wd:
I tried the official SourceForge v0.7 with it's .conf, sdl.dll, and most other files (03/02/2007).
I set ems=false and added my autoexec lines to start PADS.
I had to set PADS to VGA 640x480.
Without the mapper file (everything default), it acted the same as HAL9000's "CAD Edition" without the mapper file.
With the mapper file, it acted the same as HAL9000's "CAD Edition" with the mapper file.
(Two of Gulikoza's (May and July) and one of ykhwong's (June) acted slightly different. See below).
Qbix:
The mapper made the lines I posted above. I just copied them and bolded the important ones. I'll explain what I did in the mapper after some background info:
PADS supported the XT keyboard. The DOS version never added support for the 101-key keyboard, so the regular nav keys Insert-Page Down don't do anything in an MSDOS or Win 98 setup. The inverted-T cursors cause the screen cursor to scroll slowly and smoothly for small motions.
The very important functions for the numeric pad are as follows:
PADS-PCB provides a complete set of window control commands, based on
the numeric keypad located to the right of the main keyboard. (Note:
Your keyboard must have NUM LOCK turned off in order to access the
Windowing Keys.) The function of each key is as follows:
Num 1: Redraw the screen
Num 2: Move the window down
Num 3: Zoom out
Num 4: Move the window left
Num 6: Move the window right
Num 7: Show the entire board
Num 8: Move the window up
Num 9: Zoom in about the cursor
Num 0: Reposition window with cursor in center
Num . : Start creating a window box
With Num Lock on, the numeric keypad should generate numbers as expected rather than the windowing functions.
OK, here's what happens with HAL9000's CAD Edition and no mapper file:
1. When Num Lock is on, the number keys work. But when Num Lock is off; 1,3, 7, and 9 do nothing and the cursor keys (2, 4, 6, and 8) move the screen cursor in a slow smooth motion instead of the normal function which is to pan about a third of the screen. In fact, the numpad cursor keys are doing what the inverted-T cursor keys normally do.
The number pad won't ever start working for the nav functions. I get numbers or nothing. The numpad navigation functions never work with this build if there's no mapper file.
(With 2 versions of Gulikoza's and 1 of ykhwong's, the numeric pad is quirky. I usually have to turn the Num Lock key on and off several times and use the left or right cursor key (4/6) to move the screen cursor a ways and then toggle the Num Lock again. On ykhwong's, when it starts working, the Num Lock will be off as expected. On Gulikoza's, when it starts working, the Num Lock is on which is the opposite of what it should be).
2. It makes no difference if I start DOSBox with the Num Lock on or off.
But here is something strange: if I leave the program running and free the cursor by pressing the Windows key twice or Cntrl-F10, then make sure the Num Lock is ON, then click in the desktop or any other window, when I click back in my program to capture the cursor, the Num Lock is still on AND the numpad keys are all working as the windowing navigation functions should when the Num Lock is off! In other words, they are working properly, but they are doing what they should when the Num Lock is off even though the Num Lock is actually on. The bad thing is if I toggle the Num Lock even once, the numpad operation goes back to what I described in number 1.
So, I thought I should try the mapper. I exit PADS and go back to the DOSBox command prompt. At this point, sometimes I have to toggle the Num Lock to get it back in sync with the proper numpad function (numbers vs. nav.).
In the mapper I noticed that the numpad keys have only one thing bound to each. Example:
key_kp_7
Key [7]
I thought the Num Lock state or code was always read along with any numpad code so its proper function can be determined. But then I thought, maybe I'm wrong. Or maybe the XT keyboard did something different.
So, I thought I would add the nav code to each numpad key, . to 9, except 5.
I turned Num Lock off.
I highlighted the key, selected Add, and then pressed the corresponding physical key.
Ex.:
key_kp_7
Key [7]
Key home
I didn't add any codes to the regular nav keys (inv-T and Insert-Page Down).
I did add the Rt-Alt key to the Cap Mouse function so I can escape the window easily (same as VPC). And F10 is used to exit PADS, so I wanted to get away from it. One of the Windows keys seemed intuitive since mouse capture is really changing windows, but they weren't in the mapper.
(Request: could you add support for the two Windows keys in the mapper)?
I saved the mapper file and I'm ready to try it (making sure it's in the correct directory).
Here's what happens with HAL9000's CAD Edition and the mapper file:
The numeric keypad works as it should, generating numbers when Num Lock is on and nav. functions when Num Lock is off. However, many times the Num Lock must be toggled once or twice to get it in sync with the actual state or even to get the nav functions working. Once in sync, it stays unless I escape from the window, click in another window, and recapture in PADS. Then it may get out of sync again until toggled once or twice.
Here's the oddity:
The regular navigation keys, the inverted-T cursors and the Insert-Page Down keys, act as number keys when Num Lock is on. They provide the normal navigation functions when Num Lock is off. For example: the inverted-T cursors generate 2, 4, 6, and 8 when the Num Lock is on.
That shouldn't happen.
Even after I exit PADS and return to the DOSBox command prompt, this behavior continues. If Num Lock is on, pressing an inv-T cursor generates a number. So does Insert-Page Down. If Num Lock is turned off, they function normally.
So what is going on? Why do the regular nav. keys generate numbers when Num Lock is on? I only added an additional code to the numpad keys.
Since PADS didn't use the regular Insert-Page Down, I can keep Num Lock off in PADS and get the nav. functions on the numpad as is normal. If I want the Num Lock LED to be correct, I toggle the Num Lock key a couple of times and it's in sync. I just need to avoid using the inv-T cursors when Num Lock is on so I lose my small motion cursors for that state. So it's liveable but you might want to see what's happening in case it affects anyone else where they can't work around it.
I know this is long and I hope you can follow the steps I went through.
Thanks