Ocassional disappearing writes to MOTHERBOARD parallel port, on Windows 10

[cappy2112]

Hello,


We have been using inpout32.dll on Windows 10 to write to motherboard parallel ports, without issue until very recently.
We have an LED connected to the parallel port pin being written to so we can visually verify when the writes are taking place.
The writes are to the data register only, the motherboard bios/UEFI are always configured to the most basic configuration possible,
which is SPP or Output Only. This has proven to be quite reliable.

The writes to the data register are very low frequency, there are usually seconds-minutes in between each write.

Within the last few months, we have seen that writes to the parallel port sometimes "get lost", or swallowed by Windows,
but we don't know why. We often need to do several writes to the data register before we see the LED (mentioned above) turn on or off.

We look for the parallel port address in control panel, and use that address as the Base register & Data register address.
This has worked reliably for us since Windows XP.

Chances are this is more of a Microsoft issue than a DLL issue, but I don't want to make that assumption without checking
here first.

Thanks

[phil]

I'm sorry, I don't really know what to say. I haven't used InpOut myself for years, and I am unable to continue support for it (software changes wise) as to make changes to the driver, requires new signing certificates that I can not even purchase as an individual!

But I see no obvious reason for it to stop working, unless changing Windows versions - I mean I tested it on Win 10 ages ago and it worked but Ive never tested it on Win 11 (for example). I'm afraid I cant really help much more that that as I don't use it myself anymore (and none of my machines even have a parallel port!)

[dbjh]

Just wanted to chime in to let you know I have experienced similar behaviour on 64-bit Windows 7. My hunch is that it is associated with 64-bit Windows. On the same machine that I noticed that behavior, things worked fine with an old Fedora version, suggesting that it had nothing to do with the hardware.
In my case I could use giveio64 as a replacement (using direct/raw I/O), seemingly confirming my suspicion that it was Windows that was interfering. You can find it here: https://www-user.tu-chemnitz.de/~heha/hs/giveio64.zip/.