[GRASS5] fifo problems opening X0 display (CVS compile) (fwd)

[Tim Sutton]

Hi

Just a quick note to say thank you for your help. I had indeed checked
out the dev branch and not the release branch.
As recommended, I did:

cvs update -r releasebranch_11_april_2001_5_0_0

to check out the correct branch of cvs, then uncommented:

USE_G_SOCKS defined in src/include/config.h

Then did a

make uninstall && make clean

To clean up my previous botched installation, and finally

./configure --with-socket
make
su - root
make install

Everything works fine now. I am pleased to see the shapefile importer
seems to work quite now - I never had much luck with it before.

Regards

Tim Sutton


On Wed, 12 Sep 2001, Glynn Clements wrote:

>
> Tim Sutton wrote:
>
> > My apologies if this is ground ground that has been covered already. I
> > have just compiled grass CVS from source (checked out Monday night).
>
> Did you check out the release branch (releasebranch_11_april_2001_5_0_0 tag)
> or the head (no tag)? I don't know whether the head actually works at
> present, so if you have that checked out, get the release branch instead
> with "cvs update -r releasebranch_11_april_2001_5_0_0".
>
> > The
> > compilation process seems to have gone fine, and I can start grass and
> > tcltkgrass no problem. However opening a display fails:
> >
> > Usage: x0 [-] "input_fifo output_fifo" [nlev]
> > Please start graphics monitor <x0>
>
> The text of the error message indicates that the monitor was compiled
> for FIFOs, but the fact that the error message is printed suggests
> that mon.start was compiled for sockets.
>
> If you are using the release branch, my first suspicion would be that an
> incomplete version of GRASS was installed over an existing version which
> used a different
>
> Was this the first time you installed GRASS, or was there another
> version already present? Did you get any build errors in error.log?
> Is USE_G_SOCKS defined in src/include/config.h?
>
>