Not as far as I know, but like him my family immigrated from the Netherlands. In my case, though, my grandparents immigrated in the 1880s. He came on his own as an adult after he was already well-known in computer science.

So far, I appear to be the eldest, with a birth year of 1944. I'd be happy to teach any of you young'uns what I remember offhand about the early days of FORTRAN and COBOL! Of course, eventually I progressed from there to Basic, then C, C++, and finally C# with a bit of JavaScript and a few others thrown in for good measure. Please wake me in time for my nap! :zzz:

But if you're going to use the MFC Feature Pack you have to be cautious about the fact that Microsoft simply bought it from an outside vendor and added it to the feature set with most of its rough edges still intact. See for example, this[^] exploration of the CMFCLinkCtrl class intended for use with hyperlinks. Among other things it included hard-coded fonts.

I suggest you investigate DiscountASP.Net (http://www.discountasp.net[^]). I've used their service for years and have always been extremely satisfied. They're located in Pasadena, California. They know ASP.Net backwards and forwards, and the prices are excellent, with many options from which you can choose.

The easiest way to find your articles is to go to your Code Project profile. You'll find a button for your articles and that will list all of them. You can then use the links to get to where ever they've been moved.

I'm curious to know how you managed to get it to work. I've been trying to activate Hyper-V on my Windows 8 Pro laptop for several months without success. I've followed all of the suggestions I've found on the Internet, to no avail. I've done all the testing on my laptop's setup, including the BIOS, and everything that's required seems to be in place. But when I try to activate Hyper-V, after the reboot I always get an unexplained error followed by "Reverting ...". I'd be grateful for some information about the problems you encountered and how you overcame them. As a followup, I finally gave up and installed VirtualBox. It works great, so I can now run a program I use occasionally that will not work on anything later than Windows XP. It's an old MS-DOS program that takes over the entire screen, so Vista and later prevent it from running. VirtualBox was easy to install under Win 8 Pro and seems to work fine.

I have some additional information about this situation. As I thought, IE was timing out before the credit-card issuer was able to respond with a confirmation. The same thing has happened several times since then, including once as I was attempting to purchase tickets for a flight to South America. I prefer to use IE for such things because Adobe Acrobat has added a "Convert" toolbar button to IE9 so that I can get a PDF file as proof of payment (with Vista or Win7 you can no longer use the "print to PDF" option due to security issues). As far as I can determine, the convert button isn't offered with any other browser. Eventually I found a setting in the registry that affects the amount of time IE9 allows for responses. Here is the registry key:

HKEY_CurrentUser\Software\Microsoft\Windows\CurrentVersion\Internet Settings

The value that controls the setting of interest is a REG_DWORD named "ReceiveTimeout". The default setting is 10000 ms, or 10 seconds. That's too short for most credit-card processing. By changing it to 120000 (equivalent to two minutes) I was able to overcome the problem. It's odd that Microsoft set this value so low, and even odder that you can only change it through the registry. I suspect that it will cause a lot of people to stop trying to use IE9 for online purchases.

No, you have to go a bit further along the process than that, and you wouldn't have a million dollars in your bank account but rather a million, one-dollar charges against the credit card (assuming your credit-card issuer let it go that far). But you don't get there just from clicking the renew button; you would have to proceed to the next page and then click (or have your AutoIT script simulate clicking) the "Continue" button after filling out the form with your name and credit-card information. Then, if the process times out, the one-dollar approval charge won't be cancelled and replaced by the actual renewal charge, but instead will remain as an apparent charge against the card, to Microsoft's benefit. The system allows Microsoft to confirm that the card is legitimate before moving ahead with the actual charge. You don't get anything out of it in any case; Microsoft does. But even if you do this a million times without ever getting to the final renewal page, the one-dollar charges appear to the credit-card issuer as authorization charges and they wouldn't pay those amounts to Microsoft.

I had an odd experience yesterday. It was time to renew my annual Visual Studio/MSDN subscription so I signed onto MSDN with my main browser, IE9. I clicked the "Renew" button, then after a couple of timeouts that I attributed to a lot of traffic on the site, eventually got to a page where my credit-card information was requested. I entered it, then hit the "Continue" button and after a timeout got IE's standard "Unable to connect" message. Hitting reload put me back at the credit-card page again. I tried this several more times, hoping that I'd get lucky and hit "Continue" just as there was a lull in traffic. It never happened. Finally I gave up and logged off without completing the renewal. I tried again several hours later, thinking that because it had then gotten to be evening on the West Coast the traffic might have abated somewhat. Still no luck, just several more tries with the same lack of success. Then I launched Firefox 5, logged onto MSDN again, and hit the renewal button. This was only a minute our so after my previous attempt with IE9. With Firefox, everything went through without pause. No timeouts, no lags, just steady progress through the series of web pages and the final result that I was looking for. This morning I got a call from the credit-card issuer wanting to know why I had seven $1 charges to Microsoft plus one for the price of the VS renewal. I explained what I had done and he told me it was because Microsoft sends a $1 charge to the account to make sure it's legitimate, then later backs out that charge and revises it with the correct amount. When the request times out before completion the correction is never made. He said they would remove all of the $1 charges and approve the renewal charge. I'm not sure whether this is an IE9 issue or a website issue but it seems to me most likely that it results from some problem within IE9 since the website performed correctly with Firefox.

I've just tried it with an MFC dialog application. It's not perfect because the color change doesn't "stick" when the edit control has the focus but it comes back when you tab away from the combo box. You'd need to do some more work to make it work the way you want when the edit control has the focus, but this should give you a starting point. The class that I derived from CComboBox is named CColorEditCombo. In the Visual Studio designer I added a CComboBox control to the dialog window. With the ClassWizard I added a CComboBox control variable named m_ctlComboColorEdit. Then in the header file for the dialog class I changed the declaration for this variable to the following:

CColorEditCombo m_ctlComboColorEdit;

This associates that control with the new CColorEditCombo class rather than the parent CComboBox class. In the class file for CColorEditCombo I added a handler for the WM_CTLCOLOR message as follows:

HBRUSH CColorEditCombo::OnCtlColor(CDC* pDC, CWnd* pWnd, UINT nCtlColor)
{
HBRUSH hbr = CComboBox::OnCtlColor(pDC, pWnd, nCtlColor);

switch (nCtlColor)
{
case CTLCOLOR\_EDIT:
case CTLCOLOR\_MSGBOX:
    // Subclass the edit control.
    if (m\_edit.GetSafeHwnd() == NULL)
        {
        m\_edit.SubclassWindow(pWnd->GetSafeHwnd());
        }
    // Set the background color for the edit control.
    pDC->SetBkColor(RGB(255, 255, 0));
    // Recolor the edit control.
    m\_brush.DeleteObject();
    m\_brush.CreateSolidBrush(RGB(255, 255, 0));
    hbr = (HBRUSH) m\_brush.GetSafeHandle();
    return hbr;

case CTLCOLOR\_LISTBOX:
    // Subclass the listbox control.
    if (m\_listBox.GetSafeHwnd() == NULL)
        {
        m\_listBox.SubclassWindow(pWnd->GetSafeHwnd());
        }
    // Add recoloring actions for the listbox here if desired.
 }

return hbr;
}

Note the member variable named m_brush. This has to be declared in the header file for the CColorEditCombo class as a CBrush object. It should be deleted in the class destructor as well as just before the CreateSolidBrush statement shown above. If you'd like, I can send you the solution for the very simple MFC dialog application I put together as a test to make sure this procedure works. You'll have to give me your email address and

You have to subclass the edit control of the combo box, since a combo box is a combined control. Here is an article[^] from Microsoft support describing how to do so.

Fair enough. But why not at least fix the version in the archive? It doesn't work either.

I'm not able to access any of the May 10 newsletter articles. When I click a link I go to this CodeProject error page[^]. Here's an example of one of the links from the newsletter: http://www.codeproject.com/Article.aspx?tag=1928097915813829[^] The same problem occurs when I access the newsletter from the newsletter archive at http://www.codeproject.com/script/Mailouts/View.aspx?mlid=7882[^]. Clicking any article link within that newsletter gets me the error page. Any suggestions on how to overcome this? Thus far, I'm puzzled. :confused: I've tried this on three difference computers using IE6 (yes I know but it's still mandatory at the agency where I work), IE8, and the latest version of FireFox. Same result on all of them.

According to Wikipedia[^], "a web application is an application that is accessed via a web browser over a network such as the Internet or an intranet. The term may also mean a computer software application that is hosted in a browser-controlled environment (e.g. a Java applet) or coded in a browser-supported language (such as JavaScript, possibly combined with a browser-rendered markup language like HTML) and reliant on a common web browser to render the application executable." By comparison, a browser session starts when an individual user accesses or logs onto a particular website and concludes when the user logs off, closes the browser, or clicks away from that website. The session may or may not involve authentication.

Thank you for this reply. However, it doesn't really help because as noted in the original question I have to do the detection without javascript. AJAX won't be much help without javascript. But thanks anyway for the thought. :)

I'm hoping to cash in on the well-known ASP.Net expertise of CodeProject members. Currently I'm working on an ASP.Net website for a nonprofit organization with users in many developing countries, including locations where only dialup is available. The organization has determined that many of its users have disabled javascript for security reasons. The organization wants the new website to be graphically appealing, but also wants to be able to turn off most of the graphics for dialup users. This means we have to have some way of determining whether a request comes from a user on dialup or on broadband. The organization wants its home page to be filled with graphics for broadband users, which means there has to be a way to detect the connection speed in the background before sending a visible response to the browser. At first I thought this wouldn't be possible from the server side. Although the advice I've found by searching many forums (including this one) suggested largely that this is the case, a few people thought it should be possible. Most of the suggestions involved timing downloads through WebClient or WebRequest/WebResponse. By running a number of tests on both dialup and broadband connections I've determined that the download method doesn't work because what's being timed is the length of time needed to send the download; without javascript there's no way to tell when it has actually been received by the browser. A different approach was suggested via a PHP script posted by Emanuele Feronato on her blog[^]. This approach involves sending some text to the browser, one KB at a time, and flushing the response after each KB. She suggested sending 512 KB to get a good estimate of the connection speed, but that's far too much for dialup users to endure. I tested her approach with 16 KB and it seemed to give reasonable results. Since PHP is a server-side technology, I thought this might work in ASP.Net as well. My C# code for the approach is below. The text I'm sending is just a collection of dots (ASCII 0x2e). To hide the text being pushed to the browser I've used a <span style="display:none;"></span> around it. Emmanuele used comment tags around the text to accomplish the same thing. Here are a few things I've noticed: 1. The method doesn't work

You can find a step-by-step example at http://msdn2.microsoft.com/en-us/library/aa454896.aspx (see Exercise 2 in the example). The example uses CF1.0 but it should work under 2.0.

:confused:I'm using C# to develop an application for an ARM handheld that uses PocketPC 2003. Part of the time I work on an XP machine at home, and part of the time on a Win2K SR4 machine at work. The application runs under the .NET Compact Framework 2.0. It includes a set of Windows forms for data entry. The application looks fine in the VS design view on the XP machine and deploys properly to the handheld. However, when I tried to open the project's design view under Visual Studio 2005 Professional on the W2K machine at work, I received an error message stating that an input string was improperly formatted and the form did not load. I found a reference to what appeared to be an identical error in Microsoft's Knowledge Base[^]. However, the error persisted even after I downloaded and installed the hotfix developed to fix the error. I'm wondering whether anyone else has encountered this problem, and if so, whether they might have found a workaround. Incidentally, I can build and deploy the solution from my Win2K machine and it runs properly, either under the PocketPC 2003 emulator or on the handheld itself. I just can't make any changes in the design editor on the Win2K machine because the design editor won't load. This problem occurs even if I create a new smart device project in either C# or VB, so it's not just this specific project. It doesn't occur if I create a smart device project in VC++, however. I've repaired VS 2005 through the Add/Remove Programs utility, to no avail. I'm afraid the answer might be a complete re-install of VS 2005 but I dread doing that. I'd be grateful for any ideas or suggestions.