I am using Embedded VC++ 4.0 with MFC. I build your basic app-wizard generated SDI application with all the default settings -- just run appwizard and compile. It works great in the simulator. Then I go to test on an Intel based processor (xscale I think). I build for the ARM processor. The application runs, but the menu does not display. For example, on the PDA, I open Explorer, browse to the application I just made, double tap on it to run it. My app runs and the screen client area show my app, but the menu at the bottom of the screen is still the Explorer menu! My apps menu is never seen or available. How can I fix this? Do I need to buy the $1500 development package from Intel? How is it that an straight-from-appwizard program won't run correctly? By the way, I tested this on two PDS of different manufacturer but both with Intel processors and the result is the same.

I am writing software that parses out an RTF file. One of the things I want to do is extract pictures and save them to a file. The following is taken directly from Microsoft's RTF FAQ (version 1.3): ---- From RTF FAQ --- The picture in hexadecimal or binary format follows the picture-destination control words. The following example illustrates the destination format: {\pict\wbitmap0\picw170\pich77\wbmbitspixel1\wbmplanes1\wbmwidthbytes22 \picwgoal505 \pichgoal221 \picscalex172 \picscaley172 49f2000000000273023d1101a030 3901000a000000000273023d98 0048000200000275 02040000200010275023e000000000 273023d000002b90002b90002 b90002b90002b9 0002b90002b90002b90002b90002b90002 b92222b90002b90002b90 002b90002b9 0002b90002b90002b90002b9000 ---- End of Stuff from RTF FAQ --- I am able to parse out all of the bytes (in this example, beginning 49f2...) and I convert them from their text representation to real bytes (i.e., the character sequence "49" is converted to 0x49), but what I get is not a readable picture. Here is more detail: Suppose I have a bitmap file. I take the same picture and embed it in an RTF document. I can open up the document in Notepad and view the hexidecimal represeantion of the image. I have also written a program that can read a binary file such as a bitmap and convert it to the text equivalent of a hexidecimal as it would appear embedded in the RTF document. I then visually compare the hex I generated directly from the hex embeded in the RTF file. The hex from the RTF file is the same as the hex I generate directly from the disk file... EXCEPT for the first several hundred bytes! Each source starts out different, but after a hundred byts or so, each is then the same, so I know my encode/decode should be working. But why is the hexidecimal of the embedded picture not exactly the same as the hexidecmal of the disk file? According to the RTF FAQ, they should be. What can I do to extract the picture? Did Microsoft do something to encode the first series of bytes for a picture so that no one else can extract it? I have found this problem with both bitmaps and GIF files. I am using MS Word 95 to generate the RTF files with pictures embedded.

This should give you all something to mull over. http://www.washtimes.com/commentary/20040810-100240-4474r.htm[^]

Tomaž Štih wrote: I thought you still have these in some parts of the United States? It's a big country -- there is a lot of variation. I grew up in a small town. For some strange reason, the small locally owned stores in that town were all closed on Wednesday as well as Sunday. Within the last ten years, the town got a Wal-Mart. I don't what happened to the Wednesday closing thing. Wal-Mart must have been a big influence in changing these closing time laws. Wal-Mart got started in small towns. All of a sudden, here was Wal-Mart selling everything cheap, and in many cases open 24-hours if not until Midnight. I hate shopping in Wal-Mart but I have no problem with their business practices.

You may know that here in the U.S., we have stores such as Wal-Mart that are open 24 hours a day. Is it true that there are laws in Germany that prohibit a store, such as a clothing store, from being open past a certain time? I heard that there are regulations the restrict the hours of German stores because of labor policy -- that is to keep workers from having to work certain hours. Is this true? What are the details? What other countries in Europe have this policy? I am just curious.

Ted Ferenc wrote: It's called marketing/sales, it has nothing to do with the quality or otherwise of the product. It is where is software guys fail! If you think that is true, go and use WordPerfect or Lotus 123 for awhile. Oh, and use Paradox for your database. Delphi is a good product, you can write code in that. Oh, and in case you are thinking that MS rolled over OS/2 with marketing hype, then think about this... OS/2 had a bug that made it crash requiring a hard disk formatting and reinstallation (according to John Dvorak). How's that for a bug? You can do everything a normal computer user does without using any MS products (word processing, spreadsheet, email, contact management, database). Have fun!

There are three kinds of lies: lies, damned lies and statistics. Benjamin Disraeli (1804-81), English statesman, author. Quoted in: Mark Twain, Autobiography, ch. 29 (ed. by Charles Neider, 1959). You can prove anything with with statistics nowadays. 38% of all people know that. Homer Simpson (Just thought I would help edumacate you'alls)

OK, this is kind of a rant... I am currently writing a web app using MSDev 6 technology. That includes VBScript for the ASP pages, Java Script for client-side scripting, and C++ for components. Obviously, there is an inefficiency in using multiple languages for development. For example, my VBscript never compiles because I keep putting ; at the end of each line. Just out of curiousity, is there any web environment, such as .NET, where you can use a single language for all development? As I understand it, client-side scripting can only be done in Java Script, (or maybe if Netscape support is required) which would mean that Java or Java Script would have to be used on the server side. That is the only single language environment I can think of.

So how do I get the IID?

So how do I do that?

Here is the code:

// this line gets a valid ADO database connection object
_ConnectionPtr pConnection = m_AdoDatabase.GetActiveConnection();

// Now I will get a pUknown pointing to the database connection
IUnknown* pUnk;

// this is successful
HRESULT hr = pConnection->QueryInterface(IID_IDispatch, (void**) &pUnk);

// Now I will attempt to get an ADO connection object from pUnknown
_ConnectionPtr pConnection2;

// this line fails
hr = pConnection->QueryInterface(__uuidof(Connection), (void**) &pConnection2);

if (FAILED(hr))
{
// this is where the failure is!
// hr is returned with a value E_NOINTERFACE
TRACE("Failed to get pointer the second time.\n");
TRACE("Error: %s\n\n", GetComError(hr));
}

Shouldn't I be able to go from an ADO Connection object, to IUnknown, and then back again? How do I get back my ADO Connection from IUnkown?

I am writing an ATL component to do some data access using ATO. I want to pass a _RecordsetPtr pointer or a _ConnectionPtr to a ATL function. I use code like this: STDMETHODIMP CMyAtlObject::ReadData(LPDISPATCH pAdoRecordset, BOOL *pbSuccess) { _RecordsetPtr pRecordsetPtr; if (FAILED(pAdoRecordset->QueryInterface(IID_IDispatch, (void**) &pRecordsetPtr))) { return S_FALSE; } // do somethign here with the recordset return S_OK; } But this doesn't work. The recordset pointer gets closed during the function call. If I pass an _ConnectionPtr as LPDISPATCH, the pointer becomes invalid. So how can I pass valid ADO _RecordsetPtr's and _ConnectionPtr's to an ATL Object?

I am writing an ATL component to do some data access using ATO. I want to pass a _RecordsetPtr pointer or a _ConnectionPtr to a ATL function. I use code like this: STDMETHODIMP CMyAtlObject::ReadData(LPDISPATCH pAdoRecordset, BOOL *pbSuccess) { _RecordsetPtr pRecordsetPtr; if (FAILED(pAdoRecordset->QueryInterface(IID_IDispatch, (void**) &pRecordsetPtr))) { return S_FALSE; } // do somethign here with the recordset return S_OK; } But this doesn't work. The recordset pointer gets closed during the function call. If I pass an _ConnectionPtr as LPDISPATCH, the pointer becomes invalid. So how can I pass valid ADO _RecordsetPtr's and _ConnectionPtr's to an ATL Object?

Mark Nischalke wrote: I'm curious as to the requirement for COM and no .NET? We have existing C++ classes that are required for some of the data processing, and these do not use .NET (and we use VC++ 6.0 only, to keep cost down and productivity up). So to leverage the existing code, they will be wrapped in COM objects. Mark Nischalke wrote: If this is a web app what does it matter what you use on the backend? Mainly what I stated above, but it is also because the web app may actually be sold to customers on installed on their servers, so it affects the customers system requirements. I am curious your perspective on those answers.

I need to write a pro-quality database web application, using all MS technologies. I can't emphasize enough that The app must have COM controls that do not use .NET! Well, I am thinking about ASP.NET. How difficult is that to use with COM objects in a web application? What is the development environment for ASP.NET? I assume VS 7.0? (I am a VC 6.0 guy). Can Visual Studio 7.0 be used to develop for ASP and COM without using .NET?

I am writing an application that uses the CrytpoAPI. In the CryptAcquireContext function call, I request the PROV_RSA_FULL provider. In my Windows 2000 development machine, it works, but on many other systems, I get the error for invalid keyset, meaning that a Cryptographic provider supporting RSA_FULL does not exist. Windows 98 does not work Windows XP clean install does not work. Windows XP with Office XP DOES work Windows 2000 clean does not work Windows 2000 SP 3 does not work Windows 2000 128-bit encryption does not work Windows 2000 with Office 97 does not work IE 6 SP 1 does not work All of the above should have the MS Base cryptographic provider How the fudge do I get RSA_FULL CryptoAPI support installed on a system?