[Idle-dev] [ idlefork-Bugs-747772 ] RedHat Linux 8.0 problems
SourceForge.net
noreply@sourceforge.net
Thu, 05 Jun 2003 19:15:06 -0700
- Previous message: [Idle-dev] CVS: idle PyShell.py,1.73,1.74 rpc.py,1.25,1.26 run.py,1.21,1.22
- Next message: [Idle-dev] CVS: idle EditorWindow.py,1.49,1.50 FileList.py,1.5,1.6 WindowList.py,1.4,1.5
- Messages sorted by:
[ date ]
[ thread ]
[ subject ]
[ author ]
Bugs item #747772, was opened at 2003-06-02 16:41
Message generated for change (Comment added) made by kbk
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=109579&aid=747772&group_id=9579
Category: None
Group: None
>Status: Closed
>Resolution: Fixed
Priority: 5
Submitted By: Bruce Sherwood (bsherwood)
>Assigned to: Kurt B. Kaiser (kbk)
Summary: RedHat Linux 8.0 problems
Initial Comment:
When installing the latest CVS on RedHat Linux 8.0,
setup failed because it was looking for CREDITS.txt,
HISTORY.txt, INSTALL.txt, LICENSE.txt, NEWS.txt, and
README.txt, all of which were present but with
lower-case file names (which didn't matter on Windows).
The Help for Idle displays "\r" at the end of each line.
I always get the "cannot open socket" message,
sometimes when trying to start Idle, but sometimes not
until I press F5 to run. I don't have this problem with
the old original idlefork.
I had a nonreproducible problem with keyboard shortcuts
such as Ctrl-C for copy. For a while they worked only
for items on the Format menu. Now they all work.
Thought I should at least report it.
----------------------------------------------------------------------
>Comment By: Kurt B. Kaiser (kbk)
Date: 2003-06-05 21:15
Message:
Logged In: YES
user_id=149084
Linux issue was 'localhost' not resolving to
'127.0.0.1' . Avoided by defining a LOCALHOST
variable.
----------------------------------------------------------------------
Comment By: Bruce Sherwood (bsherwood)
Date: 2003-06-04 11:19
Message:
Logged In: YES
user_id=34881
Here is what I get both before and after running idlefork
from lsof -i:
COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
portmap 530 rpc 3u IPv4 926 UDP *:sunrpc
portmap 530 rpc 4u IPv4 927 TCP *:sunrpc
(LISTEN)
rpc.statd 549 rpcuser 4u IPv4 1000 UDP *:1024
rpc.statd 549 rpcuser 6u IPv4 1003 TCP *:1024
(LISTEN)
sshd 678 root 3u IPv4 1324 TCP *:ssh
(LISTEN)
xinetd 692 root 5u IPv4 1364 TCP
127.0.0.1:1025 (LISTEN)
X 853 root 1u IPv4 1589 TCP *:x11
(LISTEN)
fam 943 root 0u IPv4 1364 TCP
127.0.0.1:1025 (LISTEN)
fam 943 root 1u IPv4 1364 TCP
127.0.0.1:1025 (LISTEN)
fam 943 root 2u IPv4 1364 TCP
127.0.0.1:1025 (LISTEN)
I don't see how to get this info at the instant that I press
F5, though I did clumsily try to do this. Is there some test
of that kind you'd like me to do? How do I do it?
----------------------------------------------------------------------
Comment By: Kurt B. Kaiser (kbk)
Date: 2003-06-03 18:44
Message:
Logged In: YES
user_id=149084
Please start IDLEfork from the command line and
post all the error messages here.
If you have lsof installed, please post the output
of lsof -i after the error occurs, also.
----------------------------------------------------------------------
Comment By: Bruce Sherwood (bsherwood)
Date: 2003-06-03 14:07
Message:
Logged In: YES
user_id=34881
Guido's right -- my fault for passing through Windows on the
way to Linux. And I should have checked CVS to see the names
there.
Today I used IDLEfork-0.9b1 to install on Linux without
difficulty, but I still can't run a program. If I set
preferences to start in an edit window, I can edit, but as
soon as I press F5 I get the socket error. If on the other
hand I set preferences to start in the shell, I get the
socket error immediately, before any window opens. ps fx
shows no instances of Python lying around. I don't know what
to try. I repeat that the VPython version of Idle continues
to work on my Linux machine.
----------------------------------------------------------------------
Comment By: Guido van Rossum (gvanrossum)
Date: 2003-06-03 12:24
Message:
Logged In: YES
user_id=6380
Looks to me like maybe Bruce got the files on Windows and
transferred them to Linux from there (or maybe shared them
using VMWARE or SAMBA).
----------------------------------------------------------------------
Comment By: Kurt B. Kaiser (kbk)
Date: 2003-06-03 12:14
Message:
Logged In: YES
user_id=149084
First, the CVS issues: If you have lower case .txt
files (except for help.txt) there is something wrong
with your cvs client or other setup. The repository
is correct. Also, the line endings for help.txt in the
repository are 0x0a (\n) with no \r involved. So
again, it sounds like a cvs client problem.
The socket problem: Since you are struggling a bit
you may have moribund IDLEfork processes. Do a
ps fx and kill any you see, then try again. Once the
first instance of IDLEfork is started I would like all
subsequent attempts to fail cleanly, but that is still
not completely robust. Once it is, we can move on
to Michael William's request to run multiple copies
of IDLEfork simultaneously. (To give you an idea of
how flakey this is, /once/ on WinXP I had two copies
running with one printing a string of 1's and
the other a string of 2's. But it's not supposed to be
possible to have two processes setting up connections
on the same socket. And I can't reproduce it.)
When this kind of thing happens to me it's usually
because I created a bug that caused IDLE to fail while
initializing, and the stuck process prevents the next
instance from initializing. ps fx and kill are your friends.
I haven't seen anything like the last problem. It's
likely related to incomplete initialization of IDLE on
startup.
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=109579&aid=747772&group_id=9579
- Previous message: [Idle-dev] CVS: idle PyShell.py,1.73,1.74 rpc.py,1.25,1.26 run.py,1.21,1.22
- Next message: [Idle-dev] CVS: idle EditorWindow.py,1.49,1.50 FileList.py,1.5,1.6 WindowList.py,1.4,1.5
- Messages sorted by:
[ date ]
[ thread ]
[ subject ]
[ author ]