[GRASS-user] mysterious error (WARNING: Cannot open old vector on level 2)

GRASS GRASS user
,
1

I'm having mysterious errors trying to use various modules in my
database (GRASS 6.1 running on Debian), which contains mapsets
PERMANENT and hajhouse.

example:
   GRASS 6.1.cvs (sacramento_sanjoaquin):~/grass > ps.map in=Scripts/psmap_script output=out.ps
   
   WARNING: Cannot open old vector soil@hajhouse on level 2
   ERROR: soil in hajhouse : can't open vector file

g.list vect reports:
   vector files available in mapset hajhouse:
   soil
   
   vector files available in mapset PERMANENT:
   cont_10 cty_bounds hydro plsa rd_local
   rd_state
   rd_usa soilmu_a_sac soilmu_l_sac soilmu_p_sac soilsa_a_sac
   soilsf_l_sac
   soilsf_p_sac urban

Trying to use v.out.ogr fails with the same error. Trying to display a
layer in QGIS (via grass plugin) fails with a rather unhelpful error
that the layer is not a valid layer. The error presented is the same
for soilmu_a_sac@PERMANENT and soil@hajhouse but some other layers in
this database work fine with all GRASS modules. All the layers
including the problematic ones display fine in the GRASS GIS manager,
however.

What is puzzling and frustrating about this is that it worked fine a
month or so ago. Then I set work in GRASS aside for a month and now
that I have resumed the errors are here. I have not altered my GRASS
binaries or support files in the meanwhile, and AFAIK my OS library
versions have not been updated either.

--
Henry House
+1 530 753 3361 ext. 13
Please don't send me HTML mail! My mail system frequently rejects it.
The unintelligible text that may follow is a digital signature.
See <http://hajhouse.org/pgp&gt; to find out how to use it.
My OpenPGP key: <http://hajhouse.org/hajhouse.asc&gt;\.

2

Henry House wrote:

I'm having mysterious errors trying to use various modules in my
database (GRASS 6.1 running on Debian), which contains mapsets
PERMANENT and hajhouse.

example:
   GRASS 6.1.cvs (sacramento_sanjoaquin):~/grass > ps.map
   in=Scripts/psmap_script output=out.ps
   
   WARNING: Cannot open old vector soil@hajhouse on level 2
   ERROR: soil in hajhouse : can't open vector file

g.list vect reports:
   vector files available in mapset hajhouse:
   soil
   
   vector files available in mapset PERMANENT:
   cont_10 cty_bounds hydro plsa rd_local
   rd_state
   rd_usa soilmu_a_sac soilmu_l_sac soilmu_p_sac soilsa_a_sac
   soilsf_l_sac
   soilsf_p_sac urban

Trying to use v.out.ogr fails with the same error. Trying to display a
layer in QGIS (via grass plugin) fails with a rather unhelpful error
that the layer is not a valid layer. The error presented is the same
for soilmu_a_sac@PERMANENT and soil@hajhouse but some other layers in
this database work fine with all GRASS modules. All the layers
including the problematic ones display fine in the GRASS GIS manager,
however.

What is puzzling and frustrating about this is that it worked fine a
month or so ago. Then I set work in GRASS aside for a month and now
that I have resumed the errors are here. I have not altered my GRASS
binaries or support files in the meanwhile, and AFAIK my OS library
versions have not been updated either.

rebuilt topology with:

v.build soils
  or
v.build.all # to rebuild all vector maps in current mapset

then it should work again.

(unless files are totally corrupt on the disk)

Hamish

3

On Wednesday 22 November 2006 17:53, Hamish wrote:

Henry House wrote:
> I'm having mysterious errors trying to use various modules in my
> database (GRASS 6.1 running on Debian), which contains mapsets
> PERMANENT and hajhouse.
>
> example:
> GRASS 6.1.cvs (sacramento_sanjoaquin):~/grass > ps.map
> in=Scripts/psmap_script output=out.ps
>
> WARNING: Cannot open old vector soil@hajhouse on level 2
> ERROR: soil in hajhouse : can't open vector file
>
> g.list vect reports:
> vector files available in mapset hajhouse:
> soil
>
> vector files available in mapset PERMANENT:
> cont_10 cty_bounds hydro plsa rd_local
> rd_state
> rd_usa soilmu_a_sac soilmu_l_sac soilmu_p_sac soilsa_a_sac
> soilsf_l_sac
> soilsf_p_sac urban
>
> Trying to use v.out.ogr fails with the same error. Trying to display a
> layer in QGIS (via grass plugin) fails with a rather unhelpful error
> that the layer is not a valid layer. The error presented is the same
> for soilmu_a_sac@PERMANENT and soil@hajhouse but some other layers in
> this database work fine with all GRASS modules. All the layers
> including the problematic ones display fine in the GRASS GIS manager,
> however.
>
> What is puzzling and frustrating about this is that it worked fine a
> month or so ago. Then I set work in GRASS aside for a month and now
> that I have resumed the errors are here. I have not altered my GRASS
> binaries or support files in the meanwhile, and AFAIK my OS library
> versions have not been updated either.

rebuilt topology with:

v.build soils
  or
v.build.all # to rebuild all vector maps in current mapset

then it should work again.

(unless files are totally corrupt on the disk)

Hamish

_______________________________________________
grassuser mailing list
grassuser@grass.itc.it
http://grass.itc.it/mailman/listinfo/grassuser

Hi Henry!

I have encountered this type of error when digitizing GRASS vectors in QGIS-
and something bizzare happens. Normally re-building the topology with v.build
solved the problem.

Another question- are you working with the latest NRCS SSURGO database? If you
would like some notes on importing all of the tabular data into postgres let
me know and I can send my notes.

Cheers,

--
Dylan Beaudette
Soils and Biogeochemistry Graduate Group
University of California at Davis
530.754.7341

4

Dylan Beaudette napisał(a):

On Wednesday 22 November 2006 17:53, Hamish wrote:
  

Henry House wrote:
    

I'm having mysterious errors trying to use various modules in my
database (GRASS 6.1 running on Debian), which contains mapsets
PERMANENT and hajhouse.

example:
   GRASS 6.1.cvs (sacramento_sanjoaquin):~/grass > ps.map
   in=Scripts/psmap_script output=out.ps

   WARNING: Cannot open old vector soil@hajhouse on level 2
   ERROR: soil in hajhouse : can't open vector file

g.list vect reports:
   vector files available in mapset hajhouse:
   soil

   vector files available in mapset PERMANENT:
   cont_10 cty_bounds hydro plsa rd_local
   rd_state
   rd_usa soilmu_a_sac soilmu_l_sac soilmu_p_sac soilsa_a_sac
   soilsf_l_sac
   soilsf_p_sac urban

Trying to use v.out.ogr fails with the same error. Trying to display a
layer in QGIS (via grass plugin) fails with a rather unhelpful error
that the layer is not a valid layer. The error presented is the same
for soilmu_a_sac@PERMANENT and soil@hajhouse but some other layers in
this database work fine with all GRASS modules. All the layers
including the problematic ones display fine in the GRASS GIS manager,
however.

What is puzzling and frustrating about this is that it worked fine a
month or so ago. Then I set work in GRASS aside for a month and now
that I have resumed the errors are here. I have not altered my GRASS
binaries or support files in the meanwhile, and AFAIK my OS library
versions have not been updated either.
      

rebuilt topology with:

v.build soils
  or
v.build.all # to rebuild all vector maps in current mapset

then it should work again.

(unless files are totally corrupt on the disk)

Hamish

_______________________________________________
grassuser mailing list
grassuser@grass.itc.it
http://grass.itc.it/mailman/listinfo/grassuser
    
Hi Henry!

I have encountered this type of error when digitizing GRASS vectors in QGIS- and something bizzare happens. Normally re-building the topology with v.build solved the problem.

Another question- are you working with the latest NRCS SSURGO database? If you would like some notes on importing all of the tabular data into postgres let me know and I can send my notes.

Cheers,

1. this error also heppens when you create file and leave it empty - it is normal - no topology can be build
2. I also had a problem with vectors after editing with qgis (file totally corupted v.build not helped). So I resign to use qgis as a editor for grass vector. I see, more people has the same problem

5

[...]

rebuilt topology with:

v.build soils
  or
v.build.all # to rebuild all vector maps in current mapset

then it should work again.

(unless files are totally corrupt on the disk)

Thanks, that works like a charm!

--
Henry House
+1 530 753 3361 ext. 13
Please don't send me HTML mail! My mail system frequently rejects it.
The unintelligible text that may follow is a digital signature.
See <http://hajhouse.org/pgp&gt; to find out how to use it.
My OpenPGP key: <http://hajhouse.org/hajhouse.asc&gt;\.

6

Jaros³aw Jasiewicz wrote:

Dylan Beaudette napisa³(a):

I have encountered this type of error when digitizing GRASS vectors in
QGIS- and something bizzare happens. Normally re-building the topology
with v.build solved the problem.

1. this error also heppens when you create file and leave it empty - it
is normal - no topology can be build
2. I also had a problem with vectors after editing with qgis (file
totally corupted v.build not helped). So I resign to use qgis as a
editor for grass vector. I see, more people has the same problem

Well, I for one never had such a problem as Jarek and Dylan describe -
QGIS never corrupted any GRASS vector for me. Even if QGIS crashes
during digitizing, v.build always helps, and QGIS doesn't crash that
easily these days; v.digit freezes or crashes for me more often AAMOF.

Honestly, QGIS 0.8 provides a really good vector digitizing tool for
GRASS. Using it is more convenient and time efficient than using
v.digit, and it doesn't suffer it's numerous problems (has it's own
problems though, yet less harmfull that those in v.digit; and they are
being fixed).

If you found bugs in QGIS GRASS digitizer please report them. There is
a major bug squashing ongoing in QGIS currently, your input could help!

Cheers,
Maciek

7

Maciej Sieczka napisa³(a):

Jaros³aw Jasiewicz wrote:

Dylan Beaudette napisa³(a):
   

I have encountered this type of error when digitizing GRASS vectors in
QGIS- and something bizzare happens. Normally re-building the topology
with v.build solved the problem.
     
1. this error also heppens when you create file and leave it empty - it
is normal - no topology can be build
2. I also had a problem with vectors after editing with qgis (file
totally corupted v.build not helped). So I resign to use qgis as a
editor for grass vector. I see, more people has the same problem
   
Well, I for one never had such a problem as Jarek and Dylan describe -
QGIS never corrupted any GRASS vector for me. Even if QGIS crashes
during digitizing, v.build always helps, and QGIS doesn't crash that
easily these days; v.digit freezes or crashes for me more often AAMOF.

Honestly, QGIS 0.8 provides a really good vector digitizing tool for
GRASS. Using it is more convenient and time efficient than using
v.digit, and it doesn't suffer it's numerous problems (has it's own
problems though, yet less harmfull that those in v.digit; and they are
being fixed).

If you found bugs in QGIS GRASS digitizer please report them. There is
a major bug squashing ongoing in QGIS currently, your input could help!

Cheers,
Maciek

Hi
Its about QGIS 0.8 preview 2 (TITAN) installed from deb packages
Well, most of bugs I saw with qgis happens during workshops with students, so that errors shuldn't be mentioned enywhere. In fact I have almost 20 students in workshop group so I cannot control everyone, so cannot reproduce errors respectively.
Most of them (during digitalisation) are simple run-time errors.
Addationally qgis is the only program I know which can crash linux totaly without particular reasons, (only hard reset can help)

I check qgis news regularry, And all errors I had been mentioned, so I didn't see reasons to mention them again.

Addationally on my own computer qgis works extremly slow. I work on digitzing soil map of Great Poland Lowland (over 2000 areas) And opposite Maciek I never had problems with v.digit (in spite of lack of experience) but qgis became solower and slower when number of obiects incresed.
I thing that one of reasons of my problems is that I install it form packages - but it shouldn't be an argument. And second is that I use gnome (GTK), but qgis is based on Qt (KDE) it also shouldn't be a problem.

It not a criticism - both programs (grass and qgis) are open source so I use them as they came And I happy they are - but in my private experience v.digit is much faster and stable than qgis

regards
Jarek

8

Jaros³aw Jasiewicz wrote:

Jaroslaw,

Its about QGIS 0.8 preview 2 (TITAN) installed from deb packages
Well, most of bugs I saw with qgis happens during workshops with
students, so that errors shuldn't be mentioned enywhere. In fact I have
almost 20 students in workshop group so I cannot control everyone, so
cannot reproduce errors respectively.

I see.

Most of them (during digitalisation) are simple run-time errors.
Addationally qgis is the only program I know which can crash linux
totaly without particular reasons, (only hard reset can help)

There's got to be a reason. Can you elaborate? (we should consider
moving the thread to QGIS user list then).

I check qgis news regularry, And all errors I had been mentioned, so I
didn't see reasons to mention them again.

Did you report them in their bugtracker, in a way they can make some
use of the report?

Addationally on my own computer qgis works extremly slow. I work on
digitzing soil map of Great Poland Lowland (over 2000 areas)

That's not really big. I've been editing in QGIS a GRASS vector a map
of similar size *daily*, for few last months, without any slowdown;
here's the vectors details:

$ v.info -t parcels_15_2zl
nodes=3097
points=0
lines=0
boundaries=3091
centroids=1175
areas=1175
islands=6
faces=0
kernels=0
primitives=4266
map3d=0

Can you specify what is the nature of the "qgis works extremly slow"?
Is that the vectors are rendered slowly (and under what circumstances),
or the backdrop raster, else?

And opposite Maciek I never had problems with v.digit (in spite of lack of
experience) but qgis became solower and slower when number of obiects
incresed.

I've had a major problem with the speed of rendering vectors in QGIS
too (if that's what you mean), but it showed to be due to X server
issues, *maybe* limited to the distro I'm using (Ubuntu 6.06 Dapper).
After fixing my Xorg setup the problem vanished; please see the thread
on QGIS user list [1] and in the QGIS bugtracker [2] (the 2 last
messages). Please let me know whether this is the same issue.

I thing that one of reasons of my problems is that I install it form
packages - but it shouldn't be an argument. And second is that I use
gnome (GTK), but qgis is based on Qt (KDE) it also shouldn't be a problem.

It not a criticism

But yes it is. And it's a very good thing, if you can present
argument's supporting your criticism, so that others can benefit from
it (the users - be informed, the devs - be able to fix the problem).
However, if you can't provide details and arguments, you can't have a
clue what is the possible reason of the issue, and you can't be sure it
is in QGIS, can you? I made this mistake in case of the issue with
vectors rendering speed in QGIS myself :).

Maciek

[1]http://www.nabble.com/rendering-current-line-in-Grass-digitizer-is-very-slow-tf1908709.html#a6825237
[2]https://svn.qgis.org/trac/ticket/295