Search Results

Search found 8219 results on 329 pages for 'less'.

Page 154/329 | < Previous Page | 150 151 152 153 154 155 156 157 158 159 160 161  | Next Page >

  • Returning a flexible datatype from a C++ function

    - by GavinH
    I'm developing for a legacy C++ application which uses ODBC for it's data access. Coming from a C# background, I really miss the ADO style of data access. I'm writing a wrapper (because we can't actually use ADO) to make our data access less painful. This means no char arrays, no manual text blob streaming, and no declaritive column binding. I'm struggling with how to store / return data values. In C# at least, you can declare an object and cast it to whatever (as long as the type is convertable). My current C++ solution is to use boost::any to store the data value in a custom DataColumnValue object. This class has conversion and assignment operators to the various types used in our app (more than 10). There's a bit of complexity here because if you store an int in the boost::any and try to boost::any_cast<long> you get a boost::bad_any_cast. Client objects shouldn't have to know how the value is stored internally. Does anyone have any experience trying to store / return values whose types are only known at runtime? Is there a better / cleaner way?

    Read the article

  • PHP Debugging

    - by Bob Porter
    Originally posted on: http://geekswithblogs.net/blogofbob/archive/2013/06/25/php-debugging.aspxI have been experimenting setting up a PHP development environment. I have been trying on Windows, Linux (Ubuntu) and Mac OS X. So far my favorite environment is on Mac OS X. I have tried a number of IDE's and debuggers as well.  IDE's Eclipse with the PDT Add On The PDT version of Eclipse Aptana Zend Netbeans  Debuggers Zend XDebug So far the only environments that I could get running quickly were Zend and Netbeans. Eclipse is a nightmare of versions and capabilities. I could only get Eclipse working well on Windows. On Ubuntu I was able to get the debugger working once. Thats it, one session, then it never worked again. I love the Zend tools and environment and it worked well everywhere I tried it, but it was beyond my budget.  Aptana also worked best on Windows, on Mac OS X it was fragile and I never could get debugging to work.  Netbeans worked first time, every time, every where. With one oddity, after several debugging sessions the debugger would refuse to connect. On every platform, I would end having to reboot to restore debugging, which would then work correctly for quite some time. I am sure I will discover that some process is hanging and there is a less intrusive way to clear the issue, but for now rebooting always works. In a future post I will go over how exactly I set my environment up, for now I have decided to stay with OS X. By the way, I did NOT use MAMP or the Zend Server, I stuck with PHP compiled and built from source, as well as Apache and MySQL installed locally. I use Homebrew as a package manager for OS X. I tried PORT but did not like the fact I had to sudo all the time to use it, and it installed things in /opt which I was not used to. Homebrew does sandbox the apps but it is nice enough to symlink them to their "normal" locations usually in /usr/local.

    Read the article

  • Why is a software development life-cycle so inefficient?

    - by user87166
    Currently the software development lifecycle followed in the IT company I work at is: The "Business" works with a solution manager to build a Business Requirement document The solution manager works with the Program manager to build a Functional Spec The PM works with the engineering lead to develop a release plan and with the engineering team to develop technical specifications If there are any clarifications required, developers contact the PM who contacts the solution manager who contacts the business and all the way back introducing a latency of nearly 24 hours and massive email chains for any clarifications By the time the tech spec is made, nearly 1 month has passed in back and forth Now, 2 weeks go to development while the test writes test cases Code is dropped formally to test, test starts raising bugs. Even if there is 1 root cause for 10 different issues, and its an easily fixed one, developers are not allowed to give fresh code to test for the next 1 week. After 2-3 such drops to test the code is given to the ops team as a "golden drop" ( 2 months passed from the beginning) Ops team will now deploy the code in a staging environment. If it runs stable for a week, it will be promoted to UAT and after 2 weeks of that it will be promoted to prod. If there are any bugs found here, well, applying for a visa requires less paperwork This entire process is followed even if a single SSRS report is to be released. How do other companies process such requirements? I'm wondering why, the business cannot just drop the requirements to developers, developers build and deploy to UAT themselves, expose it to the business who raise functional bugs and after fixing those promote to prod. (even for more complex stuff)

    Read the article

  • Fosfor EPUB Reader

    - by Geertjan
    Instead of creating a fullblown NetBeans Platform application for doing WYSIWYG editing for EPUB, similar to Sigil, I decided to focus purely on the very narrow scope of EPUB reading. The scope is narrower and, since the application will be a lot less ambitious and smaller, a pure JavaFX implementation makes sense. When you somehow get, e.g., buy, an EPUB file, you typically read it on a tablet or mobile device. However, some people in the world, e.g., me, still have laptops. Therefore, I'm creating a small JavaFX application that unzips EPUB files, into a temp directory, and then loads them into a JavaFX WebView. Arabic support: For an application like this, simplicity is the most important thing. Very few buttons, very few options, preferably no configuration of anything. Just let the user open the EPUB file and read it, that's it, nothing fancy. CSS stylesheets and images are correctly read. It's exactly what it looks like, a reader for EPUB files. The back and forward buttons are working and you can also switch to the table of contents. When it is complete, which it pretty much is right now, publishers of EPUB files can make this small app available from their site, to simplify life for their readers, since it will run easily and well on all operating systems.

    Read the article

  • Why is my USB data transfer so slow?

    - by Dave M G
    Whenever I do any kind of file transfer using USB, whether to a USB stick, or with my Android phone, or anything else, it is ridiculously slow. It says 59.8 KB/sec, which would be an awesome speed if this were 1991 and I was using a modem to dial up to my local BBS. Surely USB technology is better than that...? 37 seconds to move less data than the equivelent of 1 MP3 file? Also, regardless of what it says about speed and time, the reality is much, much slower. I routinely see it say something like "37 seconds left" and have to wait for minutes. Sometimes, if I want to move large amounts of files, it can say it will take 8 hours or more. Is this normal? My computer may not be the most awesome on the market, and about a year old, but it's an i5 with 4GB RAM and modern components, so surely this isn't the hardware's fault. What can I do to get better USB data transfer performance? Also, I did look at this question, but my newbie eyes don't see anything that look like an actual solution, just a lot of discussion about what transfer rates could or should be. Update: As requested in the comments, I've generated a whole bunch of output from the command line, and put it on Ubuntu Pastebin. Please see it here.

    Read the article

  • Bayesian content filter for vbulletin [on hold]

    - by mc0e
    I've been tasked with coming up with a tool to automatically flag some posts for moderator attention on a large vbulletin forum. It's not spam per se, but the task has a lot in common with the sort of handling that might be done by a spam protection plugin (a mod in vbulletin speak). There's only so much I can say, but the task does not involve bad users, so much as particular kinds of posts which the moderators need to be aware of. Filtering out user registrations and links is therefore not useful, and we are talking about posts by real human users. What I'm looking for is an existing bayesian classification plugin, or something that I can study to get an understanding of how to do the vbulletin side of the interface in order to build such a thing. Ie I'd need ways for moderators to list flagged posts, and to correct the classification of posts which have been mis-classified. Ideally I want a 3 way split with an "unsure" category in order to reduce what has to be reviewed to find any mis-classifications. Any pointers? I've searched around a bit, and so far what I've found has been more or less entirely targetted at intervening in sign-ups (mostly using stopforumspam), captchas, and use of external services like akismet which are spam specific. I'm also considering an external solution, which might be ableto be interfaced i

    Read the article

  • IDirect3DDevice9::GetRenderTargetData() returns no data

    - by P. Avery
    I've got a simple function to get the rendertarget data of an RT( w/default pool ). This particular RT has a resolution of 1x1( it's the 10'th and final mip of a texture ). Here is my code to get data for IDirect3DSurface9 *pTargetSurface: IDirect3DSurface9 *pSOS = NULL; pd3dDevice->CreateOffScreenPlainSurface( 1, 1, D3DFMT_A8R8G8B8, D3DPOOL_SYSTEMMEM, &pSOS, NULL ); // get residual energy if( FAILED( hr = pd3dDevice->GetRenderTargetData( pTargetSurface, pSOS ) ) ) { DebugStringDX( ClassName, "Failed to IDirect3DDevice9::GetRenderTargetData() at DownsampleArea()", __LINE__, hr ); goto Exit; } // lock surface if( FAILED( hr = pSOS->LockRect( &rct, NULL, D3DLOCK_READONLY ) ) ) { DebugStringDX( ClassName, "Failed to IDirect3DSurface9::LockRect() at DownsampleArea()", __LINE__, hr ); goto Exit; } // get residual energy from downsampled texture pByte = ( BYTE* )rct.pBits; D3DXVECTOR4 vEnergy; vEnergy.z = ( float )pByte[ 0 ] / 255.0f; vEnergy.y = ( float )pByte[ 1 ] / 255.0f; vEnergy.x = ( float )pByte[ 2 ] / 255.0f; vEnergy.w = ( float )pByte[ 3 ] / 255.0f; V( pSOS->UnlockRect() ); All formatting and settings are correct, directx in debug mode shows no errors... The problem is that the 4 bytes above are 0...I know this to be incorrect by using PIX to debug...PIX shows that RGB bytes are 0.078 and Alpah is 1. These values are not less than that which can be represented by a single byte( 1 / 255 ). Any ideas? Am I copying rendertarget data correctly?

    Read the article

  • I am confused between PHP and ASP.NET to choose as a career in Indian software development context

    - by Confused_Guy
    I need your help(specially from the software professionals of India). I have completed MCA in 2009. After that instead of joining software company I did a Teaching job nearby by home. In the mean time I prepared myself for public sector jobs(bank). I continued my job for 1 year more and left it in 2010. Now in 2012 ,I feel that I should have done the software jobs,so that I could earn my bread and butter and in the mean time I could have prepared for the job.Because,according to my qualification it will give me the best salary. Now I want to go back in software industries. Now all of them are asking for experiences.And I don't have any.....So which language should I learn? And what should I do,because I have two year gap. Some of my friends suggested me to go with PHP as its easier and quicker to get job in India. But Here the PHP guys are getting less salary as compared to ASP.NET. I am planning to begin with PHP and but is it possible to switch to ASP.NET after two years experience. JAVA: I know upto servlet & JSP. Which is nothing in current market. ASP.NET: I know the basics of asp.net upto database connection ie(Gridview). PHP: Only the basics. So what should I do now. Which is most demanding. Does PHP is good, I feel its more like JSP pages. Please guide me, All your suggestions are needed for me.

    Read the article

  • Why does there seem to be a lot of fear in choosing the "wrong" language to learn?

    - by Shewbox
    Perhaps its just me, but as a current CS student I have already come across many questions on this site and elsewhere about not just "Which language should I use for x?" but also "Does anyone still use language Y?" My first CS class was taught in Scheme, which, if I'm not mistaken, isn't used widely (at least in comparison to languages like Java, PHP, Python, etc). Many of my classmates balked at the idea of having to learn a language they would never have to use again, but I don't quite understand where so much of this fear of learning less popular languages comes from. No, I may not use Scheme in any job I get, but I certainly don't regret having learned to use it (albeit in a very beginner, not very in-depth manner in that one semester). I am taking a search engines class this semester, which is done in Perl and again I am seeing classmates complaining about the language choice. I can understand having a favorite language and disliking others but why do some get worked up over learning it in the first place? Can you really learn the "wrong" language? Isn't learning something like Scheme or Haskell good mental exercise if nothing else, and useful at least to exposure to different ways of solving problems?

    Read the article

  • QueryUnit 0.0.0.8 – Trust No One

    - by Davide Mauri
    Yesterday I’ve release an updated version of QueryUnit, the version 0.0.0.8. QueryUnit now supports AreNotEqual, Greater, and Less assertions and is more capable of managing strings results. I must say that I cannot live anymore without a proper Unit Testing of a BI solution. Just yesterday happened that one of the unit tests at a customer site failed showing a subtle situation where the release of a new version of custom application would have corrupted the source of BI data with a very low chance that someone would have noticed it before several days. It may happen when you have more the 15 systems that handles the data needed by your BI solution. The key message of this situation is “Trust No One”: if your data hasn’t passed quality testing it’s not trustable. Period. QueryUnit is now officialy an hero :) No superpowers still, but useful above all. http://queryunit.codeplex.com/ Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • What is the value of checking in failing unit tests?

    - by user20194
    While there are ways of keeping unit tests from being executed, what is the value of checking in failing unit tests? I will use a simple example: Case Sensitivity. The current code is case sensitive. A valid input into the method is "Cat" and it would return an enum of Animal.Cat. However, the desired functionality of the method should not be case sensitive. So if the method described was passed "cat" it could possibly return something like Animal.Null instead of Animal.Cat and the unit test would fail. Though a simple code change would make this work, a more complex issue may take weeks to fix, but identifying the bug with a unit test could be a less complex task. The application currently being analyzed has 4 years of code that "works". However, recent discussions regarding unit tests have found flaws in the code. Some just need explicit implementation documentation (ex. case sensitive or not), or code that does not execute the bug based on how it is currently called. But unit tests can be created executing specific scenarios that will cause the bug to be seen and are valid inputs. What is the value of checking in unit tests that exercise the bug until someone can get around to fixing the code? Should this unit test be flagged with ignore, priority, category etc, to determine whether a build was successful based on tests executed? Eventually the unit test should be created to execute the code once someone fixes it. On one hand it shows that identified bugs have not been fixed. On the other, there could be hundreds of failed unit tests showing up in the logs and weeding through the ones that should fail vs. failures due to a code check-in would be difficult to find.

    Read the article

  • two thoughts about career excellence

    - by john.rose
    I love Dickens, warts and all. Sometimes he is sententious, and (like the mediocre modern I am) at such points I am willing to listen non-ironically. This bit here struck me hard enough to stop and write it down: I mean a man whose hopes and aims may sometimes lie (as most men's sometimes do, I dare say) above the ordinary level, but to whom the ordinary level will be high enough after all if it should prove to be a way of usefulness and good service leading to no other. All generous spirits are ambitious, I suppose, but the ambition that calmly trusts itself to such a road, instead of spasmodically trying to fly over it, is of the kind I care for. It is Woodcourt's kind. (John Jarndyce to Esther Summerson, Bleak House, ch. 60) Woodcourt is, of course, one of the heroes of the story. It is a heroism that is attractive to me. Here is a similar idea, from the Screwtape Letters. In the satirically inverted logic of that book, the “Enemy” is God, the enemy of the devils but the author of good: The Enemy wants to bring the man to a state of mind in which he could design the best cathedral in the world, and know it to be the best, and rejoice in the, fact, without being any more (or less) or otherwise glad at having done it than he would be if it had been done by another. (C.S. Lewis, Screwtape Letters, ch. 14) Though I will be happy with a good Bazaar, I also dream of Cathedrals. Put whatever name you like on it, as long as I get some part in the fun of building a good one.

    Read the article

  • New whitepaper: Evolution from the Traditional Data Center to Exalogic: An Operational Perspective

    - by Javier Puerta
    IT organizations are struggling with the need to balance the day-to-day concerns of data center management against the business level requirements to deliver long-term value. This balancing act has proven difficult and inefficient: systems and application management tools are resource intensive and traditional infrastructure management architectures have developed over time on a project by project basis. These traditional management systems consist of multiple tools that require administrators to waste time performing too many steps to handle routine administrative tasks. Operational efficiency and agility in your enterprise are directly linked to the capabilities provided by the management layer across the entire stack, from the application, middleware, operating system, compute, network and storage. Only when this end to end capability is provided will we experience the full benefit of a scalable, efficient, responsive and secure datacenter. Managing Exalogic is substantially less complex and error prone than managing traditional systems built from individually sourced, multi-vendor components because Exalogic is designed to be administered and maintained as a single, integrated system (Figure 1). It is at the forefront of the industry-wide shift away from costly and inferior one-off platforms toward private clouds and Engineered Systems. Read the full whitepaper "Evolution from the Traditional Data Center to Exalogic: An Operational Perspective". Full document is available for download at the Exadata Partner Community Collaborative Workspace (for community members only - if you get an error message, please register for the Community first).

    Read the article

  • New Whitepaper: Evolution from the Traditional Data Center to Exalogic: An Operational Perspective

    - by Javier Puerta
    IT organizations are struggling with the need to balance the day-to-day concerns of data center management against the business level requirements to deliver long-term value. This balancing act has proven difficult and inefficient: systems and application management tools are resource intensive and traditional infrastructure management architectures have developed over time on a project by project basis. These traditional management systems consist of multiple tools that require administrators to waste time performing too many steps to handle routine administrative tasks. Operational efficiency and agility in your enterprise are directly linked to the capabilities provided by the management layer across the entire stack, from the application, middleware, operating system, compute, network and storage. Only when this end to end capability is provided will we experience the full benefit of a scalable, efficient, responsive and secure datacenter. Managing Exalogic is substantially less complex and error prone than managing traditional systems built from individually sourced, multi-vendor components because Exalogic is designed to be administered and maintained as a single, integrated system (Figure 1). It is at the forefront of the industry-wide shift away from costly and inferior one-off platforms toward private clouds and Engineered Systems. Read the full whitepaper "Evolution from the Traditional Data Center to Exalogic: An Operational Perspective". Full document is available for download at the Exadata Partner Community Collaborative Workspace (for community members only - if you get an error message, please register for the Community first).

    Read the article

  • When decomposing a large function, how can I avoid the complexity from the extra subfunctions?

    - by missingno
    Say I have a large function like the following: function do_lots_of_stuff(){ { //subpart 1 ... } ... { //subpart N ... } } a common pattern is to decompose it into subfunctions function do_lots_of_stuff(){ subpart_1(...) subpart_2(...) ... subpart_N(...) } I usually find that decomposition has two main advantages: The decomposed function becomes much smaller. This can help people read it without getting lost in the details. Parameters have to be explicitly passed to the underlying subfunctions, instead of being implicitly available by just being in scope. This can help readability and modularity in some situations. However, I also find that decomposition has some disadvantages: There are no guarantees that the subfunctions "belong" to do_lots_of_stuff so there is nothing stopping someone from accidentally calling them from a wrong place. A module's complexity grows quadratically with the number of functions we add to it. (There are more possible ways for things to call each other) Therefore: Are there useful convention or coding styles that help me balance the pros and cons of function decomposition or should I just use an editor with code folding and call it a day? EDIT: This problem also applies to functional code (although in a less pressing manner). For example, in a functional setting we would have the subparts be returning values that are combined in the end and the decomposition problem of having lots of subfunctions being able to use each other is still present. We can't always assume that the problem domain will be able to be modeled on just some small simple types with just a few highly orthogonal functions. There will always be complicated algorithms or long lists of business rules that we still want to correctly be able to deal with. function do_lots_of_stuff(){ p1 = subpart_1() p2 = subpart_2() pN = subpart_N() return assembleStuff(p1, p2, ..., pN) }

    Read the article

  • Massive 404 attack with non existent URLs. How to prevent this?

    - by tattvamasi
    The problem is a whole load of 404 errors, as reported by Google Webmaster Tools, with pages and queries that have never been there. One of them is viewtopic.php, and I've also noticed a scary number of attempts to check if the site is a WordPress site (wp_admin) and for the cPanel login. I block TRACE already, and the server is equipped with some defense against scanning/hacking. However, this doesn't seem to stop. The referrer is, according to Google Webmaster, totally.me. I have looked for a solution to stop this, because it isn't certainly good for the poor real actual users, let alone the SEO concerns. I am using the Perishable Press mini black list (found here), a standard referrer blocker (for porn, herbal, casino sites), and even some software to protect the site (XSS blocking, SQL injection, etc). The server is using other measures as well, so one would assume that the site is safe (hopefully), but it isn't ending. Does anybody else have the same problem, or am I the only one seeing this? Is it what I think, i.e., some sort of attack? Is there a way to fix it, or better, prevent this useless resource waste? EDIT I've never used the question to thank for the answers, and hope this can be done. Thank you all for your insightful replies, which helped me to find my way out of this. I have followed everyone's suggestions and implemented the following: a honeypot a script that listens to suspect urls in the 404 page and sends me an email with user agent/ip, while returning a standard 404 header a script that rewards legitimate users, in the same 404 custom page, in case they end up clicking on one of those urls. In less than 24 hours I have been able to isolate some suspect IPs, all listed in Spamhaus. All the IPs logged so far belong to spam VPS hosting companies. Thank you all again, I would have accepted all answers if I could.

    Read the article

  • Touchpad issues on HP Pavilion dm4 (can't right click)

    - by Habstinat
    Can anyone help me with my touchpad issues? I have a HP Pavilion dm4 and it has two areas on the bottom of the touchpad to designate right and left clicks. This mostly doesn't work on Ubuntu in the fact that it recognizes any taps on either tap zone as a left click. Instead, I have it set so if I tap anywhere on the pad it makes a left click. There should be, and there are, many ways in the mouse configuration window to simulate a right click using only a touchpad. None of these work. Changing mouse orientation doesn't do anything, "dwell click" also does nothing, and, the oddest part of this problem, whenever I try to turn "Simulated Secondary Click" off (it doesn't work anyways, but just to try to toggle it), the entire theme of my desktop changes to a gray Windows '95ey look. The only way to get rid of this is to close and reopen the mouse preferences window. My computer is fairly new and the Ubuntu installation is less than a day old. I didn't do anything that I think could cause this. The problem is that I can't right click. Help, please?s that I can't right click. Help, please? Afterword: I installed two scripts from http://sansmicrosoft.blogspot.com/2010/10/pavilion-dm4-1160-erratic-touchpad.html . They didn't do anything I couldn't already do, and they did not make it possible for me to right click. :(

    Read the article

  • When creating a GUI wizard, should all pages/tabs be of the same size? [closed]

    - by Job
    I understand that some libraries would force me to, but my question is general. If I have a set of buttons at the bottom: Back, Next, Cancel?, (other?), then should their location ever change? If the answer is no, then what do I do about pages with little content? Do I stretch things? Place them in the lone upper left corner? According to Steve Krug, it does not make sense to add anything to GUI that does not need to be there. I understand that there are different approaches to wizards - some have tabs, others do not. Some tabs are lined horizontally at the top; others - vertically on the left. Some do not show pages/tabs, and are simply sequences of dialogs. This is probably a must when the wizard is "non-linear", e.g. some earlier choices can result in branching. Either way the problem is the same - sacrifice on the consistency of the "big picture" (outline of the page/tab + location of buttons), or the consistency of details (some tabs might be somewhat packed; others having very little content). A third choice, I suppose is putting extra effort in the content in order to make sure that organizing the content such that it is more or less evenly distributed from page to page. However, this can be difficult to do (say, when the very first tab contains only a choice of three things, and then branches off from there; there are probably other examples), and hard to maintain this balance if any of the content changes later. Can you recommend a good approach? A link to a relevant good blog post or a chapter of a book is also welcome. Let me know if you have questions.

    Read the article

  • Programming During a Crisis

    - by Duracell
    Hello, I'm having difficulty turning this into a proper question, but here goes... Some of you may have heard about the flooding happening in Queensland, Australia. Well, I'm in the inner suburbs of Brisbane right now; the river has been slowly creeping toward my house since Tuesday. When I left for work this morning it was twenty meters down the road when it is normally kilometers away. Within hours of the distater striking, the government already had some pretty good web applications available for people to get information about what was happening and where the flood was predicted to rise. They also set up a database for people to search for the whereabouts of relatives or could register their location for others to see. Has anyone been involved in the development of these kinds of projects before? It's interesting that they could churn out this software in what appeared to be less than a day when the average development house could take weeks at best. In what ways did it differ from a 'normal' project? Any other thoughts?

    Read the article

  • Been doing .NET for several years and am thinking about a platform change. Where do people suggest I go?

    - by rsteckly
    Hi, I've been programming in .NET for several years now and am thinking maybe its time to do a platform switch. Any suggestions about which platform would be the best to learn? I've been thinking about going back to C++ development or just focusing on T-SQL within the Microsoft stack. I'm thinking of switching because: a) I feel that the .NET platform is increasingly becoming commodified--meaning that its more about learning a GUI and certain things to click around than really understanding programming. I'm concerned that this will lend itself to making developers on that stack increasingly paid less. b) It's very frustrating to spend your entire day essentially debugging something that should work but doesn't. Usually, Microsoft releases something that suggests anyone can just click here and there and poof there's your application. Most of the time it doesn't work and winds up sucking so much more time than it was supposed to save. c) I recently led a team in a small startup to build a WPF application. We were really hit hard with people complaining about having to download the runtime. Our code was also not portable to any other platform. Added to which, the ram usage and slowness to load of the app was remarkable for its size. I researched it and we could not find a way to optimize it. d) I'm a little concerned about being wedded to the Windows platform. What are the pros and cons of adding another platform and which platform do people suggest? Thanks!

    Read the article

  • Noise Canceling Earphones

    - by Mark Treadwell
    I travel a lot. The hours spent droning through the sky can be made more tolerable with an MP3 player and a set of noise-cancelling headphones. Reducing the sound of the airflow and engines is a great relief. For a year or two, I used a pair of folding Sony MDR-NC5 Noise Canceling Headphones, the ear foam covers self-destructed. I replaced them with old washcloth material and was happy, but the DW thought it looked bad.  I switched to a new set of Sony MDR-NC6 Noise Canceling Headphones.  These worked equally well, although they did not fold as small as the MDR-NC5 headphones. Over four years of use, the MDR-NC6 headphones started cutting out and making popping noises.  This was not surprising considering the beating they took on travel in my backpack.  It looked like I needed another new set. The older MDR-NC5 headphones were still on the shelf with the hated washcloth covers.  A quick search online showed a vibrant business in selling replacement ear foams, often at exorbitant prices.  Nowhere did I see ear foam covers made for the oblong MDR-NC5.  I then realized that foam is stretchable and that the shape should not matter.  After another search and some consideration, I purchased 2-5/16" foam pad ear covers that were able to stretch over the MDR-NC5's strange shape.  Problem solved for less than $5.

    Read the article

  • Is it okay to use a language that isn't supported by your company for some tasks?

    - by systempuntoout
    I work for a company that supports several languages: COBOL, VB6, C# and Java. I use those languages for my primary work, but I often find myself to coding some minor programs (e.g. scripts) in Python because I found it to be the best tool for that type of task. For example: An analyst gives me a complex CSV file to populate some DB tables, so I would use Python to parse it and create a DB script. What's the problem? The main problem I see is that a few parts of these quick & dirty scripts are slowly gaining importance and: My company does not support Python They're not version controlled (I back them up in another way) My coworkers do not know Python The analysts have even started referencing them in email ("launch the script that exports..."), so they are needed more often than I initially thought. I should add that these scripts are just utilities that are not part of the main project; they simply help to get trivial tasks done in less time. For my own small tasks they help a lot. In short, if I were a lottery winner to be in a accident, my coworkers would need to keep the project alive without those scripts; they would spend more time in fixing CSV errors by hand for example. Is this a common scenario? Am I doing something wrong? What should I do?

    Read the article

  • Fastest approach to 3D animation

    - by HappyFerret
    I'm currently tasked with designing a small HTML5 game. Having done everything by myself so far (3D models, codebase, game design, etc) I'm now at a point where I'm running out of time. I've less than a day to animate and bind everything together. However, that's exactly my problem. I was under the naive impression that everything would be easier if I went with pre-rendered 3D models. However, I didn't consider the most difficult part. Animation. After having spent over an hour trying to figure out messiahStudio, I figured it's time to ask for outside help. Is there any easier solution to 3D animation than 3D rigging? What I'm basically looking for is some sort of tool that allows me to simply grab and move/deform select polygons. It doesn't have to be as life-like and accurate as rigging, just efficient enough. Were the circumstances any different, I might just learn how to rig. But that's sorely out of scope right now. PS:The models were created in Sculptris but are fairly low-poly.

    Read the article

  • Motivation for a service layer (instead of just copying dlls)?

    - by BornToCode
    I'm creating an application which has 2 different UIs so I'm making it with a service layer which I understood is appropriate for such scenario. However I found myself just creating web methods for every single method I have in the BL layer, so the services basically built from methods that looks like this: return customers_bl.Get_Customer_Prices(customer_id); I understood that a main point of the service layer is to prevent duplication of code so I asked myself - why not just import the BL.DLL (and the dal.dll) to the other UI, and whenever making a change re-copy the dlls, it might not be so 'neat', but still less hassle than one more layer? {I know something is wrong in my approach, I'm probably missing the importance of service layer, I'd like to get more motivation to create another layer, especially because as it is I found that many of my BL functions ALREADY looks like: return customers_dal.Get_Customer_Prices(cust_id) which led me to ask: was it really necessary to create the BL just because on several functions I actually have LOGIC inside the BL?} so I'm looking for more motivation to creating ONE MORE layer, I'm sure it's not just to make it more convenient that I won't have to re-copy the dlls on changes? Am I grasping it wrong? Any simple guidelines on how to design service layer (corresponding to all the BL layer functions or not? any simple example?) any enlightenment on the subject?

    Read the article

  • Free Xsigo Technical Pre-sales workshop for Selected Partners !

    - by mseika
    In 2012 Oracle acquired Xsigo, a developer of network I/O virtualisation solutions. This acquisition compliments Oracle’s extensive virtualisation portfolio. With Oracle Virtual Networking products (Xsigo) you can: Virtualise connectivity from any server to any storage and any network. Reduce datacentre complexity by 70% Cut infrastructure expenses by up to 50% Benefits to Channel Partners: Offer a unique proposition that your competitors can’t match. Provide an innovative solution that delivers more performance at less cost. High margins that help sell more products and services. This course is aimed at Technical Pre-Sales Consultants equipping them to provide detailed demos, and architect RFP feedback and customer solutions. The language of this event is French. WHEN24th September 2013 WHEREOracle France 15, boulevard Charles De Gaulle92715 COLOMBES FEESFree of charge 09.00: Welcome, Coffee & Introduction 09.30: Value Propositions, Architecture & Use Cases 11.30: Build a OVN Web Quote & TCO 12.30: Lunch 13.30: Competitive Summary 14.00: Design Scenario Workshop 15.45: Questions/Opportunities  REGISTRATION: Register via this link as soon as possible, 14th june, latest. Note that we have only 20 seats in total for this event. Note that after 14th june we will release free seats for other organizations to register. We look forward to your participation! What we expect from you: You will bring your own laptop. Recommended browser is Firefox 10 ESR. You have checked the material and conducted the assessments. You will be flexible in terms of Agenda and Progress as we intend this to be more of a Workshop having Dialogue rather than sticking tightly into the tentative timeline. What this is not: This PartnerLab does not replace Oracle University Trainings. This PartnerLab does not lead to a Certification as such. This PartnerLab does not enable Partners to full and complete implementation skills.

    Read the article

< Previous Page | 150 151 152 153 154 155 156 157 158 159 160 161  | Next Page >