Did you note that the "picture" is actually 9 pictures and that the airplane and its shadow are on different segments -- the center segment (airplane in air) is probably a double exposure.

Bite me -- the problem is NOT the public education system but the national administration, ie Bush, who is driving all to mediocrity with such ill-conceived programs as "no child left behind" and "mainstreaming" -- we drag the more average to guifted down to the lowest common denominator, the also-ran, thank you for playing. You want to know what's wrong with the present public education system, contrast K-12 with colleges, trade schools, and other specialized institutions, and, if you dare, contrast public education in the US with most other countries who do not relish mediocrity and actually reward superior achievement. You don't achieve, you don't go on to higher education. Why do we teach Java instead of C++ -- why not ask why do most colleges not have programs in assembly language and programming process controllers? It is EASIER to teach to the "mainstream", it is EASIER to recruit mediocre "instructors" rather than professors, and it is EASIER for the average student to comprehend. Who gives a damn how the machine does it or how to optimize it -- just code i++ or ++i and the "system" takes care of whatever variable size you are using, including range checking? Who gives a damn about raw IO buffer sizes compared to actual disk sectors? The "system" will save your butt. Memory is cheap so why bother with memory utilization -- let the "system" manage a larger swap file, let the "system" page for you, let the "system" provide all means of native support -- all you have to do is code simplicity and let the program take all the memory and overhead it wants. Its so much easier to "box" things and cram them on the stack/heap rather than create optimized data structures. By the way, how many programmers know what variables go on the stack versus the heap? How many know how garbage collection is managed and the overhead added to your program by garbage collection or range checking? Why worry about managing the life cycle of your pointers, just use "smart pointers" and the "system" will save your butt. When our national leaders finally realize, if they ever do, that we should not embrace mediocrity but achievement, then we will go back to teaching the core basics and REQUIRING competency to progress, not pass merely because the mediocre also-ran. As a parting thought, have you ever checked what we pay public school teachers compared to the actual amount of work they perform and the goals and standards we set for them in terms of the mediocre student advancement? It takes a pretty damned dedicated t

DirectShow "useless" -- totally agree. Want to see a 3G P4 with 2G memory choke? Try using live & encoded video, sequentially accessed, transitions between videos, multiple VMR's (one for each of 3 screens) each with different mixed content (text and images), on live video, insert second video for hearing-impaired "signer", allow input cropping rectangle and output size and position adjustments, and for one customer, split a video stream between two display devices, oh yes, and cut the video latency for live video to minimum -- gives new meaning to "100% CPU utilization" and "frame drop". DirectShow support in C# -- only if you want to Marshall the hell out of the interfaces, again without support of any kind from MS. Use DirectShow in C# through COM -- Hi ho, hi ho, its Marshalling we go ..... While this is a little off-base from my original post, it has been quite lively. My original objective had something to do with draining the swamp ..... I just wanted a quicker way to scan articles on CodeProject and ignore VS.NET 2005 articles, most of which do not indicate the target in the article name, only in the text, or, in some cases, only when you try and run the sample code in VS.NET 2003 (no, I don't mind the converstion from 2002, etc, at least most of the time).

These have nothing to do with DirectShow -- they are used for access to WMP10 ProcAmp functions by interfacing through a custom "skin" -- however, MS declines to provide any sample. If you check http://msdn.microsoft.com/library/en-us/wmplay10/mmp\_sdk/usingskinswiththewindowsmediaplayercontrol.asp, you will find that a partial reference is made to this access but no sample, either here or in the SDK's. Because of the lack of DirectShow support in C#, I write all my DirectShow FilterGraphs in C++ (ugh -- wrapper) and the remainder of the programming in C#. While DirectShow does provide access to the ProcAmp, WMP10 access is not available.

Fair comment, "so long as the framework behaves as documented" -- perhaps you could answer a question that MS cannot -- where is the documentation for IWMPVideoSettingsCtrl/WMPVideoSettingsCtrl and IWMPVideoSettingsCtrlClass/WMPVideoSettingsCtrlClass, why are these elements of WMPLib shown as having different attributes between C++, C#, J# and VB, and why is ProcAmp access to WMP10 in NET, although advertised as being available by MS, not available (except as shown in the object browser) except for "internal use by MS"?

You have much more confidence in MS and their new releases than I do, particularly where much of the inner workings either is not documented, incorrectly documented or not available except for internal use by MS.

How about separating VS.NET 2005 projects from all other versions of VS.NET, particularly where 2005 is "beta" code, exclusive of production versions, and not particularly well suited for release code (forget the "go live" krud)?