Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com To: cygwin AT cygwin DOT com From: Rui-Tao Dong Subject: Re: xemacs shell problem Date: Wed, 19 May 2004 14:53:36 -0700 Lines: 82 Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Complaints-To: usenet AT sea DOT gmane DOT org X-Gmane-NNTP-Posting-Host: 216-237-9-34.orange.nextweb.net User-Agent: Gnus/5.090008 (Oort Gnus v0.08) XEmacs/21.4 (Security Through Obscurity, i686-pc-cygwin) Some old XEmacs (21.1 era) was handling this correctly, so I believe the problem is in XEmacs. I haven't had a chance to trace XEmacs code yet. I have been using the following work around. None of which is perfect. 1) Stick with an old XEmacs. 2) Do run /usr/local/bin/xemacs. You can search google for cygwin run. However, although it works the first a few times, the problem re-appears after a while. 3) Instead of compile with nmake -k (e.g.), do ssh localhost "cd $PWD; nmake -k" -- Rui-Tao Dong rdong AT trestlecorp DOT com >>>>> "sly5+gmane" == Levent Yilmaz writes: sly5+gmane> Hi! After posting this message to emacs.xemacs.windows sly5+gmane> and following the advices I have been told, I found out sly5+gmane> that this happens only with Xemacs Cygwin build. Thus, I sly5+gmane> hope it would be a relevant discussion for this group. sly5+gmane> *the problem setting* The shell-mode(using bash) at sly5+gmane> Xemacs(21.4.15) called from Bash (2.05b.0(1)-release) sly5+gmane> running Cygwin(1.42) on WindowsXP-SP1 sly5+gmane> *the ISSUE* Error messages from certain Windows console sly5+gmane> programs (such as the Fortran compilers LaheyF95 and sly5+gmane> Compaq Visual Fortran 6.6) (hereafter I will refer these sly5+gmane> as Compiler) are not displayed at the sly5+gmane> shell-buffer. Instead the errors are flushed to a DOS sly5+gmane> console that dissappears immidiately after the sly5+gmane> compilation is finished and the regular messages (logo sly5+gmane> and stuff) are displayed at the shell-buffer as usual. sly5+gmane> *complications* - When the Compiler is invoked directly sly5+gmane> from Bash, everything works fine, all messages are there sly5+gmane> and in fact, output redirection works as expected. The sly5+gmane> error mesages can be redirected with the usual 2>myfile. sly5+gmane> - This issue does not occur with all console applications sly5+gmane> (for instance I tried with MSVC++6/7 and they are working sly5+gmane> fine). Among the ones I tested, only these fortran sly5+gmane> Compilers has this behaviour. This is also why I fail to sly5+gmane> find a reproducible example for submitting as a test case sly5+gmane> here (though, you may download a trial version from sly5+gmane> http://www.lahey.com/ and test with a erroneous Fortan sly5+gmane> code, if you really agree that this is a significant sly5+gmane> issue and worth the trouble). Here is a sample failing sly5+gmane> FORTRAN code !====================== program fail end sly5+gmane> program nail !====================== - As I mentioned sly5+gmane> before this problem happens only with Xemacs Cygwin sly5+gmane> build. Native-Windows build is working fine. sly5+gmane> My question is, it seems that the output from the sly5+gmane> Compiler is flushed to that blinking console, what may be sly5+gmane> the cause of this? Are there other outputs other than sly5+gmane> STDERR and STDOUT, that Cygwin is handling properly, but sly5+gmane> the shell-mode in Xemacs is failing? sly5+gmane> Thank you so very much!! - Levent -- Rui-Tao Dong (949) 673-1907 x116 (949) 673-1058(Fax) rdong AT trestlecorp DOT com ALL INFORMATION IN THIS EMAIL OR ATTACHED TO THIS EMAIL IS CONFIDENTIAL. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/