W00T! Windows 95 runs on it!

[NothingMuchHereToSay]

I thought I did on the forums, but apparently I didn't look hard enough, I searched google for a patch for it though.

[rloew]

You didn't search the forum, did you?

I use the free hex editor XVI32 to do that. Most often to change the API call IsDebuggerPresent (not present in Win95, and most programs don't even use it) into ReadFile.

PassingBy is not around anymore, but BenoitRen remains with us, so if the above info is not enough (but it *is* a good start), you should ask him.

I'm not sure Patching "IsDebuggerPresent" to "ReadFile" is a good idea. The two API Calls have different numbers of arguments. This will cause a Stack error when the API Call returns. It is also possible that an incorrect File Read command could be executed. I would suggest "GetLastError" instead.

[SomeGuy]

To patch OpenOffice 2.3 or 2.4 to run on 95: Use a hex editor on sal3.dll to replace "IsDebuggerPresent" with "ReadFile" (or "GetLastError" or whatever fits). Then on vcl680mi.dll replace "EnumDisplayMonitors" with "EnumDesktopsA". Put hex zeros over the characters that are not written over. You will also need to obtain msimg32.dll, and install Java Runtime 1.4.2 if you want the wizards to run.

In the case of IsDebuggerPresent, it never gets called. Newer MSVC puts it there even when it is not needed. EnumDisplayMonitors, don't know if that gets called but never ran in to it. It is nicer to have something that matches up parameters, but that is not always possible, and will sometimes crash anyway as the results might be different.

[Mijzelf]

You didn't search the forum, did you?

I use the free hex editor XVI32 to do that. Most often to change the API call IsDebuggerPresent (not present in Win95, and most programs don't even use it) into ReadFile.

PassingBy is not around anymore, but BenoitRen remains with us, so if the above info is not enough (but it *is* a good start), you should ask him.

I'm not sure Patching "IsDebuggerPresent" to "ReadFile" is a good idea. The two API Calls have different numbers of arguments. This will cause a Stack error when the API Call returns. It is also possible that an incorrect File Read command could be executed. I would suggest "GetLastError" instead.

GetLastError() output is unpredictable. I don't know why OOo would call IsDebuggerPresent(), but it could be bad news when two subsequent calls give a different result. I recommend GetCurrentProcess(), which always evaluates to TRUE, or GetDaylightFlag(), which gives FALSE for half of a year, and TRUE the other half, and can be forced to always give FALSE. (Just disable daylight saving time)

[cov3rt]

On 12/2/2007 at 7:42 AM, PassingBy said:

Hmmm .....

Can you fill in the blanks please ?

1. Motherboard model

2. Chipset

3. VGA chipset

4. Audio chipset

5. NIC if available

On VIA chipsets, older drivers version 4.43 seems to works best (may need to patch in the irq driver manually) ... Nvidia drivers can be made to work fully (driver+cpl) on win95 upto version 61.77, above that requires some patching upto 81.98 ... Haven't tried ATI or Unichrome drivers yet ... Audio for realtek is version 4.03 vxd ...

Rgds

i know this thread is very old but i didn't want to make a new thread on working out the issues i need with a particular system. i have used via 4in1 version 4.43 on the laptop im working with which is averatec 3200 on windows 95, i believe it uses the kt, kn or km400 chipset or some via chipset, anyways, the problem is that although the 4.43 driver package does install important things like the agp 3.0 driver where i can get the gpu driver to work, etc, however, it doesn't seem to update the ide driver, or if it does, it shows the name as via hard disk controller, but doesn't show any date or updated file driver, there is other issues with it too. 

another issue is the cardbus devices will not properly work / configure. installing the drivers gives a code 29 error in device manager and i know the drivers are for windows 95. i've tried researching this issue, saying it may be a bios or allocation issue, however i couldn't find anything in the bios that would affect this. i've looked up pci routing as it seemed to explain similar issues i was having or possibly the same issues which the pci bus handles such as code 29 error with cardbus devices, unfortunately, applying the changes given by the website did not change anything. 

i also had problems with the usb controllers, which i seem to have with other computers that also used via usb controllers with dev 3038. it will install the controllers, but will give a yellow exclamation mark and it wont install the drivers, it just shows the controllers in yellow markings in device manager. 

back to the cardbus issue, it is dev 6972 and ac50 cardbus, the 6972 is the one the laptop internally uses and the ac50 is a supplement to the toshiba e5d7 mini pci wireless adapter im trying to use, in other words, to detect the toshiba card, the ac50 cardbus driver has to first be installed and properly working which it doesn't. 

i was gonna start fresh again, wipe the hard drive and install with ACPI disabled this time, im hoping maybe combining that and applying some other changes, i may be able to get the system to properly work or configure things better. one of which to be specific and why i replied to this post is to patch in the updated irq driver which i was gonna ask if anyone knows how to that as i don't know how to do that. i feel as though this could be the reason why everything in the system isnt running the way it should, being that i always install chipset drivers first, then the unofficial usb update ( xusbsupp ), and so on.