Some external testing can be done by standard usability tests and surveys, however, there are some hypotheses that can only be tested by building a version of SmartAssembly with some things in the UI or implementation changed. We'll then be able to look at how the experimental build is used compared to the 'mainline' build, which forms our baseline or control group, and use this data to confirm or refute the relevant hypotheses. However, there are several issues we need to consider before running experiments using separate builds: Ideally, the user wouldn't know they're running an experimental SmartAssembly. We don't want users to use the experimental build like it's an experimental build, we want them to use it like it's the real mainline build. Only then will we get valid, useful, and informative data concerning our hypotheses. There's no point running the experiments if we can't find out what happens after the download. To confirm or refute some of our hypotheses, we need to find out how the tool is used once it is installed. Fortunately, we've applied feature usage reporting to the SmartAssembly codebase itself to provide us with that information. Of course, this then makes the experimental data conditional on the user agreeing to send that data back to us in the first place. Unfortunately, even though this does limit the amount of useful data we'll be getting back, and possibly skew the data, there's not much we can do about this; we don't collect feature usage data without the user's consent. Looks like we'll simply have to live with this. What if the user tries to buy the experiment? This is something that isn't really covered by the Lean Startup book; how do you support users who give you money for an experiment? If the experiment is a new feature, and the user buys a license for SmartAssembly based on that feature, then what do we do if we later decide to pivot & scrap that feature? We've either got to spend time and money bringing that feature up to production quality and into the mainline anyway, or we've got disgruntled customers. Either way is bad. Again, there's not really any good solution to this. Similarly, what if we've removed some features for an experiment and a potential new user downloads the experimental build? (As I said above, there's no indication the build is an experimental build, as we want to see what users really do with it). The crucial feature they need is missing, causing a bad trial experience, a lost potential customer, and a lost chance to help the customer with their problem. Again, this is something not really covered by the Lean Startup book, and something that doesn't have a good solution. So, some tricky issues there, not all of them with nice easy answers. Turns out the practicalities of running Lean Startup experiments are more complicated than they first seem! Cross posted from Simple Talk.

    Read the article

  • Why Apple’s New SDK Limitation is So Offensive

    - by TStewartDev
    I am not an Apple fanboy, nor have I ever been. However, I have owned a Mac, an iPod, and an iPhone in my lifetime, and for more than a decade, I have defended Apple against the untruths that the haters so enjoy spewing. I encouraged my wife to buy a MacBook when she needed a new laptop two years ago, and I often recommend them to my friends and relatives. I have proudly and happily used my first generation iPhone for nearly three years. Now, for the first time in well over ten years, I find myself ready to swear off Apple and encourage everyone I know to do the same. I was disappointed when Apple wouldn't allow native apps, but I still bought the iPhone. I've stomached their ambiguous app approval process even though it's apparent that Steve may just reject your app because he doesn't like you or feels threatened by you (I'm still lamenting the rejection of the Google Voice app). But, as a developer, I can no longer tolerate Apple's terms and the kind of totalitarian control they indicate Apple wants. In case you are not already familiar, Apple has dictated in their OS 4.0 SDK license agreement (the now infamous Section 3.3.1) that all apps developed for the iPhone must be coded in C, C++, or Objective C, and moreover, that using any cross-compiling platforms is a violation of the agreement. For those of you who aren't developers, let me try to illustrate why this angers those of us who are. Imagine you're a professional writer. You've had articles published in some journals and magazines, and you've got a couple popular books out there, too. You've got an idea for a new book, and so you take it to your publisher. Your publisher agrees that it's a good idea. "But," says the publisher, "we want to hold our books to a tighter standard so that our readers get the experience we want them to have. Therefore, from now on, all our writers may only use words from this list of the 10,000 most common English words. Furthermore, if you cite any other works or quote anyone, they must comply with that same list, or you'll have to rewrite the entire work as well in case our readers want to look up your citation." What do you do? If your work is a children's book, this probably isn't a big deal to you. If it's an autobiography, textbook, or even a novel, though, you're going to have a lot of trouble describing your content with only common words. It's going to take you longer to complete your book, too, since you'll be looking up less common words frequently to see if you can use them. You could always go to another publisher, but this one has the best ability to distribute your book. The next largest distributor can only do a quarter as much. You could abandon the project altogether, but then everyone loses. Isn't this a silly scenario? Who would put such a limitation on writers? Yet this is very much what Apple is doing. They are using their dominant position in the market to coerce developers to write their apps exclusively for the iPhone OS by making it too expensive to write for multiple platforms. It is at least a threefold attack, striking at Adobe who is set to release a compiler that lets Flash source be compiled to iPhone binaries; striking at Google whose Android platform stands the best chance at the moment of providing serious competition to the iPhone; and reinforcing their own strong position by keeping popular apps exclusively to iPhone. And while developers are already very upset about this, the sad fact is that most of us will cave and give in to Apple because consumers don't know any better. They will continue to buy Apple's toy forcing developers to play Apple's maniacal game in order to make any money, at least until Steve Jobs decides he doesn't like them or he intends to release a competing application (bye-bye OpenFeint). Apple has been kept in check on the desktop front by a very dominant Microsoft, but I'm afraid that their success with iPods, iTunes, and iPhones has created a monster that we may have to bear until it is slain by an anti-trust suit or dies with the retirement of Steve Jobs.

    Read the article

  • Monitoring Windows Azure Service Bus Endpoint with BizTalk 360?

    - by Michael Stephenson
    I'm currently working with a customer who is undergoing an initiative to expose some of their line of business applications to external partners and SAAS applications and as part of this we have been looking at using the Windows Azure Service Bus. For the first part of the project we were focused on some synchronous request response scenarios where an external application would use the Service Bus relay functionality to get data from some internal applications. When we were looking at the operational monitoring side of the solution it was obvious that although most of the normal server monitoring capabilities would be required for the on premise components we would have to look at new approaches to validate that the operation of the service from outside of the organization was working as expected. A number of months ago one of my colleagues Elton Stoneman wrote about an approach I have introduced with a number of clients in the past where we implement a diagnostics service in each service component we build. This service would allow us to make a call which would flex some of the working parts of the system to prove it was working within any SLA. This approach is discussed on the following article: http://geekswithblogs.net/EltonStoneman/archive/2011/12/12/the-value-of-a-diagnostics-service.aspx In our solution we wanted to take the same approach but we had to consider that the service clients were external to the service. We also had to consider that by going through Windows Azure Service Bus it's not that easy to make most of your standard monitoring solutions just give you an easy way to do this. In a previous article I have described how you can use BizTalk 360 to monitor things using a custom extension to the Web Endpoint Manager and I felt that we could use this approach to provide an excellent way to monitor our service bus endpoint. The previous article is available on the following link: http://geekswithblogs.net/michaelstephenson/archive/2012/09/12/150696.aspx   The Monitoring Solution BizTalk 360 currently has an easy way to hook up the endpoint manager to a url which it will then call and if a successful response is returned it then considers the endpoint to be in a healthy state. We would take advantage of this by creating an ASP.net web page which would be called by BizTalk 360 and behind this page we would implement the functionality to call the diagnostics service on our Service Bus endpoint. The ASP.net page could include logic to work out how to handle the response from the diagnostics service. For example if the overall result of the diagnostics service was successful but the call to the diagnostics service was longer than a certain amount of time then we could return an error and indicate the service is taking too long. The following diagram illustrates the monitoring pattern.   The diagnostics service which is hosted in the line of business application allows us to ping a simple message through the Azure Service Bus relay to the WCF services in the LOB application and we they get a response back indicating that the service is working fine. To implement this I used the exact same approach I described in my previous post to create a custom web page which calls the diagnostics service and then it would return an HTTP response code which would depend on the error condition returned or a 200 if it was successful. One of the limitations of this approach is that the competing consumer pattern for listening to messages from service bus means that you cannot guarantee which server would process your diagnostics check message but with BizTalk 360 you could simply add multiple endpoint checks so that it could access the individual on-premise web servers directly to ensure that each server is working fine and then check that messages can also be processed through the cloud. Conclusion It took me about 15 minutes to get a proof of concept of this up and running which was able to monitor our web services which had been exposed via Windows Azure Service Bus. I was then able to inherit all of the monitoring benefits of BizTalk 360 to provide an enterprise class monitoring solution for our cloud enabled API.

    Read the article

  • Some PowerShell goodness

    - by KyleBurns
    Ever work somewhere where processes dump files into folders to maintain an archive?  Me too and Windows Explorer hates it.  Very often I find myself needing to organize these files into subfolders so that I can go after files without locking up Windows Explorer and my answer used to be to write a program in something like C# to do the job.  These programs will typically enumerate the files in a folder and move each file to a subdirectory named based on a datestamp.  The last such program I wrote had to use lower-level Win32 API calls to perform the enumeration because it appears the standard .Net calls make use of the same method of enumerating the directories that Windows Explorer chokes on when dealing with a large number of entries in a particular directory, so a simple task was accomplished with a lot of code. Of course, this little utility was just something I used to make my life easier and "not a production app", so it was in my local source folder and not source control when my hard drive died.  So... I was getting ready to re-create it and thought it might be a good idea to play with PowerShell a bit - something I had been wanting to do but had not yet met a requirement to make me do it.  The resulting script was amazingly succinct and even building the flexibility for parameterization and adding line breaks for readability was only about 25 lines long.  Here's the code with discussion following: param(     [Parameter(         Mandatory = $false,         Position = 0,         HelpMessage = "Root of the folders or share to archive.  Be sure to end with appropriate path separator"     )]     [String] $folderRoot="\\fileServer\pathToFolderWithLotsOfFiles\",       [Parameter(         Mandatory = $false,         Position = 1     )]     [int] $days = 1 ) dir $folderRoot|?{(!($_.PsIsContainer)) -and ((get-date) - $_.lastwritetime).totaldays -gt $days }|%{     [string]$year=$([string]$_.lastwritetime.year)     [string]$month=$_.lastwritetime.month     [string]$day=$_.lastwritetime.day     $dir=$folderRoot+$year+"\"+$month+"\"+$day     if(!(test-path $dir)){         new-item -type container $dir     }     Write-output $_     move-item $_.fullname $dir } The script starts by declaring two parameters.  The first parameter holds the path to the folder that I am going to be sorting into subdirectories.  The path separator is intended to be included in this argument because I didn't want to mess with determining whether this was local or UNC and picking the right separator in code, but this could be easily improved upon using Path.Combine since PowerShell has access to the full framework libraries.  The second parameter holds a minimum age in days for files to be removed from the root folder.  The script then pipes the dir command through a query to include only files (by excluding containers) and of those, only entries that meet the age requirement based on the last modified datestamp.  For each of those, the datestamp is used to construct a folder name in the format YYYY\MM\DD (if you're in an environment where even a day's worth of files need further divided, you could make this more granular) and the folder is created if it does not yet exist.  Finally, the file is moved into the directory. One of the things that was really cool about using PowerShell for this task is that the new-item command is smart enough to create the entire subdirectory structure with a single call.  In previous code that I have written to do this kind of thing, I would have to test the entire tree leading down to the subfolder I want, leading to a lot of branching code that detracted from being able to quickly look at the code and understand the job it performs. Overall, I have to say I'm really pleased with what has been done making PowerShell powerful and useful.

    Read the article

  • Using HTML5 Today part 3&ndash; Using Polyfills

    - by Steve Albers
    Shims helps when adding semantic tags to older IE browsers, but there is a huge range of other new HTML5 features that having varying support on browsers.  Polyfills are JavaScript code and/or browser plug-ins that can provide older or less featured browsers with API support.  The best polyfills will detect the whether the current browser has native support, and only adds the functionality if necessary.  The Douglas Crockford JSON2.js library is an example of this approach: if the browser already supports the JSON object, nothing changes.  If JSON is not available, the library adds a JSON property in the global object. This approach provides some big benefits: It lets you add great new HTML5 features to your web sites sooner. It lets the developer focus on writing to the up-and-coming standard rather than proprietary APIs. Where most one-off legacy code fixes tends to break down over time, well done polyfills will stop executing over time (as customer browsers natively support the feature) meaning polyfill code may not need to be tested against new browsers since they will execute the native methods instead. Your should also remember that Polyfills represent an entirely separate code path (and sometimes plug-in) that requires testing for support.  Also Polyfills tend to run on older browsers, which often have slower JavaScript performance.  As a result you might find that performance on older browsers is not comparable. When looking for Polyfills you can start by checking the Modernizr GitHub wiki or the HTML5 Please site. For an example of a polyfill consider a page that writes a few geometric shapes on a <canvas> <script src="jquery-1.7.1.min.js"><script> <script> $(document).ready(function () { drawCanvas(); }); function drawCanvas() { var context = $("canvas")[0].getContext('2d'); //background context.fillStyle = "#8B0000"; context.fillRect(5, 5, 300, 100); // emptybox context.strokeStyle = "#B0C4DE"; context.lineWidth = 4; context.strokeRect(20, 15, 80, 80); // circle context.arc(160, 55, 40, 0, Math.PI * 2, false); context.fillStyle = "#4B0082"; context.fill(); </script>   The result is a simple static canvas with a box & a circle:   …to enable this functionality on a pre-canvas browser we can find a polyfill.  A check on html5please.com references  FlashCanvas.  Pull down the zip and extract the files (flashcanvas.js, flash10canvas.swf, etc) to a directory on your site.  Then based on the documentation you need to add a single line to your original HTML file: <!--[if lt IE 9]><script src="flashcanvas.js"></script><![endif]—> …and you have canvas functionality!  The IE conditional comments ensure that the library is only loaded in browsers where it is useful, improving page load & processing time. Like all Polyfills, you should test to verify the functionality matches your expectations across browsers you need to support.  For instance the Flash Canvas home page advertises 70% support of HTML5 Canvas spec tests.

    Read the article

  • ADO and Two Way Storage Tiering

    - by Andy-Oracle
    Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 We get asked the following question about Automatic Data Optimization (ADO) storage tiering quite a bit. Can you tier back to the original location if the data gets hot again? The answer is yes but not with standard Automatic Data Optimization policies, at least not reliably. That's not how ADO is meant to operate. ADO is meant to mirror a traditional view of Information Lifecycle Management (ILM) where data will be very volatile when first created, will become less active or cool, and then will eventually cease to be accessed at all (i.e. cold). I think the reason this question gets asked is because customers realize that many of their business processes are cyclical and the thinking goes that those segments that only get used during month end or year-end cycles could sit on lower cost storage when not being used. Unfortunately this doesn't fit very well with the ADO storage tiering model. ADO storage tiering is based on the amount of free and used space in the source tablespace. There are two parameters that control this behavior, TBS_PERCENT_USED and TBS_PERCENT_FREE. When the space in the tablespace exceeds the TBS_PERCENT_USED value then segments specified in storage tiering clause(s) can be moved until the percent of free space reaches the TBS_PERCENT_FREE value. It is worth mentioning that no checks are made for available space in the target tablespace. Now, it is certainly possible to create custom functions to control storage tiering, but this can get complicated. The biggest problem is insuring that there is enough space to move the segment back to tier 1 storage, assuming that that's the goal. This isn't as much of a problem when moving from tier 1 to tier 2 storage because there is typically more tier 2 storage available. At least that's the premise since it is supposed to be less costly, lower performing and higher capacity storage. In either case though, if there isn't enough space then the operation fails. In the case of a customized function, the question becomes do you attempt to free the space so the move can be made or do you just stop and return false so that the move cannot take place? This is really the crux of the issue. Once you cross into this territory you're really going to have to implement two-way hierarchical storage and the whole point of ADO was to provide automatic storage tiering. You're probably better off using heat map and/or business access requirements and building your own hierarchical storage management infrastructure if you really want two way storage tiering. /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:11.0pt; font-family:"Calibri","sans-serif"; mso-ascii-font-family:Calibri; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:"Times New Roman"; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Calibri; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:"Times New Roman"; mso-bidi-theme-font:minor-bidi;}

    Read the article

  • Modernizr Rocks HTML5

    - by Laila
    HTML5 is a moving target.  At the moment, we don't know what will be in future versions.  In most circumstances, this really matters to the developer. When you're using Adobe Air, you can be reasonably sure what works, what is there, and what isn't, since you have a version of the browser built-in. With Metro, you can assume that you're going to be using at least IE 10.   If, however,  you are using HTML5 in a web application, then you are going to rely heavily on Feature Detection.  Feature-Detection is a collection of techniques that tell you, via JavaScript, whether the current browser has this feature natively implemented or not Feature Detection isn't just there for the esoteric stuff such as  Geo-location,  progress bars,  <canvas> support,  the new <input> types, Audio, Video, web workers or storage, but is required even for semantic markup, since old browsers make a pigs ear out of rendering this.  Feature detection can't rely just on reading the browser version and inferring from that what works. Instead, you must use JavaScript to check that an HTML5 feature is there before using it.  The problem with relying on the user-agent is that it takes a lot of historical data  to work out what version does what, and, anyway, the user-agent can be, and sometimes is, spoofed. The open-source library Modernizr  is just about the most essential  JavaScript library for anyone using HTML5, because it provides APIs to test for most of the CSS3 and HTML5 features before you use them, and is intelligent enough to alter semantic markup into 'legacy' 'markup  using shims  on page-load  for old browsers. It also allows you to check what video Codecs are installed for playing video. It also provides media queries  and conditional resource-loading (formerly YepNope.js.).  Generally, Modernizr gives you the choice of what you do about browsers that don't support the feature that you want. Often, the best choice is graceful degradation, but the resource-loading feature allows you to dynamically load JavaScript Shims to replace the standard API for missing or defective HTML5 functionality, called 'PolyFills'.  As the Modernizr site says 'Yes, not only can you use HTML5 today, but you can use it in the past, too!' The evolutionary progress of HTML5  requires a more defensive style of JavaScript programming where the programmer adopts a mindset of fearing the worst ( IE 6)  rather than assuming the best, whilst exploiting as many of the new HTML features as possible for the requirements of the site or HTML application.  Why would anyone want the distraction of developing their own techniques to do this when  Modernizr exists to do this for you? Laila

    Read the article

  • Why is x=x++ undefined?

    - by ugoren
    It's undefined because the it modifies x twice between sequence points. The standard says it's undefined, therefore it's undefined. That much I know. But why? My understanding is that forbidding this allows compilers to optimize better. This could have made sense when C was invented, but now seems like a weak argument. If we were to reinvent C today, would we do it this way, or can it be done better? Or maybe there's a deeper problem, that makes it hard to define consistent rules for such expressions, so it's best to forbid them? So suppose we were to reinvent C today. I'd like to suggest simple rules for expressions such as x=x++, which seem to me to work better than the existing rules. I'd like to get your opinion on the suggested rules compared to the existing ones, or other suggestions. Suggested Rules: Between sequence points, order of evaluation is unspecified. Side effects take place immediately. There's no undefined behavior involved. Expressions evaluate to this value or that, but surely won't format your hard disk (strangely, I've never seen an implementation where x=x++ formats the hard disk). Example Expressions x=x++ - Well defined, doesn't change x. First, x is incremented (immediately when x++ is evaluated), then it's old value is stored in x. x++ + ++x - Increments x twice, evaluates to 2*x+2. Though either side may be evaluated first, the result is either x + (x+2) (left side first) or (x+1) + (x+1) (right side first). x = x + (x=3) - Unspecified, x set to either x+3 or 6. If the right side is evaluated first, it's x+3. It's also possible that x=3 is evaluated first, so it's 3+3. In either case, the x=3 assignment happens immediately when x=3 is evaluated, so the value stored is overwritten by the other assignment. x+=(x=3) - Well defined, sets x to 6. You could argue that this is just shorthand for the expression above. But I'd say that += must be executed after x=3, and not in two parts (read x, evaluate x=3, add and store new value). What's the Advantage? Some comments raised this good point. It's not that I'm after the pleasure of using x=x++ in my code. It's a strange and misleading expression. What I want is to be able to understand complicated expressions. Normally, a complicated expression is no more than the sum of its parts. If you understand the parts and the operators combining them, you can understand the whole. C's current behavior seems to deviate from this principle. One assignment plus another assignment suddenly doesn't make two assignments. Today, when I look at x=x++, I can't say what it does. With my suggested rules, I can, by simply examining its components and their relations.

    Read the article

  • Generic Adjacency List Graph implementation

    - by DmainEvent
    I am trying to come up with a decent Adjacency List graph implementation so I can start tooling around with all kinds of graph problems and algorithms like traveling salesman and other problems... But I can't seem to come up with a decent implementation. This is probably because I am trying to dust the cobwebs off my data structures class. But what I have so far... and this is implemented in Java... is basically an edgeNode class that has a generic type and a weight-in the event the graph is indeed weighted. public class edgeNode<E> { private E y; private int weight; //... getters and setters as well as constructors... } I have a graph class that has a list of edges a value for the number of Vertices and and an int value for edges as well as a boolean value for whether or not it is directed. The brings up my first question, if the graph is indeed directed, shouldn't I have a value in my edgeNode class? Or would I just need to add another vertices to my LinkedList? That would imply that a directed graph is 2X as big as an undirected graph wouldn't it? public class graph { private List<edgeNode<?>> edges; private int nVertices; private int nEdges; private boolean directed; //... getters and setters as well as constructors... } Finally does anybody have a standard way of initializing there graph? I was thinking of reading in a pipe-delimited file but that is so 1997. public graph GenereateGraph(boolean directed, String file){ List<edgeNode<?>> edges; graph g; try{ int count = 0; String line; FileReader input = new FileReader("C:\\Users\\derekww\\Documents\\JavaEE Projects\\graphFile"); BufferedReader bufRead = new BufferedReader(input); line = bufRead.readLine(); count++; edges = new ArrayList<edgeNode<?>>(); while(line != null){ line = bufRead.readLine(); Object edgeInfo = line.split("|")[0]; int weight = Integer.parseInt(line.split("|")[1]); edgeNode<String> e = new edgeNode<String>((String) edges.add(e); } return g; } catch(Exception e){ return null; } } I guess when I am adding edges if boolean is true I would be adding a second edge. So far, this all depends on the file I write. So if I wrote a file with the following Vertices and weights... Buffalo | 18 br Pittsburgh | 20 br New York | 15 br D.C | 45 br I would obviously load them into my list of edges, but how can I represent one vertices connected to the other... so on... I would need the opposite vertices? Say I was representing Highways connected to each city weighted and un-directed (each edge is bi-directional with weights in some fictional distance unit)... Would my implementation be the best way to do that? I found this tutorial online Graph Tutorial that has a connector object. This appears to me be a collection of vertices pointing to each other. So you would have A and B each with there weights and so on, and you would add this to a list and this list of connectors to your graph... That strikes me as somewhat cumbersome and a little dismissive of the adjacency list concept? Am I wrong and that is a novel solution? This is all inspired by steve skiena's Algorithm Design Manual. Which I have to say is pretty good so far. Thanks for any help you can provide.

    Read the article

  • Using Private Extension Galleries in Visual Studio 2012

    - by Jakob Ehn
    Note: The installer and the complete source code is available over at CodePlex at the following location: http://inmetavsgallery.codeplex.com   Extensions and addins are everywhere in the Visual Studio ALM ecosystem! Microsoft releases new cool features in the form of extensions and the list of 3rd party extensions that plug into Visual Studio just keeps growing. One of the nice things about the VSIX extensions is how they are deployed. Microsoft hosts a public Visual Studio Gallery where you can upload extensions and make them available to the rest of the community. Visual Studio checks for updates to the installed extensions when you start Visual Studio, and installing/updating the extensions is fast since it is only a matter of extracting the files within the VSIX package to the local extension folder. But for custom, enterprise-specific extensions, you don’t want to publish them online to the whole world, but you still want an easy way to distribute them to your developers and partners. This is where Private Extension Galleries come into play. In Visual Studio 2012, it is now possible to add custom extensions galleries that can point to any URL, as long as that URL returns the expected content of course (see below).Registering a new gallery in Visual Studio is easy, but there is very little documentation on how to actually host the gallery. Visual Studio galleries uses Atom Feed XML as the protocol for delivering new and updated versions of the extensions. This MSDN page describes how to create a static XML file that returns the information about your extensions. This approach works, but require manual updates of that file every time you want to deploy an update of the extension. Wouldn’t it be nice with a web service that takes care of this for you, that just lets you drop a new version of your VSIX file and have it automatically detect the new version and produce the correct Atom Feed XML? Well search no more, this is exactly what the Inmeta Visual Studio Gallery Service does for you :-) Here you can see that in addition to the standard Online galleries there is an Inmeta Gallery that contains two extensions (our WIX templates and our custom TFS Checkin Policies). These can be installed/updated i the same way as extensions from the public Visual Studio Gallery. Installing the Service Download the installler (Inmeta.VSGalleryService.Install.msi) for the service and run it. The installation is straight forward, just select web site, application pool and (optional) a virtual directory where you want to install the service.   Note: If you want to run it in the web site root, just leave the application name blank Press Next and finish the installer. Open web.config in a text editor and locate the the <applicationSettings> element Edit the following setting values: FeedTitle This is the name that is shown if you browse to the service using a browser. Not used by Visual Studio BaseURI When Visual Studio downloads the extension, it will be given this URI + the name of the extension that you selected. This value should be on the following format: http://SERVER/[VDIR]/gallery/extension/ VSIXAbsolutePath This is the path where you will deploy your extensions. This can be a local folder or a remote share. You just need to make sure that the application pool identity account has read permissions in this folder Save web.config to finish the installation Open a browser and enter the URL to the service. It should show an empty Feed page:   Adding the Private Gallery in Visual Studio 2012 Now you need to add the gallery in Visual Studio. This is very easy and is done as follows: Go to Tools –> Options and select Environment –> Extensions and Updates Press Add to add a new gallery Enter a descriptive name, and add the URL that points to the web site/virtual directory where you installed the service in the previous step   Press OK to save the settings. Deploying an Extension This one is easy: Just drop the file in the designated folder! :-)  If it is a new version of an existing extension, the developers will be notified in the same way as for extensions from the public Visual Studio gallery: I hope that you will find this sever useful, please contact me if you have questions or suggestions for improvements!

    Read the article

  • Criminals and Other Illegal Characters

    - by Most Valuable Yak (Rob Volk)
    SQLTeam's favorite Slovenian blogger Mladen (b | t) had an interesting question on Twitter: http://www.twitter.com/MladenPrajdic/status/347057950470307841 I liked Kendal Van Dyke's (b | t) reply: http://twitter.com/SQLDBA/status/347058908801667072 And he was right!  This is one of those pretty-useless-but-sounds-interesting propositions that I've based all my presentations on, and most of my blog posts. If you read all the replies you'll see a lot of good suggestions.  I particularly like Aaron Bertrand's (b | t) idea of going into the Unicode character set, since there are over 65,000 characters available.  But how to find an illegal character?  Detective work? I'm working on the premise that if SQL Server will reject it as a name it would throw an error.  So all we have to do is generate all Unicode characters, rename a database with that character, and catch any errors. It turns out that dynamic SQL can lend a hand here: IF DB_ID(N'a') IS NULL CREATE DATABASE [a]; DECLARE @c INT=1, @sql NVARCHAR(MAX)=N'', @err NVARCHAR(MAX)=N''; WHILE @c<65536 BEGIN BEGIN TRY SET @sql=N'alter database ' + QUOTENAME(CASE WHEN @c=1 THEN N'a' ELSE NCHAR(@c-1) END) + N' modify name=' + QUOTENAME(NCHAR(@c)); RAISERROR(N'*** Trying %d',10,1,@c) WITH NOWAIT; EXEC(@sql); SET @c+=1; END TRY BEGIN CATCH SET @err=ERROR_MESSAGE(); RAISERROR(N'Ooops - %d - %s',10,1,@c,@err) WITH NOWAIT; BREAK; END CATCH END SET @sql=N'alter database ' + QUOTENAME(NCHAR(@c-1)) + N' modify name=[a]'; EXEC(@sql); The script creates a dummy database "a" if it doesn't already exist, and only tests single characters as a database name.  If you have databases with single character names then you shouldn't run this on that server. It takes a few minutes to run, but if you do you'll see that no errors are thrown for any of the characters.  It seems that SQL Server will accept any character, no matter where they're from.  (Well, there's one, but I won't tell you which. Actually there's 2, but one of them requires some deep existential thinking.) The output is also interesting, as quite a few codes do some weird things there.  I'm pretty sure it's due to the font used in SSMS for the messages output window, not all characters are available.  If you run it using the SQLCMD utility, and use the -o switch to output to a file, and -u for Unicode output, you can open the file in Notepad or another text editor and see the whole thing. I'm not sure what character I'd recommend to answer Mladen's question.  I think the standard tab (ASCII 9) is fine.  There's also several specific separator characters in the original ASCII character set (decimal 28-31). But of all the choices available in Unicode whitespace, I think my favorite would be the Mongolian Vowel Separator.  Or maybe the zero-width space. (that'll be fun to print!)  And since this is Mladen we're talking about, here's a good selection of "intriguing" characters he could use.

    Read the article

  • Validation and authorization in layered architecture

    - by SonOfPirate
    I know you are thinking (or maybe yelling), "not another question asking where validation belongs in a layered architecture?!?" Well, yes, but hopefully this will be a little bit of a different take on the subject. I am a firm believer that validation takes many forms, is context-based and varies at each level of the architecture. That is the basis for the post - helping to identify what type of validation should be performed in each layer. In addition, a question that often comes up is where authorization checks belong. The example scenario comes from an application for a catering business. Periodically during the day, a driver may turn in to the office any excess cash they've accumulated while taking the truck from site to site. The application allows a user to record the 'cash drop' by collecting the driver's ID, and the amount. Here's some skeleton code to illustrate the layers involved: public class CashDropApi // This is in the Service Facade Layer { [WebInvoke(Method = "POST")] public void AddCashDrop(NewCashDropContract contract) { // 1 Service.AddCashDrop(contract.Amount, contract.DriverId); } } public class CashDropService // This is the Application Service in the Domain Layer { public void AddCashDrop(Decimal amount, Int32 driverId) { // 2 CommandBus.Send(new AddCashDropCommand(amount, driverId)); } } internal class AddCashDropCommand // This is a command object in Domain Layer { public AddCashDropCommand(Decimal amount, Int32 driverId) { // 3 Amount = amount; DriverId = driverId; } public Decimal Amount { get; private set; } public Int32 DriverId { get; private set; } } internal class AddCashDropCommandHandler : IHandle<AddCashDropCommand> { internal ICashDropFactory Factory { get; set; } // Set by IoC container internal ICashDropRepository CashDrops { get; set; } // Set by IoC container internal IEmployeeRepository Employees { get; set; } // Set by IoC container public void Handle(AddCashDropCommand command) { // 4 var driver = Employees.GetById(command.DriverId); // 5 var authorizedBy = CurrentUser as Employee; // 6 var cashDrop = Factory.CreateCashDrop(command.Amount, driver, authorizedBy); // 7 CashDrops.Add(cashDrop); } } public class CashDropFactory { public CashDrop CreateCashDrop(Decimal amount, Employee driver, Employee authorizedBy) { // 8 return new CashDrop(amount, driver, authorizedBy, DateTime.Now); } } public class CashDrop // The domain object (entity) { public CashDrop(Decimal amount, Employee driver, Employee authorizedBy, DateTime at) { // 9 ... } } public class CashDropRepository // The implementation is in the Data Access Layer { public void Add(CashDrop item) { // 10 ... } } I've indicated 10 locations where I've seen validation checks placed in code. My question is what checks you would, if any, be performing at each given the following business rules (along with standard checks for length, range, format, type, etc): The amount of the cash drop must be greater than zero. The cash drop must have a valid Driver. The current user must be authorized to add cash drops (current user is not the driver). Please share your thoughts, how you have or would approach this scenario and the reasons for your choices.

    Read the article

  • No access to Samba shares

    - by koanhead
    I have three shared folders in my local home directory- that is to say, on my Ubuntu desktop's /home/me/. All were set up using "Sharing Options" in Nautilus' right-click menu. The standard "Music" and "Videos" folders are configured identically: the "Guest Access" box is checked, but the "Allow others to create and delete" is not. The third folder, called "shared", is configured to not allow Guest access but to allow others to modify files. I have not altered /etc/samba/smb.conf by hand, I have only used Sharing Options to create and modify these so-called "shares". My roommates have two Windows 7 computers and one Ubuntu Netbook Remix netbook. I have the aforementioned desktop machine and laptop running 10.04. None of these machines can access any of the shares. Attempts to access the Guest shares result in the message \\machine\directory is not accessible. The network name could not be found. This is the error message generated by a VM running Windows 2000. The other Windows machines generate a similar error. The Ubuntu laptop gives the error Unable to mount location: Failed to mount Windows share. Hurrah, once again, for informative error messages. That really helps a lot. When attempting to browse the folder called "shared" from the laptop, I'm confronted with a password dialog. This behavior is the same will all machines I've tried in the situation. On entering my username and password for the account to which the shares belong, the password dialog briefly disappears and is replaced with an identical dialog. No error message, useful or not, appears. When attempting to browse this folder with the VM, the outcome is the same except that the password dialog helpfully states "incorrect username or password". My assumption is that the username and password in question is that of the user which owns the shares. I have tried all other username and password combinations available in this context and the outcome is the same. I would like to be able to share files. Sharing them with Windows machines is a nice feature, or would be if it was available. Really I consider sharing files between two machines with the same version of the same operating system kind of a minimum condition for network usability. Samba last functioned reliably for me more than ten years ago. I have attempted to use it on and off since then with only intermittent success. Oh, and "Personal File Sharing" from the Preferences menu does not result in an entry in Places → Network → my-server. In fact, the old entry "MY-SERVER" goes away and is replaced by "koanhead's public files on my-server", which when I attempt to open it from the laptop gives a "DBus.Error.NoReply: Message did not receive a reply." I know I come here and gripe about Ubuntu a lot, but on the other hand I spend literally hours every day trying to fix things in Ubuntu. It's a good system which aspires to greatness, which is why things like this either Need to work; or Be adequately documented. Ideally both would be the case. Anyway, rant over. Hopefully someone will have some insight on this issue. Thanks all who bother to read this wall o'text for your time.

    Read the article

  • Add/ End Date Responsibility For Oracle FND User

    - by PRajkumar
    API - fnd_user_pkg.addresp Example -- -- ------------------------------------------------------------------------- -- Add/ End Date Responsibility to Oracle FND User -- ------------------------------------------------------------------------- DECLARE     lc_user_name                        VARCHAR2(100)    := 'PRAJ_TEST';     lc_resp_appl_short_name   VARCHAR2(100)    := 'FND';     lc_responsibility_key          VARCHAR2(100)    := 'APPLICATION_DEVELOPER';     lc_security_group_key        VARCHAR2(100)    := 'STANDARD';     ld_resp_start_date                DATE                        := TO_DATE('25-JUN-2012');     ld_resp_end_date                 DATE                        := NULL; BEGIN      fnd_user_pkg.addresp      (   username           => lc_user_name,         resp_app             => lc_resp_appl_short_name,         resp_key             => lc_responsibility_key,         security_group  => lc_security_group_key,         description         => NULL,         start_date           => ld_resp_start_date,         end_date            => ld_resp_end_date     );  COMMIT; EXCEPTION             WHEN OTHERS THEN                         ROLLBACK;                         DBMS_OUTPUT.PUT_LINE(SQLERRM); END; / SHOW ERR;  

    Read the article

  • Simple Navigation In Windows Phone 7

    - by PeterTweed
    Take the Slalom Challenge at www.slalomchallenge.com! When moving to the mobile platform all applications need to be able to provide different views.  Navigating around views in Windows Phone 7 is a very easy thing to do.  This post will introduce you to the simplest technique for navigation in Windows Phone 7 apps. Steps: 1.     Create a new Windows Phone Application project. 2.     In the MainPage.xaml file copy the following xaml into the ContentGrid Grid:             <StackPanel Orientation="Vertical" VerticalAlignment="Center"  >                 <TextBox Name="ValueTextBox" Width="200" ></TextBox>                 <Button Width="200" Height="30" Content="Next Page" Click="Button_Click"></Button>             </StackPanel> This gives a text box for the user to enter text and a button to navigate to the next page. 3.     Copy the following event handler code to the MainPage.xaml.cs file:         private void Button_Click(object sender, RoutedEventArgs e)         {             NavigationService.Navigate(new Uri( string.Format("/SecondPage.xaml?val={0}", ValueTextBox.Text), UriKind.Relative));         }   The event handler uses the NavigationService.Navigate() function.  This is what makes the navigation to another page happen.  The function takes a Uri parameter with the name of the page to navigate to and the indication that it is a relative Uri to the current page.  Note also the querystring is formatted with the value entered in the ValueTextBox control – in a similar manner to a standard web querystring. 4.     Add a new Windows Phone Portrait Page to the project named SecondPage.xaml. 5.     Paste the following XAML in the ContentGrid Grid in SecondPage.xaml:             <Button Name="GoBackButton" Width="200" Height="30" Content="Go Back" Click="Button_Click"></Button>   This provides a button to navigate back to the first page. 6.     Copy the following event handler code to the SecondPage.xaml.cs file:         private void Button_Click(object sender, RoutedEventArgs e)         {             NavigationService.GoBack();         } This tells the application to go back to the previously displayed page. 7.     Add the following code to the constructor in SecondPage.xaml.cs:             this.Loaded += new RoutedEventHandler(SecondPage_Loaded); 8.     Add the following loaded event handler to the SecondPage.xaml.cs file:         void SecondPage_Loaded(object sender, RoutedEventArgs e)         {             if (NavigationContext.QueryString["val"].Length > 0)                 MessageBox.Show(NavigationContext.QueryString["val"], "Data Passed", MessageBoxButton.OK);             else                 MessageBox.Show("{Empty}!", "Data Passed", MessageBoxButton.OK);         }   This code pops up a message box displaying either the text entered on the first page or the message “{Empty}!” if no text was entered. 9.     Run the application, enter some text in the text box and click on the next page button to see the application in action:   Congratulations!  You have created a new Windows Phone 7 application with page navigation.

    Read the article

  • No access to Samba shares

    - by koanhead
    I have three shared folders in my local home directory- that is to say, on my Ubuntu desktop's /home/me/. All were set up using "Sharing Options" in Nautilus' right-click menu. The standard "Music" and "Videos" folders are configured identically: the "Guest Access" box is checked, but the "Allow others to create and delete" is not. The third folder, called "shared", is configured to not allow Guest access but to allow others to modify files. I have not altered /etc/samba/smb.conf by hand, I have only used Sharing Options to create and modify these so-called "shares". My roommates have two Windows 7 computers and one Ubuntu Netbook Remix netbook. I have the aforementioned desktop machine and laptop running 10.04. None of these machines can access any of the shares. Attempts to access the Guest shares result in the message \\machine\directory is not accessible. The network name could not be found. This is the error message generated by a VM running Windows 2000. The other Windows machines generate a similar error. The Ubuntu laptop gives the error Unable to mount location: Failed to mount Windows share. Hurrah, once again, for informative error messages. That really helps a lot. When attempting to browse the folder called "shared" from the laptop, I'm confronted with a password dialog. This behavior is the same will all machines I've tried in the situation. On entering my username and password for the account to which the shares belong, the password dialog briefly disappears and is replaced with an identical dialog. No error message, useful or not, appears. When attempting to browse this folder with the VM, the outcome is the same except that the password dialog helpfully states "incorrect username or password". My assumption is that the username and password in question is that of the user which owns the shares. I have tried all other username and password combinations available in this context and the outcome is the same. I would like to be able to share files. Sharing them with Windows machines is a nice feature, or would be if it was available. Really I consider sharing files between two machines with the same version of the same operating system kind of a minimum condition for network usability. Samba last functioned reliably for me more than ten years ago. I have attempted to use it on and off since then with only intermittent success. Oh, and "Personal File Sharing" from the Preferences menu does not result in an entry in Places → Network → my-server. In fact, the old entry "MY-SERVER" goes away and is replaced by "koanhead's public files on my-server", which when I attempt to open it from the laptop gives a "DBus.Error.NoReply: Message did not receive a reply." I know I come here and gripe about Ubuntu a lot, but on the other hand I spend literally hours every day trying to fix things in Ubuntu. It's a good system which aspires to greatness, which is why things like this either Need to work; or Be adequately documented. Ideally both would be the case. Anyway, rant over. Hopefully someone will have some insight on this issue. Thanks all who bother to read this wall o'text for your time.

    Read the article

  • Hosted Monitoring

    - by Grant Fritchey
    The concept of using services to take the place of writing a lot of your own code goes way, way back in computing history. The fundamentals of the concept go back to the dawn of computing with places like IBM hosting time-shares for computing power that you could rent for short periods of time. But things really took off with the building of the Web. Now, all the growth with virtual machines, hosted machines, hosted services from vendors like Amazon and Microsoft, the need to keep all of your software locally on physical boxes is just going the way of the dodo. There will likely always be some pieces of software that you keep on machines on your property or on your person, but the concept of keeping fundamental services locally is going away. As someone put it to me once, if you were starting a business right now, would you bother setting up an Exchange server to manage your email or would you just go to one of the external mail services for everything? For most of us (who are not Exchange admins) the answer is pretty easy. With all this momentum to having external services manage more and more of the infrastructure that’s not business unique, why would you burn up a server and license instance setting up monitoring for your SQL Servers? Of course, some of you are dealing with hyper-sensitive data that might require, through law or treaty, that you lock it down and never expose it to the intertubes, but most of us are not. So, what if someone else took on the basic hassle of setting up monitoring on your systems? That’s what we’re working on here at Red Gate. Right now it’s a private test, but we’re growing it and developing it and it’ll be going to a public beta, probably (hopefully) this year. I’m running it on my machines right now. The concept is pretty simple. You put a relay on your server, poke a hole in your firewall for it, and we start monitoring your server using SQL Monitor. It’s actually shocking how easy it is to get going. You still have to adjust your alerting thresholds, but that’s a standard part of alerting. Your pain threshold and my pain threshold for any given alert may be different. But from there, we do all the heavy lifting, keeping your data online and available, providing you with access to the information about how your servers are behaving, everything. Maybe it’s just me, but I’m really excited by this. I think we’re getting to a place where we can really help the small and medium sized businesses get a monitoring solution in place, quickly and easily. All you crazy busy, and possibly accidental, DBAs and system admins finally can set up monitoring without taking all the time to configure systems, run installs, and all the rest. You just have to tweak your alerts and you’re ready to run. If you are interested in checking it out, you can apply for the closed beta through the Monitor web page.

    Read the article

  • Disabling the right-click sub menu using JQuery

    - by nikolaosk
    Recently I needed to disable the right-click contextual menu in an HTML page for a very simple HTML application I was creating for a friend.This is going to be a short post where I will demonstrate how to disable the right-click contextual menu.I will use the very popular JQuery Library. Please download the library (minified version) from http://jquery.com/downloadPlease find here all my posts regarding JQuery.In this hands-on example I will be using Expression Web 4.0.This application is not a free application. You can use any HTML editor you like.You can use Visual Studio 2012 Express edition. You can download it here. I am going to create a very simple HTML 5 page with some text and an image. The HTML markup for the page follows. <!DOCTYPE html><html lang="en">  <head>    <title>HTML 5, CSS3 and JQuery</title>        <meta http-equiv="Content-Type" content="text/html;charset=utf-8" >    <link rel="stylesheet" type="text/css" href="style.css">     <script type="text/javascript" src="jquery-1.8.2.min.js">        </script><script type="text/javascript"> (function ($) { $(document).bind('contextmenu', function () { return false;}); })(jQuery); </script>       </head>  <body>      <div id="header">      <h1>Learn cutting edge technologies</h1>      <h2>HTML 5, JQuery, CSS3</h2>    </div>      <figure>  <img src="html5.png" alt="HTML 5"></figure>        <div id="main">          <h2>HTML 5</h2>                        <article>          <p>            HTML5 is the latest version of HTML and XHTML. The HTML standard defines a single language that can be written in HTML and XML. It attempts to solve issues found in previous iterations of HTML and addresses the needs of Web Applications, an area previously not adequately covered by HTML.          </p>          </article>      </div>             </body>  </html> This is the JQuery code, I use (function ($) { $(document).bind('contextmenu', function () { return false;}); })(jQuery); I simply disable/cancel the contextmenu event.When I load the simple page on the browser and I right-click the context menu does not appear.Hope it helps!!!

    Read the article

  • Reading graph inputs for a programming puzzle and then solving it

    - by Vrashabh
    I just took a programming competition question and I absolutely bombed it. I had trouble right at the beginning itself from reading the input set. The question was basically a variant of this puzzle http://codercharts.com/puzzle/evacuation-plan but also had an hour component in the first line(say 3 hours after start of evacuation). It reads like this This puzzle is a tribute to all the people who suffered from the earthquake in Japan. The goal of this puzzle is, given a network of road and locations, to determine the maximum number of people that can be evacuated. The people must be evacuated from evacuation points to rescue points. The list of road and the number of people they can carry per hour is provided. Input Specifications Your program must accept one and only one command line argument: the input file. The input file is formatted as follows: the first line contains 4 integers n r s t n is the number of locations (each location is given by a number from 0 to n-1) r is the number of roads s is the number of locations to be evacuated from (evacuation points) t is the number of locations where people must be evacuated to (rescue points) the second line contains s integers giving the locations of the evacuation points the third line contains t integers giving the locations of the rescue points the r following lines contain to the road definitions. Each road is defined by 3 integers l1 l2 width where l1 and l2 are the locations connected by the road (roads are one-way) and width is the number of people per hour that can fit on the road Now look at the sample input set 5 5 1 2 3 0 3 4 0 1 10 0 2 5 1 2 4 1 3 5 2 4 10 The 3 in the first line is the additional component and is defined as the number of hours since the resuce has started which is 3 in this case. Now my solution was to use Dijisktras algorithm to find the shortest path between each of the rescue and evac nodes. Now my problem started with how to read the input set. I read the first line in python and stored the values in variables. But then I did not know how to store the values of the distance between the nodes and what DS to use and how to input it to say a standard implementation of dijikstras algorithm. So my question is two fold 1.) How do I take the input of such problems? - I have faced this problem in quite a few competitions recently and I hope I can get a simple code snippet or an explanation in java or python to read the data input set in such a way that I can input it as a graph to graph algorithms like dijikstra and floyd/warshall. Also a solution to the above problem would also help. 2.) How to solve this puzzle? My algorithm was: Find shortest path between evac points (in the above example it is 14 from 0 to 3) Multiply it by number of hours to get maximal number of saves Also the answer given for the variant for the input set was 24 which I dont understand. Can someone explain that also. UPDATE: I get how the answer is 14 in the given problem link - it seems to be just the shortest path between node 0 and 3. But with the 3 hour component how is the answer 24 UPDATE I get how it is 24 - its a complete graph traversal at every hour and this is how I solve it Hour 1 Node 0 to Node 1 - 10 people Node 0 to Node 2- 5 people TotalRescueCount=0 Node 1=10 Node 2= 5 Hour 2 Node 1 to Node 3 = 5(Rescued) Node 2 to Node 4 = 5(Rescued) Node 0 to Node 1 = 10 Node 0 to Node 2 = 5 Node 1 to Node 2 = 4 TotalRescueCount = 10 Node 1 = 10 Node 2= 5+4 = 9 Hour 3 Node 1 to Node 3 = 5(Rescued) Node 2 to Node 4 = 5+4 = 9(Rescued) TotalRescueCount = 9+5+10 = 24 It hard enough for this case , for multiple evac and rescue points how in the world would I write a pgm for this ?

    Read the article

  • Best way to process a queue in C# (PDF treatment)

    - by Bartdude
    First of all let me expose what I would like to do : I already dispose of a long-time running webapp developed in ASP.NET (C#) 2.0. In this app, users can upload standard PDF files (text+pics). The app is running in production on a Windows Server 2003 and has a dedicated database server (SQL server 2008) also running Windows Server 2003. I myself am a quite experienced web developer, but never actually programmed anything non-web (or at least nothing serious). I plan on adding a functionality to the webapp for which I would need a jpg snapshot of each page of the PDF. Creating these "thumbnails" isn't the big deal as such, I already do it inside my webapp using ghostscript. I've only done it on 1 page documents for now though, and the new functionality will need to process bigger documents. In order for this process to be transparent aswell for the admins as the final users, I would like to implement some kind of queue to delay the processing of the thumbnails. There again, no problem to create the queue, it will consist of records in a table, with enough info to find the pdf document back. Then I will need to process this queue, and that's were my interrogations start. Obviously the best solution to process it isn't an ASP script or so, so I will have to get out of my known environment. No problem, but I have no idea which direction to go. Therefore, a few questions : What should I develop ? I presumably need something that is "standby" on the server, runs when needed, then returns to idle state until further notice.Should I be looking into Windows service ? Is there another more appropriate type of project ? Depending on the first answer, what will be the approach ? Should I have somehow SQL server "tell" the program/service/... to process the queue, or should I have that program/service/... periodically check the state of the queue and treat new items. In both case, which functionality can I use ? we're not talking about hundreds of PDF a day (max 50 maybe), I can totally afford to treat the queue 1 item at a time. Can you confirm I don't have to look much further on threads and so ? (I found a lot of answers talking about threads in queue treatment, but it looks quite overkill for my needs) Maybe linked to the previous question : what about concurrent call to the program, whatever it is ? Let's suppose it is currently running, and a new record comes in the queue, what should be the behaviour ? I don't need much detailed answers and would already be happy with answers like "You can do the processing with a service, and yes it's possible to have sqlserver on machine A trigger a service start on machine B" or "You have to develop xxx and then use the scheduler to run it every xxx minutes". I don't mind reading articles and so, but I can hardly afford to spend too much time learning stuff to finally realize I went the wrong way for this project, so basically I'm trying to narrow down the scope of matters I need to investigate. Thanks for reading me, I hope I'll find some helping hands on here :-)

    Read the article

< Previous Page | 433 434 435 436 437 438 439 440 441 442 443 444  | Next Page >