ATL GUI programming (preface)

Objective

Perhaps, you are a stubborn SDK surrounded by;

Perhaps, you have an extremely repulsive attitude to MFC, even like me almost ignorant of it;

Perhaps, you meet the top two, and are looking for a way out;

Maybe you found a way out--wtl, but still worried about its technical support, its upper and lower compatibility;

Perhaps, you still have more perhaps;

Perhaps, this time you see Li Ma this article.

The world's stories, often by a variety of "perhaps" interwoven.

My transformation.

History tells me that when you recommend something to someone, you should first tell others how much you have benefited from it, and if you can make it all more convincing. For example, I tell you, since I used ATL, my waist is not sour, my legs don't ache, and I walk stronger, then you may have a little affection for ATL. Yes, in this section of the preface, I will spare no effort to recommend atl--to you with my own Win32 GUI programming experience because my series is themed on it.

In 2000, I started using Visual Basic to write Win32 (strictly speaking, then Windows 95 was not a pure Win32 system, it was just a 16/32-bit Windows hybrid) program. By the year 2001, I began to tire of its runtime and operational efficiency, so I turned to Delphi. Perhaps is Borland's fan is not good, perhaps I am tired of Pascal's rigorous, perhaps is that kind of "the real programmer uses the VC" the childish thought in the mischief--in short, I also invested in the VC.

On the one hand, I am stingy son, not willing to spend money to buy books, on the other hand, China's MFC books a variety of mixed and so I can not mention the interest of learning-anyway, I do not know when to start, I began to strongly conflict with MFC this no fault of the framework, With just a few SDK articles in magazine newspapers and MSDN 6.0 in English, I started my hard long march. Fortunately, although this road stumbling, but eventually I reached my own northern Shaanxi. If you have seen my Win32 technical articles on the Internet or in other media during this period, it should be found that I rarely dabble in any content of MFC-I do not deny that I have a narrow bias against it, although the graduation project I completed in 2005 was written in MFC.

The SDK is not a comfortable destination after all, I think the SDK fans should have a deep understanding. Even if your code goes up to 1000 lines, you'll find it very difficult to maintain these things-bloated, stacked, global, coupling, poor reusability-and there are so many ugly words that can be used on this. Yes, I did say that I do not like MFC, but I also can not deny the use of MFC written program structure clear, streaming clear-of course, if you are not in-depth to the MFC inside to see. For example, the 2.11 version of the process viewer July I wrote was written in the SDK (this version is open source), just 1000 lines of code have made it difficult for me to control them, so I have more than one time to rewrite it-I want to use MFC, even the VCL of Delphi. However, the limitations of the runtime and the size of the exe still make me abandon these ideas.

If MFC and VCL are grown in the deep palace of the noble family, then WTL can be regarded as a wandering the stunning singer. When I was in 2004, I had a casual encounter with it. Excellent design, lightweight exe, unmatched efficiency-if not its compatibility problems (if not modified, WTL 7.1 even its own incidental to some of the sample can not compile), I almost bowed to its pomegranate skirt.

So I continue to be confused, this kind of confusion continues to 2005 after I graduate. In August 2005, I contacted the originator of WTL-an ATL designed for the development of COM components-in order to study the technology that calls ActiveX controls without using MFC. Today I seem to have been unable to recall the situation, only remember one months later, with ATL rewrite July v2.20 was born.

Compared to the ladylike and stunning singer, ATL is more like a gentle and virtuous woman.