[Idle-dev] Separate process; automatic save for a run

[Bruce Sherwood]

Guido said:

> In the mean time, I agree with you that this area of IDLE needs a lot
> of improvement.  I have some ideas (with Jeremy) for completely
> revamping this part of IDLE, which involve running the program in a
> separate process altogether (the only way to get everything to work
> right).  This would require a lot of restructuring, and the code you
> are patching would probably disappear completely...
> 
> I'm not sure what to do about your proposed patch, but I do want to
> say that I am *against* saving without asking the user.

Note that David Scherer has already formally submitted changes for IDLE so
that your program does run in a separate process. Some of us have been
using this version for many months with great pleasure; it is included in
the VPython package (which offers easy real-time 3D output from Python). See

	http://cil.andrew.cmu.edu/projects/visual

Another important feature of Scherer's version of IDLE is that you just
press F5 and the file is automatically saved and run (in a separate
process). This is a different protocol than I have used in other
programming environments, but I find that I like it a LOT in practice.
Asking the user whether to save is very intrusive by comparison. I do agree
that it would be even better for the default save to go to something.new,
and upon closing the file the user could be asked whether to copy
something.new to something.py. (Obviously with the current protocol I make
a backup copy before doing extensive editing.)

I can say that I really hated the old IDLE protocol of asking me about
saving every time I wanted to run. Scherer's change made a huge improvement
in interactivity.

Bruce Sherwood