Maximus5 wrote: |
---|
False, of course. |
Boy how I love it when smart people come to explain to me that things I've been doing for thirty years are impossible.
You might consider that I understand more about how the Windows Console window works than you do.
And that the thesis of my post is that
I cannot get the same behavior from ConEmu as I do the standard Windows console.
And "start" is
not more correct. It is a different way of doing things. Which, by the way, I've tried variations of as well. The "start" command isn't the problem. ConEmu is.
If you want to help, please don't belittle me or explain what I've said is happening cannot.
coder777 wrote: |
---|
Did you try the redirection? |
I have now, but I didn't expect it to work (which is why I didn't try it before).
The reason a terminal "locks" when you run a process in *nix has less to do with the standard I/O than the fact that the parent process is waiting on the child to terminate. This relationship of child to parent is what you typically call a "foreground" process. A "background" process (started with a trailing ampersand &) causes the parent (shell) to launch the child asynchronously -- meaning it isn't simply waiting for the child to terminate.
In Windows, processes don't actually have parent-child relationships, though the "parent" can still behave in much the same way with respect to "children" it starts.
It seems ConEmu tries to inject every process it runs and parent it in a tab. You can specify "special" command-line arguments to make it (not) do stuff like that, but even in *nix I don't have to type so much junk just to launch a separate process.
And it still doesn't quite behave correctly.
No one has experience with this?