[GRASS5] Let's release 5.4 asap

GRASS GRASS developer
41

On Sun, Oct 10, 2004 at 08:37:57PM +0200, Markus Neteler wrote:

On Fri, Oct 08, 2004 at 12:56:07PM +0200, Bernhard Reiter wrote:
[...]
> I wrote that a few mails ago and it was meant against "deleting 5.0"
> from the servers. (That was a missunderstanding that got clarified,
> nobody proposed to delete it.)

[also related to your latest message, Bernhard]

Just to clarify that again: Radim and me have spent much time to
find (!) all the versions. We still try to get hands on the
CERL-4.2 which was never published. Jim Westervelt is trying to
find it for me (us). So, just by entering the
web site through ftp you can see all the versions in subdirectories.

Yes, you did not delete them as I wrote above.
(clarified the third time now.) :wink:

Perhaps you didn't see this page?
http://grass.itc.it/grass_releases.html
Not sure what else could be done.
You can click on all versions and grab the source code.

I did see that page and in comparision to the table
that was there before it basically got "hidden".
What can you do:
Make a direct link to the ftp directory (ftp://grass.itc.it/pub/grass/)
which has all the versions, or better:
Have page only listing the old source tarballs in a table
clearly marking them as source tarballs.
Even better would to to improve the structure to have an
directory only containing all the sources, e.g. much like:
ftp://ftp.kde.org/pub/kde/stable/

If you want to see detailed changes (diffs), go here:
http://mpa.itc.it/radim/g50history/

Radim has spent much time to make it convenient HTML pages.
He even fetched all related emails:
http://mpa.itc.it/radim/g50history/grass50history.html

Yes, very important work.

However, to offer many different development lines is impossible
unless more developers join the team. But this has been discussed
often enough.

And I tend to disagree because keeping them available and clearly marking
their status is a matter of marking and keeping the revisions,
not of more effort. Then each person can choose what version to try.
And old one or a different one.
Of course I agree that we need more contributors
and until they come, it might well be that a stable version
is only stable in their bugs. :wink:

  Bernhard

42

> The motivation is that v.in.sites currently doesn't make a table

Realy?

Nope.

I could have sworn last time I tried this it didn't work.
Oh well, never mind.

1. How do I create a new column in a table within a dbf file? Is it
possible through GRASS? Or should I just use a seperate tool?

echo "alter table xxx add column yyy double precision" | db.execute

I see that is already in the db.execute help page;

Is it possible to rename a column after it is created?

Should we have a DB howto page in addition to the existing SQL howto
help page?

Re. broken man page generation; I see that most of the database
commands are breaking the script (if that helps the debugging):
  http://grass.ibiblio.org/grass57/manuals/html57_user/database.html

Hamish

43

On Tuesday 12 October 2004 03:57, Hamish wrote:

> echo "alter table xxx add column yyy double precision" | db.execute

I see that is already in the db.execute help page;

Is it possible to rename a column after it is created?

No.

Should we have a DB howto page in addition to the existing SQL howto
help page?

? http://grass.itc.it/grass57/manuals/html57_user/sql.html
It is not SQL but DBF driver howto.

Radim

44

Hamish,

Where would I find free OCR software?

Stephen

----- Original Message -----
From: "Hamish" <hamish_nospam@yahoo.com>
To: "Stephen Clark" <stephen.clark@focus.ca>
Cc: <grass5@grass.itc.it>
Sent: Friday, October 01, 2004 9:15 PM
Subject: Re: [GRASS5] Is there a raster to vector converter for text like
there is for lines?

> I am wondering if there is a raster to vector converter for text like
> there is for lines?

r.out.tiff and generic OCR software?
what exactly are you trying to do?

Hamish

_______________________________________________
grass5 mailing list
grass5@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass5

45

Where would I find free OCR software?

$ apt-cache search ocr

clara - Free OCR program for Unix Systems
gocr - A command line OCR
gocr-doc - gocr documentation
gocr-gtk - A GTK wrapper around gocr
gocr-tk - A tcl/tk wrapper around gocr
gstreamer-misc - Collection of various GStreamer plugins
gstreamer0.8-misc - Collection of various GStreamer plugins
kaudiocreator - CD ripper and audio encoder frontend
ksocrat - English/Russian and Russian/English Dictionary
ksocrat-data - English and Russian KSocrat data files
libgocr-dev - API set to write your own OCR engine - development files
libgocr-doc - API set to write your own OCR engine - documentation
libgocr0 - API set to write your own OCR engine - runtime libs
nec - NEC2 Antenna Modelling System
quiteinsane - A Qt based X11 frontend for SANE (Scanner Access Now Easy)
ksubtitleripper - GUI for KDE to rip DVD subtitles

that should be enough to feed into google.

Hamish

46

I noticed that s.kcv is commented out with a note that it needs update
to 5.0 API -
is that true? Markus says that it compiles (and runs?) OK for him.
I have troubles compiling it because of bad path, but otherwise it may
be
up-to-date and worth including - can somebody else try it?.
It creates a random subset of point data -
this is particularly useful for crossvalidation, as well as for data
sets
that are oversampled (e.g. lidar).
Radim, would it be possible to port it to 5.7 as v.kcv? (I noticed that
some
of the other useful site programs such as v.random are already there.

thank you for any feedback on this,

helena

         [GRASS-CVS] paul: grass/src/CMD/lists GRASS,1.132,1.133
....
#src/sites/s.kcv # needs update to 5.0 API
....

Markus Neteler wrote:

> I tried to compile it myself but I am getting a similar error as with the
> g3. command in 5.7 - it tries to write into /etc/bin which does not exist (see below)

That's odd. I get:

make[1]: Leaving directory `/hardmnt/thuille1/ssi/cvsgrass_exp/src/sites/s.kcv'

So it writes (as it should) to /nfsmnt/thuille1/ssi/cvsgrass_exp/grass-build/dist.i686-pc-linux-gnu/etc/bin/cmd/s.kcv

I feel that the part /path/to/dist.i686-pc-linux-gnu/ is missing in your case.

> gcc
> -L/home/helena/grasscvs53/grass50_exp_2003_10_04/src/libes/LIB.i686-pc-linux-gnu -o /etc/bin/cmd/s.kcv OBJ.i686-pc-linux-gnu/main.o OBJ.i686-pc-linux-gnu/dreset.o OBJ.i686-pc-linux-gnu/histo.o OBJ.i686-pc-linux-gnu/utils.o -lgis -ldatetime -lm -lz
> /usr/bin/ld: cannot open output file /etc/bin/cmd/s.kcv: No such file or
> directory
> collect2: ld returned 1 exit status
> make: *** [/etc/bin/cmd/s.kcv] Error 1

Somehow the path prefix got lost.

Markus

47

Helena,

It doesn't work for x|y|%z format - I got a segmentation fault. Maybe it works for other site types.

Jaro

BUGS:

* not fully updated to 5.0 sites API
  -> fixed MN 2/2001

* the modules crashes in
   if (!z[m].partition)
  as this is not set properly

Helena Mitasova wrote:

I noticed that s.kcv is commented out with a note that it needs update
to 5.0 API -
is that true? Markus says that it compiles (and runs?) OK for him.
I have troubles compiling it because of bad path, but otherwise it may
be
up-to-date and worth including - can somebody else try it?. It creates a random subset of point data -
this is particularly useful for crossvalidation, as well as for data
sets
that are oversampled (e.g. lidar).
Radim, would it be possible to port it to 5.7 as v.kcv? (I noticed that
some
of the other useful site programs such as v.random are already there.

thank you for any feedback on this,

helena

          [GRASS-CVS] paul: grass/src/CMD/lists GRASS,1.132,1.133
....
#src/sites/s.kcv # needs update to 5.0 API
....

Markus Neteler wrote:

I tried to compile it myself but I am getting a similar error as with the
g3. command in 5.7 - it tries to write into /etc/bin which does not exist (see below)

That's odd. I get:

make[1]: Leaving directory `/hardmnt/thuille1/ssi/cvsgrass_exp/src/sites/s.kcv'

So it writes (as it should) to /nfsmnt/thuille1/ssi/cvsgrass_exp/grass-build/dist.i686-pc-linux-gnu/etc/bin/cmd/s.kcv

I feel that the part /path/to/dist.i686-pc-linux-gnu/ is missing in your case.

gcc
-L/home/helena/grasscvs53/grass50_exp_2003_10_04/src/libes/LIB.i686-pc-linux-gnu -o /etc/bin/cmd/s.kcv OBJ.i686-pc-linux-gnu/main.o OBJ.i686-pc-linux-gnu/dreset.o OBJ.i686-pc-linux-gnu/histo.o OBJ.i686-pc-linux-gnu/utils.o -lgis -ldatetime -lm -lz
/usr/bin/ld: cannot open output file /etc/bin/cmd/s.kcv: No such file or
directory
collect2: ld returned 1 exit status
make: *** [/etc/bin/cmd/s.kcv] Error 1

Somehow the path prefix got lost.

Markus

_______________________________________________
grass5 mailing list
grass5@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass5

-=x=-
Skontrolované antivírovým programom NOD32

48

On Fri, 15 Oct 2004, Jaro Hofierka wrote:

Helena,

It doesn't work for x|y|%z format - I got a segmentation fault. Maybe it works for other site types.

Perhaps we can enable it for 5.4.0 and then fix the bug later in 5.4.1. Are there many other modules that are disabled for small reasons like this? I think the bugs are more likely to be fixed if the modules are compiled and in use and people are complaining about them.

Paul

49

On Friday 15 October 2004 21:46, Helena Mitasova wrote:

Radim, would it be possible to port it to 5.7 as v.kcv? (I noticed that
some
of the other useful site programs such as v.random are already there.

I'll look at it, i have recently ported more s.* modules.

Radim

50

It doesn't work for x|y|%z format - I got a segmentation fault. Maybe
it works for other site types.

main.c line 215:

      /* now sort these distances */
      qsort (d, k, sizeof (D), dcmp);
      z[d[0].i].z = i + 1;
      outSite->east = z[d[0].i].x;
      outSite->north = z[d[0].i].y;
      outSite->ccat = z[d[0].i].z;
      G_site_put (fdtest, outSite);

but d[0].i is -1, which is not a valid array element.

Hamish

51

On Friday 15 October 2004 21:46, Helena Mitasova wrote:

Radim, would it be possible to port it to 5.7 as v.kcv? (I noticed that
some
of the other useful site programs such as v.random are already there.

I have submitted v.kcv; random partition is written to a new column
(table must exist, no check, should be added).
It must be terribly slow for larger data sets because of the algorithm
used (from s.kcv): random point -> nearest point, i.e. O(n^2)

Radim

52

On Mon, 18 Oct 2004, Radim Blazek wrote:

On Friday 15 October 2004 21:46, Helena Mitasova wrote:
> Radim, would it be possible to port it to 5.7 as v.kcv? (I noticed that
> some
> of the other useful site programs such as v.random are already there.

I have submitted v.kcv; random partition is written to a new column
(table must exist, no check, should be added).
It must be terribly slow for larger data sets because of the algorithm
used (from s.kcv): random point -> nearest point, i.e. O(n^2)

This was what I noticed in July 2001 replying (off-line) to Anne
Holopainen (http://grass.itc.it/pipermail/grass5/2001-July/002801.html):

"I've been reading the s.kcv code and manual page. As I understand it, it
takes two arguments, k, the number of test/trials required, and sites, the
complete list of sites available, numbering say nsites in total. s.kcv
then creates k partitions, some ceiling(nsites/k) some floor(nsites/k)
large, such that the sum of the counts of members of the k test sets is
nsites. No site appears in more than one test set. The trial sets seem to
contain all the sites that are not test set members, such that the sum of
numbers of members of trial sets and test sets is always nsites.

The test sites are sampled from the sites available without replacement
using a rather odd mechanism. A point is chosen in the window region,
easting and northing at random, and the nearest site not yet assigned to a
test set is deemed to be chosen to join the test set being constructed. I
don't know why this is done in this way. Darrell actually comments that
the spatial process should not necessarily be random in space - if the
points themselves are not. Further, the region maybe ought to be defined
other than the current window.

I don't see why space is brought into this - do you have a view? Anyway,
if space isn't important, sample() can be used.

s.kcv <- function(k, sites) {
        k <- as.integer(k)
        if(k < 2) stop("k less than two")
        if(!is.points(sites))
                stop ("sites not a splancs point object")
        nsites <- npts(sites)
        if(nsites < k) stop("More partitions than sites")
        idx <- 1:nsites
        p <- integer(k)
        p <- rep(floor(nsites/k), k)
        mod.p <- nsites %% k
        if (mod.p > 0) p[1:mod.p] <- ceiling(nsites/k)
        test <- vector("list", k)
        trial <- vector("list", k)
        for (i in 1:(k-1)) {
                test[[i]] <- sort(sample(idx, p[i],
                        replace=FALSE))
                trial[[i]] <- which(!(1:nsites %in% test[[i]]))
                idx <- idx[which(!(idx %in% test[[i]]))]
        }
        test[[k]] <- idx
        trial[[k]] <- which(!(1:nsites %in% idx))
        res <- list(k=k, sites=sites, p=p, test=test,
                trial=trial)
}"

As I wrote, I have no idea why Darrell sampled in space and chose the
nearest, really. It could be done with a quadtree structure, but why?

Roger

PS. Anne got her work done three years ago using the R bridge.

Radim

_______________________________________________
grass5 mailing list
grass5@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass5

--
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Breiviksveien 40, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93
e-mail: Roger.Bivand@nhh.no

53

Hi,

Has anyone tried to use OCR software on scanned maps to distinguish text
letters or words and then export these text values to a file that could be
used in a "spatial join" to join the polygons from a raster to vector
conversion and the text point files?

Anyone with some thoughts on how to do this would be welcome.

My initial thoughts are to "train" the CLARA software to recognize the text
values in my raster file and then find some way to export the individual
text entities to a shape file.

Also are there any research links on this idea?

thanks
Stephen

----- Original Message -----
From: "Hamish" <hamish_nospam@yahoo.com>
To: "Stephen Clark" <stephen.clark@focus.ca>
Cc: <grass5@grass.itc.it>
Sent: Tuesday, October 12, 2004 2:32 PM
Subject: Re: [GRASS5] Is there a raster to vector converter for text like
there is for lines?

> Where would I find free OCR software?

$ apt-cache search ocr

clara - Free OCR program for Unix Systems
gocr - A command line OCR
gocr-doc - gocr documentation
gocr-gtk - A GTK wrapper around gocr
gocr-tk - A tcl/tk wrapper around gocr
gstreamer-misc - Collection of various GStreamer plugins
gstreamer0.8-misc - Collection of various GStreamer plugins
kaudiocreator - CD ripper and audio encoder frontend
ksocrat - English/Russian and Russian/English Dictionary
ksocrat-data - English and Russian KSocrat data files
libgocr-dev - API set to write your own OCR engine - development files
libgocr-doc - API set to write your own OCR engine - documentation
libgocr0 - API set to write your own OCR engine - runtime libs
nec - NEC2 Antenna Modelling System
quiteinsane - A Qt based X11 frontend for SANE (Scanner Access Now Easy)
ksubtitleripper - GUI for KDE to rip DVD subtitles

that should be enough to feed into google.

Hamish

_______________________________________________
grass5 mailing list
grass5@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass5

54

Markus and Radim have send me a script for
moving the files in CVS from grass5 to grass51.

I plan to do the move tomorrow during the day.
For the operation I will switch off CVS access
for approximately 30 minutes.

So watch this mailinglist, because if you are working on grass51,
because you want to update your repositories after the move.

Afterwards the list of files can be used to use cvs admin command
to set a state or so, if you want to mark it.

On Sat, Oct 02, 2004 at 05:00:24PM +0200, Bernhard Reiter wrote:

We have no schedule yet.

I cannot judge how complicated the current copying
or mix linking is and how many changes are going to be made
to the files. So the timing must be up for somebody else to decide.

On Fri, Oct 01, 2004 at 04:00:43PM +0200, Radim Blazek wrote:
> On Friday 01 October 2004 15:25, Bernhard Reiter wrote:
> > > > I can move files or directories around in CVS in principle.
> > > > Just need a good list.
> > > > (probably the right mv commands might be most instructive :wink: )
> > > > and then I need to schedule shutting down CVS
> > > > and doing the change.
> > > > (Depending on my time this might a few days, so giving me an ahead
> > > > notice is appreceated.)

> > I would appreciate if someone would make a fixed list of "cp"
> > command lines that I only need to source with the ",v" extension,
> > because that seems more transparent as a script.
> > Especially must there be checking if there is a corresponding file already?

> > During a copy operation in the repository someone cannot put in CVS
> > comments, they would be exactly the same as in the original file.
> > Probably with cvs admin you can change something.
> > Maybe basiing revision numbers starting with "2.x" ?
> >
> > When do you want this to happen?

55

Bernhard Reiter wrote:

Markus and Radim have send me a script for moving the files in CVS from grass5 to grass51.

I plan to do the move tomorrow during the day.
For the operation I will switch off CVS access
for approximately 30 minutes.

So watch this mailinglist, because if you are working on grass51,
because you want to update your repositories after the move.

Afterwards the list of files can be used to use cvs admin command
to set a state or so, if you want to mark it.

It would be good to add a new log message for each file telling where it comes from. I don't see however any cvs command which can add a log
message without changes in code. cvs admin can only change existing message.

Radim

56

Radim Blazek wrote:

Bernhard Reiter wrote:

Markus and Radim have send me a script for moving the files in CVS from grass5 to grass51.

I plan to do the move tomorrow during the day.
For the operation I will switch off CVS access
for approximately 30 minutes.

So watch this mailinglist, because if you are working on grass51,
because you want to update your repositories after the move.

Afterwards the list of files can be used to use cvs admin command
to set a state or so, if you want to mark it.

It would be good to add a new log message for each file telling where it comes from. I don't see however any cvs command which can add a log
message without changes in code. cvs admin can only change existing message.

Sorry,

cvs commit -f -m

should do that. Bernhard, could you add these messages immediately
after you copy the files. Something like
-m "copy from grass/xxx/yyy/zzz.c"

Radim

57

On Fri, Nov 05, 2004 at 02:45:57PM +0100, Radim Blazek wrote:

Radim Blazek wrote:
>Bernhard Reiter wrote:

>>Afterwards the list of files can be used to use cvs admin command
>>to set a state or so, if you want to mark it.
>
>It would be good to add a new log message for each file telling where it
>comes from. I don't see however any cvs command which can add a log
>message without changes in code. cvs admin can only change existing
>message.

cvs commit -f -m

I do not have much experience with that command.
What do you think about boosting the revision to 2.0?
Like with cvs commit -r2.0
That would also be an additional clear sign.

should do that. Bernhard, could you add these messages immediately
after you copy the files. Something like
-m "copy from grass/xxx/yyy/zzz.c"

I will have to try, because the copying is in the repository
while the cvs command has to work on a working copy.
Right after copying I do not have a working copy.

58

On Fri, Nov 05, 2004 at 10:23:37AM +0100, Bernhard Reiter wrote:

Markus and Radim have send me a script for
moving the files in CVS from grass5 to grass51.

I plan to do the move tomorrow during the day.
For the operation I will switch off CVS access
for approximately 30 minutes.

So watch this mailinglist, because if you are working on grass51,
because you want to update your repositories after the move.

Due to a delay on my side the move did not happen saturday,
but will happen today.
I will keep the list posted.

59

On Sun, Nov 07, 2004 at 11:22:19AM +0100, Bernhard Reiter wrote:

On Fri, Nov 05, 2004 at 10:23:37AM +0100, Bernhard Reiter wrote:
> Markus and Radim have send me a script for
> moving the files in CVS from grass5 to grass51.
>
> I plan to do the move tomorrow during the day.
> For the operation I will switch off CVS access
> for approximately 30 minutes.
>
> So watch this mailinglist, because if you are working on grass51,
> because you want to update your repositories after the move.

Due to a delay on my side the move did not happen saturday,
but will happen today.
I will keep the list posted.

Arg. :frowning:
Again bad circumstances prevented me from doing the move today.
So I need to re-schedule to Tuesday.
Please take my appologies if this has caused trouble of any kind.

60

On Sun, Nov 07, 2004 at 10:59:36PM +0100, Bernhard Reiter wrote:

On Sun, Nov 07, 2004 at 11:22:19AM +0100, Bernhard Reiter wrote:
> On Fri, Nov 05, 2004 at 10:23:37AM +0100, Bernhard Reiter wrote:

> > For the operation I will switch off CVS access
> > for approximately 30 minutes.

Scheduled for 10:00 UTC today (20041109) now.
Planned downtime for CVS acess: 1 hour.