Idea for new feature - discuss

[Ben <2blkbelt nemontel net>]

MC has three config options for what to do after executing files by pressing return with the selection on them:

1 Close the shell every time, returning to MC

2 Leave the window open every time, press return to... return

3 Close the window only on "dumb" terminals (what?)

Number one leaves you unable to see the output of a command. So you type ls and it's kind of useless. Or any other command or script that has output you need to see.

Number two works for those things where you need to see output, but if you don't, and there are things you routinely execute that do good things for you that simply need to be done on command (like turning on house lights, my particular thang) then it eventually becomes a bit of annoyance that it doesn't just get done.

Now, I may be missing something buried in that "dumb terminal option", but it seems to me that this would be a nice solution:

You know how the F4 editor keeps track of what line you're on in which file? That's *lovely*.

Well, how about modifying option two so that if you press return, it does what it always did, that is, close the shell and return. But if you press something else, perhaps ESC or whatever, from then on, when that command is run, the shell closes when the run is done.

This implies you'd need an operation to remove the "close" status in case the cat stepped on the selected key at the wrong time, but that doesn't seem like a high hurdle, lots of room in the menus.

Another approach would be an option to pause only if the command has output other than CR and LF, which doesn't require saving file names anywhere, but I think that might be a little unsatisfactory as if a command might create output or not, you might get a little uncertain if you saw what you thought you saw.

Number three... I use the usual terminal types. Are they "dumb"? I have no idea. I don't see anything in the OS X shell preferences to variably designate the terminal as "dumb" or not based on what's executing, so I'm guessing this can't do what I'm thinking. Is it useful to anyone? If not, perhaps it could be the "pause only if command produces output" option instead.

So, anyone? Feasibility problems? Good? Bad? Stupid? Selfish? Did I completely miss or misunderstand a feature?

--Ben