[Idle-dev] Anticipated changes for Mac OS X

[Michael Williams]

On Saturday, November 30, 2002, at 04:44  pm, Guido van Rossum wrote:
>> I've never really used idle, so don't expect great testing from me.
>
> Maybe you can give it a try.  It would be cool if it could optionally
> apply PyChecker to a module.  There's a "Check" menu entry already;
> currently it just does a syntax check and a tabnanny check.

Seconded. It would be very cool, and certainly something we'd want to 
use in the Oxford course[0].

[0] <http://users.ox.ac.uk/~sann1276/python/>

We're really looking forward to the next IDLE release here, as we've 
felt unable to use the (very active) CVS over the past few months on 
our students. I pulled it down this afternoon to see what had changed 
and have a couple of queries:

+ Where have the Run and Edit menus gone? Is this a temporary thing 
while CVS is in flux?

+ I checked whether the server-client overhaul which has (presumably!?) 
taken place solved our show-stopper (see 
<http://mail.python.org/pipermail/idle-dev/2002-May/000958.html>) 
regarding multi-user machines. It does seem to although I'm a bit 
confused as to the origin and seriousness of the error message one gets 
the second time it launches:

first time:
% ./idle
Idle accepted connection from  ('127.0.0.1', 41902)

[should it even say this? is it debugging stuff that will be removed?]
------
second time, different user, different terminal:
% ./idle
.  .  .
Idle socket error: Address already in use, retrying...

Idle socket error: Address already in use, retrying...

Idle socket error: Address already in use, retrying...
IDLE cannot run.

IDLE needs to use a specific TCP/IP port (8833) in order to execute and
debug programs. IDLE is unable to bind to this port, and so cannot
start. Here are some possible causes of this problem:

   1. TCP/IP networking is not installed or not working on this computer
   2. Another program is running that uses this port
   3. Personal firewall software is preventing IDLE from using this port

IDLE makes and accepts connections only with this computer, and does not
communicate over the internet in any way. Its use of port 8833 should 
not
be a security risk on a single-user machine.
-----

This takes several seconds on our machine (presumably probing ports). 
Despite the protestations that something is broken it then launches a 
new instance of idle in the expected way: on the second users terminal. 
What's the significance of this error message?

By the way, if anyone hasn't seen it yet, I produced a detailed report 
on the successes and failures of our trial of Python as a teaching 
language, using IDLEfork 0.8.1 as the IDE. It's here: 
<http://users.ox.ac.uk/~sann1276/python/report/>. Perhaps of particular 
interest to members of this list are:

+ <http://users.ox.ac.uk/~sann1276/python/report/node17.html>, which 
describes the temporary workaround we implemented to fix the static 
port problem.
+ <http://users.ox.ac.uk/~sann1276/python/report/node27.html>, which 
discusses idleforks strenghs for our purposes.

I would still like to see some sort of warning (that could be turned 
off in Settings, but that was on by default) when a file is saved 
without a Python extension: "You are saving this file without a .py 
extension. This extension indicates that IDLE should treat it as a 
Python program. Are you sure what you are saving is not a Python 
program?"

-- 
Michael