Saturday, February 05, 2011
Adding concurrency to an existing application can be remarkably simple. I have a demo where I change a for loop to a parallel for and -poof!- the app gets faster. The hard part of that demo is knowing that a given loop is ok to parallelize, that it doesn't matter what order things are done in as long as they all get done once. And, of course, knowing that the loop in question is responsible for your execution time. Making the right decision is the hard part - typing the code is easy.
With that in mind, you need advice about choosing your structures and algorithms. Here are some handy things that may help you. On the native parallel programming blog they will help you choose your parallel sort
. And on the VC++ blog, it's all about the parallel containers
. If you're going to write parallel code (and let's face it, eventually you will) you need to read this sort of thing.
Thursday, February 03, 2011
Here's a useful blog entry from Nish
that shows how simple it can be to handle what at first seems an intractable problem - how to get an unmanaged array into a managed array. The marshaller doesn't know the size of the array, but as long as you pass the length as a parameter, you can create the array yourself and copy the contents over. Not hard at all, once you know.
Tuesday, February 01, 2011
Like a lot of folks, I have a Windows Phone and I enjoy it a lot. I've filled it up with (free) games and they pass the time quickly when I have to wait in line or otherwise have an empty minute or two. C++ MVP Marius Bancila had the same experience, but unlike me he responded by writing a version of the game in MFC so he could play it on his own PC as well as on a phone.
You can get the source from his blog post, and another post and its comments (mostly in Romanian - Bing Translate is your friend there) have a nice discussion of performance issues. After all, whenever two or more C++ programmers gather together, someone is going to have an opinion about performance. Take a look for yourself!
Sunday, January 30, 2011
There are some really amazing WPF applications out there, and many of them are extraordinarily beautiful. I am not very good at making beautiful applications, though the WPF themes
help avoid the "grey boxes on a grey background" look I usually achieve. Now if you want to develop an application that presents a whole new paradigm, there should be someone on your team who really "gets" user experience and how colours work together and so on. But what if you just want to make the same sort of apps you've always made, using the latest technology instead of something a little older? You want to think the way you thought when you were doing Windows Forms, for example, and use the same tools, but end up with a WPF application or a Silverlight application.
Well, Pete Brown has an interesting blog post
speaking directly to this kind of developer. He makes the case that for these kinds of applications, you don't need to add a designer to the team. Plenty of screenshots to support his theory - see what you think. I think I agree, myself. Because the post was last year, there's lots of interesting conversation happening in the comments, too, so be sure to read those while you're there.
Friday, January 28, 2011
Despite this rather famous xkcd comic, most developers would rather spend as little time compiling as possible:
But most of us how no clue why a build takes as long as it does, or how to make a project build faster. I mean sure a Clean and then a Rebuild will take longer than an incremental build. But beyond that, which is quicker - a debug build or a release build? Debug puts more information in, but it doesn't optimize - which wins? Does it make a difference if you keep Visual Studio up all day or if you close and re-open it from time to time? Are there options you can set that make your builds faster? All this and more is covered in remarkable detail in a new blog post from the Visual C++ team. You need to read it.
Wednesday, January 26, 2011
Not everyone knows that the Windows SDK includes a C++ compiler and linker that you can use if you don't have Visual Studio installed. Of course, Visual Studio Express is free, and you could use that to compile and link C++ apps. The advantage of using the tools that come with the SDK is that you don't have to install and configure a lot of stuff to get started. On a new machine that may appeal to you.
Gus Class has posted a super-quick 3 minute video
that shows how to compile one of the SDK samples using only the tools that come with it. I just wanted to elaborate on one thing. The SDK installs, by default, under Program Files. As you probably know UAC prevents applications from writing to any folder under Program Files unless the application is elevated. That's why Gus launches an elevated command prompt in this video. If you don't want to elevate your command prompt, you'll need to move the sample folders and adjust your path accordingly.
Three minutes. Worth a watch. It's always good to have some tricks up your sleeve on a new machine.
Monday, January 24, 2011
Recently the East of Toronto .NET User Group had a Windows Phone 7 presentation (by Atley Hunter) that was very well done, and well attended. (He's blogged about it and included some helpful links, so you should be sure to read his post too.) You would never have known it was his first presentation - he was calm and confident and knew what he was talking about.
At the meeting I met Ashish Kaila, who is working on a toolkit for WPF developers and a Windows Phone 7 library. He showed me the docking panes and some other fun functionality.
To me this is a huge benefit of user group meetings: in addition to the presentations, usually as good as any you would see at a paying conference, you also can meet fellow attendees. I hadn't heard of these tools before, and they could save you a lot of time and effort. The next best thing to coming to meetings, I suppose, is reading blog posts about meetings. Here's where you can learn more about Ashish and his products:
If you're not going to your local user group meetings, why on earth not? We don't bite, honestly, and you have so much to learn and gain from being there. See you next time!
Saturday, January 22, 2011
The revisions to the C++ standard, headed to producing a new C++ standard which is still nicknamed C++0x, are proceeding at their stately pace, and will for months yet. But since a number of compilers are supporting
more and more of the functionality and syntax, which are pretty much locked down now, we can adopt much of it today. (I don't know why some people are running around claiming that because the standard isn't released yet, no-one should be using any of it, but when you bump into such people, you can send them to me.)
I came across an interview
with Bjarne Stroustrup, the father of C++, at CodeGuru. If you're wondering why the standard is so far behind schedule, what the very first C++ developer thinks about the very newest C++ developers (undergrad students), and how risk feels to a standards committee, then you will enjoy reading this, as I did. The lambda syntax on the first page isn't quite right, but you can fix that up if you need to. There are plenty of lambda examples here in my blog.
Give it a read!
© Copyright 2022 Kate Gregory
Theme design by Bryan Bell
newtelligence dasBlog 2.3.9074.18820
| Page rendered at Monday, May 23, 2022 7:58:50 PM (Eastern Daylight Time, UTC-04:00)
On this page....
Pluralsight Free Trial