This is rather late because I only follow the digest, but I've looked
into Perl/TK a bit with Cygwin.
Initially I found it all to be fairly disappointing, not least because
when I finally got things working Perk/TK gave limited widgets and was
prone to crashes.
My current solution is to use
I am trying to install Perl's Tcl-Tk-0.97 package off CPAN
(http://search.cpan.org/~vkon/Tcl-Tk-0.97/).
The Makefile.PL does a test for something called snit in the TCL libraires.
= it runs tclsh with a small script which does package require snit.
This fails.
I am not a regular TCL/TK user
On 10/9/07, Jason Pearce wrote:
I am trying to install Perl's Tcl-Tk-0.97 package off CPAN
(http://search.cpan.org/~vkon/Tcl-Tk-0.97/
http://search.cpan.org/%7Evkon/Tcl-Tk-0.97/).
The Makefile.PL does a test for something called snit in the TCL libraires.
= it runs tclsh with a small
This is resolved.
Thanks David for correcting my misaprehension that Tcllib was the
standard TCL library. Once I got past that things fell into place...
Tcklib is architecture independant tcl code. Its available here
This is just a post back from my earlier question.
I discovered that my direct to disc utility creates a UDF file system
on my DVD burner. I never realised these utilities used something of a
standard... Anyway that allowed me to experiment with UDF on my XP machine.
XP has the fsutil
I found this once:
http://sourceforge.net/project/showfiles.php?group_id=99645package_id=112046
I tried it back at version 800.025-2 , where both the X and native
windows bindings worked. Well, to be honest I didn't use it much but I
built some test cases with specTk and they ran OK. I was
I have a removable disc (IOMEGA REV), and I want to use cygwin to take
backups using cp -l and rsync as per
http://www.mikerubel.org/computers/rsync_snapshots/
This requires hard links to work, which they do on the NTFS partitions,
but the IOMEGA has a UDF file system and they don't seem to
Dan,
Sorry if you posted this to the list and I didn't reply. I have had very
little time to keep up with the Cygwin list lately.
I have uploaded Reini's version of Win32::API to a http server.
http://www.btinternet.com/~jasebob.pearce/cygwin/Win32-API-0.42.tar.gz
There is also the cygwin
This is just a bit of an update.
I have got Win32::SerialPort working, at least a limited capacity.
After taking a look at the source for it, I discovered that this module
is actually just straight Perl, and it relies on Win32::API. I have
already been using /Reini's cut down version of this
Igor Pechtchanski wrote:
Just FYI: /dev/ttyS* devices are numbered from 0, so /dev/com1 corresponds
to /dev/ttyS0... Also, if you have a sequence of shell commands that
should work but don't (e.g., stty followed by echo followed by read/cat),
could you please post them, along with an
manually?
If yes, How, and what must I have in consideration?
--- Jason Pearce [EMAIL PROTECTED] escribió:
Yes, use the CPAN module as Brian suggests. Just
make sure environment
variable PERLIO is unset when using -MCPAN,
otherwise it will fail. (You
might set this to CRLF for DOS
Yes, use the CPAN module as Brian suggests. Just make sure environment
variable PERLIO is unset when using -MCPAN, otherwise it will fail. (You
might set this to CRLF for DOS compatibility purposes).
Also be aware that you may have trouble building some modules that use C
code and have not been
On May 20 19:25, Jason Pearce wrote:
There was some recent discussion about serial ports and Perl.
http://sourceware.org/ml/cygwin/2005-05/msg00013.html
But I did not see any resolution.
Did anyone end up porting either Win32::SerialPort or Win32API::CommPort
as Reini suggested?
I
There was some recent discussion about serial ports and Perl.
http://sourceware.org/ml/cygwin/2005-05/msg00013.html
But I did not see any resolution.
Did anyone end up porting either Win32::SerialPort or Win32API::CommPort
as Reini suggested?
I need to talk to a thermometer via a serial link.
OK. It was just lying around in my setup for bygone days then.
Thanks for clearing it up!
Regards,
Jason.
On Thu, 12 May 2005, Jason Pearce wrote:
I have a user reporting that they are not getting a mount to the fonts
directory after installing.
Where does this mount actually come from? I cant
I have a user reporting that they are not getting a mount to the fonts
directory after installing.
Where does this mount actually come from? I cant see it by greping any
of the postinstall scripts.
Regards,
Jason.
I have been using bash within my Cygwin environment thus far.
But all are other UNIX machines are set up to use /bin/csh so I am
considering making the shift to either /bin/csh or /bin/tcsh under cygwin.
Preliminary experiments look good, but I have the feeling the majority
of Cygwin users are
Vince wrote:
o Do most new/upgraded packages work or do you always have to
port setup scripts?
Sorry, I don't get what you mean
Some packages come with install scripts. I guess running them is
independant of the shell you choose so they just work.
Other than that there are just a few
I've looked around the Cygwin site for information on installation
that doesn't use the supplied setup.exe, and I've found a lot of you
on your own so I'm hoping to get some help here. Thanks!
What so wrong with setup.exe?
I am doing a similar cygwin roll out at my work site and I use
FYI - I use rsync with checksums to synchronise from my laptop to work.
It syncs to a HP binary of rsync at the other end. I found I had to use
checksums because the time stamping was unreliable, transfering more
than it needed to. I have found it to be very reliable since switching
to
further, but for now I am removing that binary.
Jason Pearce wrote:
Here is a link to a compiled version of my patch from a few days ago.
While I can't imagine what could go wrong, I dont have any detailed
knowledge of setup.exe and I have not tested this compile very
thoroughtly yet! (ie you have
further, but for now I am removing that binary.
Jason Pearce wrote:
Here is a link to a compiled version of my patch from a few days ago.
While I can't imagine what could go wrong, I dont have any detailed
knowledge of setup.exe and I have not tested this compile very
thoroughtly yet! (ie you have
Here is a link to a compiled version of my patch from a few days ago.
While I can't imagine what could go wrong, I dont have any detailed
knowledge of setup.exe and I have not tested this compile very
thoroughtly yet! (ie you have been warned, but use if it helps).
I want to use setup.exe behind a firewall where it cant see the list of
mirrors.
This patch creates a -m option for no-mirror. In this case the
application will not try to get a list of sites and the user is obliged
to provide a valid site with the existing -s switch.
BTW - The -h help option
Mmmm. Leftovers. I'll clean that out in the next release.
CHere didn't work because you either have an old or custom /etc/profile.
I fell fowl of the customized /etc/profile too. I think it is worth a
note in the man page (or /usr/share/doc/chere?) in next release.
Regards,
Jason
--
Stephan,
I don't recall -- did you attach the output of cygcheck -svr at some
point (if so, just point to it in the archives, no need to repost)? My
guess is that this may be a textmode vs. binmode issue. Try making your
home directory a binmode mount, or setting PERLIO=crlf and see if that
Reini Urban wrote:
Maybe I'll come to the pending Win32::API problem with the callbacks
on gcc also. A simple hack30-like solution should always be possible.
Or a fallback to C::Dynalib.
Win32::OLE didn't compile OOTB, but this is easy. Just forgot where I
put the patch.
And your
Reini Urban wrote:
Jason Pearce schrieb:
Reini Urban wrote:
Maybe I'll come to the pending Win32::API problem with the callbacks
on gcc also. A simple hack30-like solution should always be
possible. Or a fallback to C::Dynalib.
Win32::OLE didn't compile OOTB, but this is easy. Just forgot where
. I dropped the Environment key from the registry
path for setting system environment variables. It should have read.
regtool set -s 'HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session
Manager\Environment\FOO' 'BAR'
Jason
Igor Pechtchanski wrote:
On Sat, 4 Dec 2004, Jason Pearce wrote
I have two seemingly simple questions that I just can't find the answers
to. Any assistance would be appreciated
1. What's the right way to add add environment variables into the user
and system maps?
Currently I am using regtool to manipulate the registry. For instance to
set an environment
: 0.42-1
install: release/perl/perl-Win32API/perl-Win32API-0.42-1.tar.bz2 41047
3042c5bc313a37c4b1da641369211965
I've atached the .tar.bz directly to this message - its not that big,
hope its OK. But note it will probably only work for perl v5.8.2.
Reini Urban wrote:
Jason Pearce schrieb
the patch.
Jason
Gerrit P. Haase wrote:
Jason Pearce wrote:
For what its worth, here is the binary install I am using. Its
working nicely via setup.exe. I just pasted this at the bottom of my
setup.exe
Would you also send me a patchfile with all changes you finally used
to build the module, please
This worked a treat Reini, thanks very much for your help.
I need this on several machines at work, so I will package it up in a
tar.bz2 for use with setup.exe. Once tested I'll post it back here.
Regards,
Jason
Reini Urban wrote:
Jason Pearce schrieb:
I have been trying to compile up Win32::API
Has anyone used WinCVS 1.3 and managed to get the Python macros working
with Cygwin's Python?
I have tried entering D:\cygwin\bin\libpython2.3.dll as the python dll
under the WinCVS's admin=preferences=wincvs configuration but it does
not seem register as having Python available.
Its just
An easy way of deleting the YODL garbage is to use the VIM editor.
The following VIM script (call it patch.vim) does the job:
Q
%s/^\.YODLTAGSTART\./\.YYODLTAGSTART\./
%s/\.YODLTAGSTART\./\r\.YODLTAGSTART\./g
%s/\.YODLTAGSTART\. roffcmd //
%s/\.YYODLTAGSTART\. roffcmd //
%s/\.YODLTAGEND\.//
I am trying to install PerlMagick.
I first tried to do so automatically via CPAN and it failed. Then I
tried doing the build manually.
First up I get a bunch of warnings about not finding libs. From
http://sources.redhat.com/ml/cygwin/2003-07/msg01352.html I gather this
is erroneous, and wont
Hi,
I still use Active State Perl rather than cygwin's native Perl, the
reason being that I have trouble intalling Perl modules via cpan in
particular the MySQL DBIs DBD, cpan keeps finding other modules
which need installing before giving up. Whereas the AS PPM (Perl
Package Manager) works
Larry Hall wrote:
I specifically reinstalled my XP onto FAT32 because I dual boot Linux and the NTFS driver is not as good as the FAT32.
Why is that? And before you say that there's no support for NTFS under
Linux, see http://www.jankratochvil.net/project/captive/.
Granted it has
I see the same thing in GNU Emacs 21.2.1 and Cygwin 1.5.7, can't say I
ever noticed until now though.
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:
I am interested in doing exactly this.
I take you to mean that you downloaded the Tk source and compiled with
cygwin as per the normal instructions but built it without the x option.
Is there a binary for this compile I can install via setup.exe (or
planned?).
Jason
I downloaded Tk-800.024 and
Just a note to anyone setting PERLIO=crlf in their environment. It
breaks the CPAN module (and probably others).
Unset PERLIO before running
perl -MCPAN -e shell
Jason
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
I've been working through the past archives and discovered the change
made to the default of PERLIO as of perl 5.8.0-3. Setting PERLIO=crlf
solves my initial problems and all my scripts now run with the new
version of perl. So far so good, but I am not sure if I should set
PERLIO=crlf
Gerrit P. Haase wrote:
I guess the alternative is to leave perl to access the files in binary
mode and then modify the perl scripts to detect and account for what
ever line ends it finds. Is that the recommended approach rather than
using PERLIO=crlf?
To get the most out of PERLIO the
This indeed solves the problem and thanks for explaining the default
shell issue. I actually thought that bash was used under cygwin, and I
often construct syntax in a bash shell before using it in a makefile. It
usually works but you've pointed out why there will be subtle differences.
BTW
I have discovered a problem in cygwin's Make.
I am calling a sub-make that goes to another directory. That bit works
fine, but the Makefiles have a variable set from a shell command. That
shell seems to be spawned from the wrong directory. I have created a
small example illustrating the
45 matches
Mail list logo
