FvwmCommandS Problem
edta...@value.net
"FvwmCommandS: dead pipe"
I assume that is bad.
Using FvwmCommand results in an error of:
"No such file or directory FvwmCommand error in opening message fifo"
This error is given with ANY command I try.
FvwmCommand "Quit"
FvwmCommand -i2 'FvwmIconBox'
FvwmCommand 'popup "Utilities"'
PS: this is version Ver. 1.5 of FvwmCommand, ver. 2.0.46 of Fvwm2, libc6
Help Me.
--
Visit the official FVWM web page at <URL:http://www.hpc.uh.edu/fvwm/>.
To unsubscribe from the list, send "unsubscribe fvwm" in the body of a
message to majo...@hpc.uh.edu.
To report problems, send mail to fvwm-...@hpc.uh.edu.
Bob Woodside
On 18-Aug-98 edta...@value.net wrote:
> When FvwmCommandS is started it exits or dies with the message:
> "FvwmCommandS: dead pipe"
> I assume that is bad.
That is bad. (That's what FvwmCommandS says when you kill it
normally with KillModule, though.)
I have no idea what's going on here. I've never had it fail to
start up, so any suggestions I make are pure guesswork.
1) Was FvwmCommandS compiled at the same time as the main part
of Fvwm - same compiler release, libs, etc.?
2) Someone else recently reported that he couldn't get FvwmCommandS
to start properly from his InitFunction (Vikas, are you following this one,
and have you made any progress?) If that's how you're starting it, try
putting it into a menu and invoke it after Fvwm gets started up. If this
makes a difference, at least we have a clue about where to look. I'm not
sure what we'd be looking for - some kind of timing quirk, I guess - I can't
make it fail this way, either.
> Using FvwmCommand results in an error of:
> "No such file or directory FvwmCommand error in opening message fifo"
> This error is given with ANY command I try.
This is the normal response from FvwmCommand if FvwmCommandS is
not active.
> PS: this is version Ver. 1.5 of FvwmCommand, ver. 2.0.46 of Fvwm2, libc6
This is curious - where did your Fvwm come from? The FvwmCommand in
my copy of Fvwm 2.0.46 identifies itself as version 1.4. Unless that's a
typo, there may be something fishy here.
edta...@value.net
>
> This is curious - where did your Fvwm come from? The FvwmCommand in
> my copy of Fvwm 2.0.46 identifies itself as version 1.4. Unless that's a
> typo, there may be something fishy here.
I got version 1.5 here:
http://www.concentric.net/~isogai/fvwm2/fvwm2.shtml
The basic fvmw2 and FvwmCommand are compiled now, with the same compiler.
gcc version egcs-2.90.29 980515 (egcs-1.0.3 release)
Still the same results: dead pipe
Method of starting has no effect. All of the follow methods fail.
# Stuff to do at start-up
AddToFunc "InitFunction" "I" Exec xterm -geometry 80x25+3+2 -ls -sl 200
+ "I" Module FvwmButtons
+ "I" Module FvwmCommandS
AddToFunc "RestartFunction" "I" Module FvwmButtons
+ "I" Module FvwmCommandS
AddToMenu "Module-Popup" "Modules" Title
+ "CommandS" Module FvwmCommandS
I even used a striped down .fvwm2rc, same results.
Bob Woodside
On 19-Aug-98 edta...@value.net wrote:
>
> I got version 1.5 here:
> http://www.concentric.net/~isogai/fvwm2/fvwm2.shtml
>
> The basic fvmw2 and FvwmCommand are compiled now, with the same compiler.
> gcc version egcs-2.90.29 980515 (egcs-1.0.3 release)
>
> Still the same results: dead pipe
>
> Method of starting has no effect. All of the follow methods fail.
I'm still at a loss to know what's happening. I grabbed Toshi's
1.5 version and installed it, and everything still works fine on my system.
Duhhh...
There could be some Glibc-related quirk, but I can't help you there.
I'm still running a libc5 system, so I'm not up on which Glibc releases had
what bugs.
Since Toshi has obviously working on FvwmCommand, you might try
e-mailing him directly to see whether he has any ideas. I don't know how
regularly he follows the mailing list, though he does post here from time
to time.
Vikas Agnihotri
> On 18-Aug-98 edta...@value.net wrote:
> > When FvwmCommandS is started it exits or dies with the message:
> > "FvwmCommandS: dead pipe"
> > I assume that is bad.
> 2) Someone else recently reported that he couldn't get FvwmCommandS
> to start properly from his InitFunction (Vikas, are you following this one,
> and have you made any progress?) If that's how you're starting it, try
Well, no progress. But just yesterday my NT machine BSOD'ed and I had to
restart everything. Again, FvwmCommandS failed to start up via the
InitFunction. I had to start it via my menu entry.
So, coming to think of it, I have _never_ gracefully shutdown and restarted
Fvwm. It has always been due to some emergency, (NT crash, fvwm freeze
during Recapture or Restart, etc). BTW, I am running NT4.0 SP3 with
Netmanage Chameleon UnixLink as my X server.
Leads me to believe that the .FvwmCommandC/S pipes are never cleaned up and
so FvwmCommandS somehow croaks in the InitFunction. Dunno why the menu
entry works though!
Later,
Vikas
Toshi Isogai
>
> On 19-Aug-98 edta...@value.net wrote:
> >
> > I got version 1.5 here:
> > http://www.concentric.net/~isogai/fvwm2/fvwm2.shtml
> >
> > The basic fvmw2 and FvwmCommand are compiled now, with the same compiler.
> > gcc version egcs-2.90.29 980515 (egcs-1.0.3 release)
> >
> > Still the same results: dead pipe
> >
> > Method of starting has no effect. All of the follow methods fail.
>
> I'm still at a loss to know what's happening. I grabbed Toshi's
> 1.5 version and installed it, and everything still works fine on my system.
> Duhhh...
>
> There could be some Glibc-related quirk, but I can't help you there.
> I'm still running a libc5 system, so I'm not up on which Glibc releases had
> what bugs.
>
> Since Toshi has obviously working on FvwmCommand, you might try
> e-mailing him directly to see whether he has any ideas. I don't know how
> regularly he follows the mailing list, though he does post here from time
> to time.
>
I have never tested FvwmCommand on glibc myself. But, I exchanged emails
with Martin Bialasinski to make it work on his system, which is a glibc
machine. Version 1.5 includes that fix. (His problem was not "dead pipe"
by the way, but related with xdm.)
Checking back those emails, I did not get a confirmation from him if
he took the tar file from my web page and run as it is.
Can anybody else confirm if it works on glib systems ?
>
> --
> Visit the official FVWM web page at <URL:http://www.hpc.uh.edu/fvwm/>.
> To unsubscribe from the list, send "unsubscribe fvwm" in the body of a
> message to majo...@hpc.uh.edu.
> To report problems, send mail to fvwm-...@hpc.uh.edu.
>
-----------------------------------------------------
Toshi Isogai 半抄网汤
KENTEK Information Systems
iso...@kentek.com
Richard Lister
Vikas Agnihotri <Vik...@att.com>
wrote concerning 'Re: FVWM: FvwmCommandS Problem':
> On Tue, Aug 18, 1998 at 06:27:40PM -0400, Bob Woodside wrote:
>
>
> > On 18-Aug-98 edta...@value.net wrote:
> > > When FvwmCommandS is started it exits or dies with the message:
> > > "FvwmCommandS: dead pipe"
> > > I assume that is bad.
>
> > 2) Someone else recently reported that he couldn't get FvwmCommandS
> > to start properly from his InitFunction (Vikas, are you following this one,
> > and have you made any progress?) If that's how you're starting it, try
I seem to remember having trouble starting it from InitFunction. I now
just put the single line:
Module FvwmCommandS
in my .fvwm2rc, and that works fine.
Ric
Richard J Lister, Georgetown Institute for Cognitive and Computational Sciences
Georgetown University Medical Center, Washington, DC 20007, USA
Phone: +1 202-687 2878
Email: r...@giccs.georgetown.edu
Web: http://www.giccs.georgetown.edu/~ric/
Toshi Isogai
> Well, no progress. But just yesterday my NT machine BSOD'ed and I had to
> restart everything. Again, FvwmCommandS failed to start up via the
> InitFunction. I had to start it via my menu entry.
>
> So, coming to think of it, I have _never_ gracefully shutdown and restarted
> Fvwm. It has always been due to some emergency, (NT crash, fvwm freeze
> during Recapture or Restart, etc). BTW, I am running NT4.0 SP3 with
> Netmanage Chameleon UnixLink as my X server.
>
> Leads me to believe that the .FvwmCommandC/S pipes are never cleaned up and
> so FvwmCommandS somehow croaks in the InitFunction. Dunno why the menu
> entry works though!
>
> Later,
> Vikas
Maybe, you have already told us, but which version of FvwmCommand are you
using?
By a slight chance if it is 1.4 and you are using xdm, version 1.5
should fix it.
I don't think fifos are the problem in your case. They will be overwritten
when a new ones are created, as you run FvwmCommand from menu.
-----------------------------------------------------
Toshi Isogai 半抄网汤
KENTEK Information Systems
iso...@kentek.com
--
Vikas Agnihotri
> > Leads me to believe that the .FvwmCommandC/S pipes are never cleaned up and
> > so FvwmCommandS somehow croaks in the InitFunction. Dunno why the menu
> > entry works though!
> Maybe, you have already told us, but which version of FvwmCommand are you
> using?
> By a slight chance if it is 1.4 and you are using xdm, version 1.5
> should fix it.
I am using the FvwmCommandS that comes bundled with fvwm 2.0.46. I think it
is 1.4. I am not using XDM.
Where can I download newer versions from? Do fvwm2 modules have their own
home pages hosted by the individual module authors? Pointers?
> I don't think fifos are the problem in your case. They will be overwritten
> when a new ones are created, as you run FvwmCommand from menu.
Hm. As someone else pointed out, I'll try putting in a line
Module FvwmCommandS
in my fvwm2rc independent of menus, InitFunction, RestartFunction.
Strange problem though since Bob is unable to reproduce it.
Regards,
Vikas
Bob Woodside
On 19-Aug-98 Richard Lister wrote:
>
> I seem to remember having trouble starting it from InitFunction. I now
>
> Module FvwmCommandS
>
Hmm.... That blows away my timing quirk theory - that some necessary
bit of setup wasn't yet done when the InitFunction was called. This command
will be executed *before* InitFunction is invoked.
So now this leaves me wondering if it's related to xdm, as Toshi
suggested.