BenoitRen

I would wait for Firefox 2.0.0.20. They forgot a last-minute fix. AGAIN. Fortunately it was a Firefox 2.0-only issue, so SeaMonkey doesn't need an additional update. Phishing doesn't require JavaScript, though it helps. NoScript is so overhyped.

Oh, so the author has been updating it again? It was stuck at one version for a long time, and that version worked on Windows 98.

Isn't CD-Burner XP such a program?

As I said above, Kompozer = nVU + a bunch of patches that fix bugs, so I don't see why not.

Yes, it's ANSI. At least until I can figure out how to properly support Unicode. That's what I tried to explain. When I use RichEdit 1.0, it reaches it and replaces the string. With RichEdit 2.0, it doesn't. Yet the CHARRANGE members are the same!

I use the free version of RealVNC. Works great.

Not correctly installed? I installed TCP/IP the standard way, through the Network panel. Installed Winsock 2 afterwards using the downloaded executable. If that's not standard, I don't know what is. There's no ipconfig.exe file on my hard drive.

Now I'm very confused. Text replacing doesn't work anymore. It does work with RichEdit version 1, but the CHARRANGE members have the same values. This is my replace function: int replaceText(char *findWhat, char *replaceWith, BOOL bWholeWord, BOOL bMatchCase, BOOL bSearchDown) { CHARRANGE cr; int foundText = 0; SendMessage(g_hRichEdit, EM_EXGETSEL, 0, (LPARAM) &cr); if (cr.cpMin == cr.cpMax) foundText = searchText(findWhat, bWholeWord, bMatchCase, bSearchDown); else { CHARRANGE crText; FINDTEXTEX ftex; ftex.chrg = cr; ftex.chrgText = crText; ftex.lpstrText = findWhat; SendMessage(g_hRichEdit, EM_FINDTEXTEX, 0, (LPARAM) &ftex); if (ftex.chrgText.cpMin == cr.cpMin && ftex.chrgText.cpMax == cr.cpMax) { MessageBox(NULL, "About to replace!", "Kladblok", MB_OK); SendMessage(g_hRichEdit, EM_REPLACESEL, TRUE, (LPARAM) replaceWith); } foundText = searchText(findWhat, bWholeWord, bMatchCase, bSearchDown); } return foundText; } The idea is that if nothing is selected, a search is made for the text so that the next time the function is called it will be selected. To be on the safe side, I do a search for the text within the selection to verify that it is indeed the string I want to replace. If that's the case, the CHARRANGE of the current selection and the one that has the search results should be the same. However, in practice, in both versions of the RichEdit control, cpMin of the selection is always lower by 2 (in my testcase of trying to replace 2 characters) than cpMin of the search result. The cpMax values are very high numbers. The strange thing is that with version 1 the replacement happens, but not with version 2, even though the values are the same and the condition should always be false. What's going on?

But is it shipped with Windows 95? It's definitely not there when I try to call it using the command line.

nVU is based on Gecko, which before 1.9 is compatible with Windows 95, Windows NT 3.51 and newer. Kompozer is an unofficial version that's basically nVU + a set of patches to fix bugs, as the original developer has abandoned it.

For Win98's IE6? Because Win98's IE6 did not get the massive IE6 patch that came with XP's Service Pack 2, thereby separating the products on both platforms. The count of vulnerabilities doesn't matter. What matters is if it gets patched quickly. Microsoft has a horrible track record in this regard with IE. What also matters is the severity of the vulnerabilities, which IE also had a bad track record with. Also worthy of mention is that Mozilla-based web browsers are open-source, and that every security bug, including internally-found ones, are reported. Minefield builds come from the trunk, which is always in heavy development, except when frozen. It's no wonder that add-ons aren't marked as compatible with it.

What, is there something else in Windows 95 that does the trick?

Is it the one version with anti-phishing as mentioned in FF2 users? Yes. And that feature is being phased out.

Any browser that does not use IE's core. SeaMonkey, Firefox, K-Meleon, Opera,... It doesn't matter, as long as it doesn't use IE. That's why there are themes. Change to the Modern theme. It's also included. It's not hysteria. It's the biggest security hole of any Windows system. Simple. Change your user agent string. No holes have been plugged since the release of Service Pack 2 for XP, with a new edition of the IE6 browser. Security is the most important argument to not use IE6. But another argument that's almost equally important is web standards support. Compared to every other mature web browser, it's downright horrible. Bugs galore.

How do you activate word-wrap in a RichEdit control? Borland's WINAPI documentation only seems to mention how to replace the function used to do the actual wordwrapping.

The only program I use that checks for updates is SeaMonkey, which it's allowed to do. Other programs don't even know how; they're too old. AIM used to download ads to display on the buddy list window, but it seems to have lost that capability since removing IE. Wrong. Infections can be everywhere. Even on trusted sites. Maxthon is based on IE. You shouldn't use anything that uses IE's core.

I recall reading that Write.exe and Paintbrush.exe are 16-bit stubs that act as a wrapper for the 32-bit versions.

NEWS FLASH: Not everyone here is using Windows 98SE!

Thanks for the heads-up. Interesting to know. I don't think most users here would need that feature, though.

Ah, I see. I've learned something new about the Windows API. code65536, this is for Windows 95, where not all the W APIs are available.

Thanks, that difference did bite me. What a nasty default behaviour. I wouldn't know how I'm exceeding any bounds. The char array I have is definitely big enough. Here's my source code if you have time to look into it. I do have the Turbo Debugger, but I can't figure out how to work with it.

When I replaced the RICHED20.DLL file I had by a more recent one, the search bug went away. But now it's back again! I've been trying, in vain, to make my program open a file specified at the command line. Here's what I've done to achieve that. In my WinMain I have: if (*lpCmdLine != NULL) strcpy(g_szFileName, lpCmdLine); If there's something in the command line, it gets copied into my global variable that holds the path to the opened file. It's defined near the top of my source as follows: char g_szFileName[MAX_PATH] = ""; This seems to work, as when I show the variable's contents using a MessageBox, I get what I inputted at the command line. Next, in my window process' WM_CREATE event, I do: if (*g_szFileName != NULL) { if (loadFile(g_szFileName) == TRUE) SetWindowText(hwnd, strcat(g_szFileName, " - Kladblok")); else MessageBox(hwnd, "Er was een fout tijdens het openen van het bestand.", "Fout", MB_ICONEXCLAMATION | MB_OK);; } But this doesn't work. For some reason when my program gets here g_szFileName is empty! I verified with a MessageBox.

Now I have a different problem. I want to disable the Save menu item when no changes have been made. First I tried to use EnableWindow for this purpose. That didn't work. So I did a Google search, and found that I needed to use EnableMenuItem. I used it like described on the page and in Borland's reference, and came up with this: EnableMenuItem(GetMenu(hwnd), ID_FILE_SAVE, MF_GRAYED); However, it still doesn't work! What am I doing wrong? By the way, since switching to Rich Edit Control 2.0, search is done backwards, with only one result found each time. What the hell is this?! EDIT: And now the menu item graying works fine. Did I forget to compile, or something? :/ EDIT: Indeed. I caught myself running my program after making changes without compiling first.

I just tried to find how to convert an ANSI string to a UTF-8 string, and there's no place that gives a clear answer in code. The closest I found was on CodeProject, and that was to represent UTF-8 in only the first 128 ASCII characters. I'd have to find out how to detect a UTF-8 encoded file upon opening, too. All this Unicode stuff is doing my head in. EDIT: You have to do MultibyteToWideChar and then WideCharToMultibyte. However, encoding to UTF-8 is only supported on Windows 98 and later, unless I use the Microsoft Layer For Unicode. I give up, unless someone has a work-around.

Which basically are the protocols I'm talking about. SMB is not enabled by default on Win9x, and it doesn't have RPC. That's only a problem if the malware gets in in the first place.