I'm sorry Eugene, but you don't seem to get reasons. Just answer "What will happen to your software if you suddenly die ?" - I know it is extreme, but it may happen to you and me.
Thanks for sharing Sigurd
- Open source clones of OS/2 parts is a good idea, may be the only long shot strategy for OS/2 - but it seems to me a mountain to high to climb. It would need reverse engineering - and the well known problems to WPS Programming
The issue I think is that we don't have a long term strategy, and people that have corporate sponsors also do not have it. I understand the dream is big "AN OPEN SOURCE OS/2 CLONE!!!"... wow that's big.... But a big dream is just made of little efforts that adds up.
It will be as nice to start with PM (Dreaming in Technicolor) following a process like this:
- "ok, let's make an open source of PM". .... (wow that is too much...Big dream)
- "Here are the components of PM API" ....
link ... (wow there are too many components)
- "uhmmm I think that "Window_Functions" API compoment may be a good place to start" ....
link ..... (wow... that is still too much)
- "I'm going to start cloning WinInitialize()".....
link.... (that seems to be more manageable)
If WinInitialize() is cloned (maybe it had been done already on the past, I just grab it as example), and the source code shared, it will a little step, maybe useless right now, but it will be an step. The sum of little steps can pull a big dream. And maybe that little sum of small steps can inspire other people to join to make the little steps move faster.
-More realisitc allways seemd to me to be a way to clone the WPS on top of a Linux system, a "OS/2 distribution" - I would guess this would need less work than reprogramming OS/2 but one would get of the most driver problems and concentrate on fewer tasks
I also thought first about WPS. The best idea will be to extend the XWorkplace project and convert that project with the goal of replacing the WPS classes one by one with open source code. But after reading more about the OS/2 architecture, WPS is builded over SOM, and SOM/WPS runs over PM.
I got to the conclusion (I don't know if I'm complete right) that the base of OS/2 GUI applications is PM. The API that constructs how the OS/2 GUI looks is part of PM and it far more important that SOM/WPS. (personal opinion).
I got to that conclusion to see the quantity of WPS based applications that it exists. (
link) . Compared to the PM applications that runs without SOM/WPS, the WPS applications quantity is very small. The soul of running OS/2 GUI application is in fact PM.
What it will be incredible is to start cloning PM, and that IBM's OS/2 SOM and WPS keep running over it. SOM/WPS may be important for the OS/2's desktop, icons and but PM is real thing that has the OS/2 GUI soul. In final words (sorry I drift too much) PM is OS/2 GUI, WPS is the desktop shell and SOM is the OO framework on which WPS was constructed.
So, cloning WPS over linux may be able to produce a similar OS/2 desktop on Linux, but PM applications (The majority of GUI OS/2 apps) will be incompatible. PM is the the real thing, but to be honestly I will prefer to have a PM open source clone running over OS/2-ecs first, and if someone wants to port it over Linux, he is welcome to do it.
But is someone will like to start with WPS open source clone, I will also support that project !! I will support and help (in whatever I can) to anyone that produces a open source DLL and EXE that can replace any IBM close source component of OS/2.
I am looking forward to our user meeting in Cologne to discuss with other users in a gentle, realistic and calm way.
....but there is no problem also discussing here, I try to be always calm

Sometime the discussion may not be complete realistic, but I had found out that it is good to "dream big" and after that you start something small, but the "big dream" gives you the goal.