This is the mail archive of the
mailing list for the Cygwin project.
Re: Change in command line behavior between cygwin 1.3.18-1 and 1.3.19-1
Yes, I am using textmode mounts:
C:\cygwin\bin on /usr/bin type system (textmode)
C:\cygwin\lib on /usr/lib type system (textmode)
C:\cygwin on / type system (textmode)
c: on /cygdrive/c type user (textmode,noumount)
d: on /cygdrive/d type user (textmode,noumount)
e: on /cygdrive/e type user (textmode,noumount)
If I invoke the script using a POSIX path:
$ /cygdrive/d/cygcheck/testscript 2>&1 | od -c
If I invoke using a mixed path:
$ d:/cygcheck/testscript/ 2>&1 | od -c
0000000 d : / c y g c h e c k / t e s t
0000020 s c r i p t [ 2 ] : s e t :
0000040 - \r : u n k n o w n o p t i
0000060 o n \n
However, this behavior for mixed paths is only with 1.3.19 and 1.3.20. If I
am using 1.3.17 or 1.3.18, it works (has worked for months) as I expect.
I have been using the same pdksh all these months, and so that is why I
suspect the change of behavior is in Cygwin. (I looked through the pdksh
distribution, and could not find automode.c or automode.o )
Also, please see the strace snippets I posted yesterday. Why is the file
(now) opened in text mode in one case and binary in the other?
pablo at projtech dot com
At 07:12 PM 2/24/03 -0500, you wrote:
>I was able to reproduce your output on Cygwin 1.3.20. The invokation
>method (i.e. POSIX vs. mixed path) did not matter, though the DOS line
>Try invoking your script as
>C:/cygtest/testscript 2>&1 | od -c
>and look at the result - it should be very informative. I got:
>$ c:/cygwin/tmp/kshtest/testscript 2>&1 | od -c
>0000000 c : / c y g w i n / t m p / k s
>0000020 h t e s t / t e s t s c r i p t
>0000040 [ 2 ] : s e t : - \r : u n
>0000060 k n o w n o p t i o n \n
>Also, what does your mount table look like? Do you, by any chance, have a
>text-type mount for /cygdrive? I'm assuming your pdksh is self-built.
>Did you link with automode.o?
>On Mon, 24 Feb 2003, Paul Coltrin wrote:
>> Thanks for the reply.
>> Unfortunately, the behavior also exists in 1.3.20. I first found it there
>> and then worked back to narrow down in which version it appeared.
>> In order to see the problem, the script needs to have DOS-style
>> line endings, and also it must invoked with an MKS-style path,
>> ie with a Windows style drive letter but with unix forward slashes -
>> C:/cygtest/testscript .
>> (These are both conditions I would like to preserve in order to minimize
>> porting and maitainability issues.)
>> Paul Coltrin
>> pablo at projtech dot com
>> At 06:27 PM 2/22/03 -0000, you wrote:
>> >> I am hoping that someone can explain this change of behavior I have
>> >> noticed that occurred between Cygwin version 1.3.18-1 and 1.3.19-1.
>> >> In order to remain compatible with MKS, and also to keep version control
>> >> as simple as possible we need to continue using ksh to execute our
>> >> On Cygwin we have been using the pdksh (
>> >> http://www.cs.mun.ca/~michael/pdksh/ ).
>> >> Here is a sample script:
>> >> ---
>> >> #!/bin/ksh
>> >> set +u
>> >> ---
>> >Hello, I just tried your testscript and it works under 1.3.20 . Perhaps
>> >upgrading to the latest version of the cygwin1.dll would fix your problem?
>> >Elfyn McBratney
>> >elfyn at exposure dot org dot uk
> |\ _,,,---,,_ pechtcha at cs dot nyu dot edu
>ZZZzz /,`.-'`' -. ;-;;,_ igor at watson dot ibm dot com
> |,4- ) )-,_. ,\ ( `'-' Igor Pechtchanski
> '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
>Oh, boy, virtual memory! Now I'm gonna make myself a really *big* RAMdisk!
> -- /usr/games/fortune
>Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
>Bug reporting: http://cygwin.com/bugs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Bug reporting: http://cygwin.com/bugs.html