[GRASS-user] Is this a bug in 8.3.dev?

hellik · May 3, 2023, 7:15pm

When I used d.mon start=wx0 grass told me it could not find a numpy for
python-2.7.x.

why not, as Anna already suggested in an earlier reply, _fix_ your operating system by removing python 2.x from your linux box and keep only python 3 there.

for a long time now, GRASS GIS 8.x switched to python 3.

kind regards
Helmut

rshepard · May 3, 2023, 7:40pm

On Wed, 3 May 2023, Helmut Kudrnovsky wrote:

why not, as Anna already suggested in an earlier reply, _fix_ your
operating system by removing python 2.x from your linux box and keep only
python 3 there.

for a long time now, GRASS GIS 8.x switched to python 3.

Helmut,

I've no idea whether any other applications still require python2.
Regardless, I removed python2 from the laptop. Now GRASS_PYTHON=python3.9.

When I restart grass there's still a monitor issue. When I enter `d.mon
start=wx0` grass tells me it's already running. When I enter `d.mon
stop=wx0` grass tells me there is no wx0 monitor running.

How do I correct this?

Thanks,

Rich

annakrat · May 4, 2023, 2:37pm

Rich, it would be extremely helpful, if you always show the errors you are getting, otherwise, it’s very difficult for us to help you.

On Wed, May 3, 2023 at 3:40 PM Rich Shepard <rshepard@appl-ecosys.com> wrote:

On Wed, 3 May 2023, Helmut Kudrnovsky wrote:

why not, as Anna already suggested in an earlier reply, fix your
operating system by removing python 2.x from your linux box and keep only
python 3 there.

for a long time now, GRASS GIS 8.x switched to python 3.

Helmut,

I’ve no idea whether any other applications still require python2.
Regardless, I removed python2 from the laptop. Now GRASS_PYTHON=python3.9.

When I restart grass there’s still a monitor issue. When I enter d.mon start=wx0 grass tells me it’s already running. When I enter d.mon stop=wx0 grass tells me there is no wx0 monitor running.

How do I correct this?

Thanks,

Rich

grass-user mailing list
grass-user@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/grass-user

rshepard · May 4, 2023, 3:19pm

On Thu, 4 May 2023, Anna Petrášová wrote:

Rich, it would be extremely helpful, if you always show the errors you are
getting, otherwise, it's very difficult for us to help you.

Anna,

My apologies; I do know better.

Here's the session:

GRASS metro100/PERMANENT:~ > g.list type=vect
met100
GRASS metro100/PERMANENT:~ > d.mon start=wx0
ERROR: Monitor <wx0> already running
GRASS metro100/PERMANENT:~ > d.mon stop=wx0
WARNING: Unable to open file
</data1/grassdata/metro100/PERMANENT/.tmp/caddis.appl-ecosys.com/MONITORS/wx0/pid>
GRASS metro100/PERMANENT:~ >

When I look for that file the MONITORS subdirectory does not exist:

[rshepard@caddis ~]$ cd /data1/grassdata/metro100/PERMANENT/.tmp/caddis.appl-ecosys.com/
[rshepard@caddis caddis.appl-ecosys.com]$ ls
[rshepard@caddis caddis.appl-ecosys.com]$

Re-install grass?

Rich

neteler · May 4, 2023, 7:28pm

On Thu, May 4, 2023 at 5:19 PM Rich Shepard <rshepard@appl-ecosys.com> wrote:

On Thu, 4 May 2023, Anna Petrášová wrote:

> Rich, it would be extremely helpful, if you always show the errors you are
> getting, otherwise, it's very difficult for us to help you.

Anna,

My apologies; I do know better.

Here's the session:

GRASS metro100/PERMANENT:~ > g.list type=vect
met100
GRASS metro100/PERMANENT:~ > d.mon start=wx0
ERROR: Monitor <wx0> already running
GRASS metro100/PERMANENT:~ > d.mon stop=wx0
WARNING: Unable to open file
</data1/grassdata/metro100/PERMANENT/.tmp/caddis.appl-ecosys.com/MONITORS/wx0/pid>
GRASS metro100/PERMANENT:~ >

When I look for that file the MONITORS subdirectory does not exist:

[rshepard@caddis ~]$ cd /data1/grassdata/metro100/PERMANENT/.tmp/caddis.appl-ecosys.com/
[rshepard@caddis caddis.appl-ecosys.com]$ ls

Are these directories writable?

ls -la /data1/grassdata/metro100/PERMANENT/.tmp
and
ls -la /data1/grassdata/metro100/PERMANENT/.tmp/caddis.appl-ecosys.com/

will tell.

[rshepard@caddis caddis.appl-ecosys.com]$

Re-install grass?

Not now, this will not necessarily help.
There is an easy debugging option integrated:

# switch on
g.gisenv set="DEBUG=1"
d.mon wx0
--> what are the messages?

# switch off
g.gisenv set="DEBUG=0"

Markus

rshepard · May 4, 2023, 8:25pm

On Thu, 4 May 2023, Markus Neteler wrote:

Are these directories writable?

Markus,

Yes, they are.

Not now, this will not necessarily help.
There is an easy debugging option integrated:

# switch on
g.gisenv set="DEBUG=1"
d.mon wx0

Very interesting:

GRASS NSR_TaxLot/PERMANENT:~ > g.gisenv set="DEBUG=1"
D1/1: G_set_program_name(): g.gisenv
GRASS NSR_TaxLot/PERMANENT:~ > d.mon wx0
D1/1: G_set_program_name(): d.mon
D1/1: Monitor width/height = 720/480
D1/1: Monitor name=wx0, pyfile = /data1/grassdata/NSR_TaxLot/PERMANENT/.tmp/caddis.appl-ecosys.com/MONITORS/wx0/render.py
D1/1: Monitor name=wx0, envfile=/data1/grassdata/NSR_TaxLot/PERMANENT/.tmp/caddis.appl-ecosys.com/MONITORS/wx0/env
D1/1: Monitor name=wx0, cmdfile = /data1/grassdata/NSR_TaxLot/PERMANENT/.tmp/caddis.appl-ecosys.com/MONITORS/wx0/cmd
D1/1: G_set_program_name(): g.gisenv
GRASS NSR_TaxLot/PERMANENT:~ > D1/1: G_set_program_name(): g.gisenv
D1/1: grass.script.core.start_command(): g.gisenv -n
D1/1: G_set_program_name(): g.gisenv
wxnviz.py: This module requires the NumPy module, which could not be imported. It probably is not installed (it's not part of the standard Python distribution). See the Numeric Python site (http://numpy.scipy.org) for information on downloading source or binaries.
D1/1: grass.script.core.start_command(): g.message -v message=Starting map display <wx0>...
D1/1: G_set_program_name(): g.message
D1/1: grass.script.core.start_command(): g.gisenv set=MONITOR_WX0_PID=27366
D1/1: grass.script.core.start_command(): g.gisenv -n
D1/1: G_set_program_name(): g.gisenv
D1/1: grass.script.core.start_command(): g.proj -p
D1/1: grass.script.core.start_command(): g.region -ugpc

D1/1: G_set_program_name(): g.gisenv
GRASS NSR_TaxLot/PERMANENT:~ >

and the GUI window displayed.

# switch off
g.gisenv set="DEBUG=0"

Why are lines repeated?

And, python3-numpy-1.22.3-x86_64-1.SBo is installed.

Thanks,

Rich

neteler · May 5, 2023, 6:48am

Rich,

On Thu, May 4, 2023 at 10:25 PM Rich Shepard <rshepard@appl-ecosys.com> wrote:

On Thu, 4 May 2023, Markus Neteler wrote:

...

> There is an easy debugging option integrated:
>
> # switch on
> g.gisenv set="DEBUG=1"
> d.mon wx0

Very interesting:

GRASS NSR_TaxLot/PERMANENT:~ > g.gisenv set="DEBUG=1"
D1/1: G_set_program_name(): g.gisenv

...

D1/1: G_set_program_name(): g.gisenv
wxnviz.py: This module requires the NumPy module, which could not be imported. It probably is not installed (it's not part of the standard Python distribution). See the Numeric Python site (http://numpy.scipy.org) for information on downloading source or binaries.

Still your numpy installation appears to be broken.
Perhaps python3-numpy-1.22.3-x86_64-1.SBo does not match your python
version at binary level?

...

GRASS NSR_TaxLot/PERMANENT:~ >

and the GUI window displayed.

Good.

> # switch off
> g.gisenv set="DEBUG=0"

Why are lines repeated?

Perhaps there are loops in the code. It is harmless and just debugging
output which can be a bit noisy.

Markus

rshepard · May 5, 2023, 1:54pm

On Fri, 5 May 2023, Markus Neteler wrote:

Still your numpy installation appears to be broken. Perhaps
python3-numpy-1.22.3-x86_64-1.SBo does not match your python version at
binary level?

Markus,

I'm rebuilding python3-numpy and will re-install it. It's not likely to be a
mis-match between the Slackware cord python3 and the SBo numpy because it's
well tested before being added to the repo.

I also have openBLAS installed; it's an optimized BLAS and includes LAPACK.

If the re-installation of numpy doesn't fix the problem I build and install
the 8.2.2 source which will replace the development code installed. I must
get hydrological analyses done by Monday so I don't really care which grass
version I run.

Thanks,

Rich

rshepard · May 5, 2023, 2:06pm

On Fri, 5 May 2023, Rich Shepard wrote:

If the re-installation of numpy doesn't fix the problem I build and
install the 8.2.2 source which will replace the development code
installed. I must get hydrological analyses done by Monday so I don't
really care which grass version I run.

Markus, et al.:

To my pleasant surprise, re-building/re-installing python3-numpy-1.22.3
fixed the monitor issue. The only reason that comes to mind is that the
first install was in parallel with python2 and python2-numpy and grass was
still looking for the latter even though it was gond and the laptop
rebooted.

I've run Slackware for 20 years now and haven't had this issue with any
SlackBuilds.org packages before. Regardless I can now meet client
expectations.

Thanks for pointing me in the right direction.

I have a question about a multi-layer FileGDB where each layer has a
different CRS and I'll aske that on gitter.

Regards,

Rich