[GRASS-dev] please help review osgeo live dvd summary docs

Hi,

we're finalizing the next version of the OSGeo live sampler DVD,
USB stick, and Virtual Machine release. It's a live boot system
with all the latest goodies, including GRASS 6.4.2 and almost
all of the OSGeo family of desktop and server apps.
( http://live.osgeo.org )

It would be really helpful if others with fresh eyes could have
a look over the grass pages and double check that they're all
still relevant for v6.4.2, no dumb mistakes, you like what you
see, etc.

  http://adhoc.osgeo.osuosl.org/livedvd/docs/en/overview/grass_overview.html

  http://adhoc.osgeo.osuosl.org/livedvd/docs/en/quickstart/grass_quickstart.html

The finalized English version will go out for translation at the
end of this week. (conspicuously there are no Italian translations
yet even though there is a strong Italian FOSS4G presence, hint
hint :slight_smile:

There's a copy of the quickstart in the GRASS wiki. Note that
when I wrote it there was a GRASS workshop at the targeted FOSS4G
conference which was going to focus on vector processing, so I
didn't bother to put much of that in the quickstart. Contributions
demonstrating some fancy vector usage that makes GRASS special,
or a common vector task to show how GRASS does it easily, would
be appreciated. Just note that the quickstarts are intended to be
quick- i.e. 5-10 minutes worth, so we've got to keep it concise.

In addition, on the disc the QGIS-GRASS bridge is broken because
of this bug:
  http://hub.qgis.org/issues/2947
  https://trac.osgeo.org/osgeo/ticket/868

(QGIS doesn't take into account GRASS's mapset ownership rules,
so the plugin fails in a multi-user environment)

thanks for any assistance,
Hamish

Hamish,

the quickstart looks very nice and I did not find any major problems (but other may have a look too)

I just have couple comments:

- given that OSgeo live will be mostly used by new users should you even mention
the tcltk interface? It sounds a little bit confusing for people who don't know what TclTk is
and makes GRASS look overly complex. Instead, it would be useful to put here a link to
the wxPython GUI Help page so that people know where to find explanation of what you
are talking about later on (especially what each button on the layer manager means
- the verbal description may not be clear enough for some users).

Also, the note about the netbook use probably cannot be avoided at this time but in future
releases should we just redesign the big image on top of startup screen into something
smaller to avoid this problem?

Note that for shaded relief (and watersheds ?) there is a button in the GUI that is turned on
and will add the resulting map to the layer manager automatically so you don't need to add
it, although I am not sure this is set as a default on the live DVD version.

As for the legend and scale - we often have trouble moving around the legend
with mouse after moving around the scale - even if I click on
the legend the mouse still picks up the scale - there is a way around it but it can be
frustrating for the first time users. It may be safer to position the legend first and
the scale second but even that may be tricky. This may be the case only for mac users
(or perhaps windows) so it may work properly on Live DVD under linux.

For the vector material, I don't have anything for spearfish, but you can find some examples that produce
nice maps for nc data set using d.vect.thematic at the end og this page (including the GUI instrcutions)
http://courses.ncsu.edu/mea582/common/GIS_anal_grass/GIS_Anal_grvisual.html

and then some database stuff here (not really cleaned up)
http://courses.ncsu.edu/mea792/common/Assign_GISamodel/Vect_GISamodel.html

Helena

P.S. We have had discussion already, but talking about scale would it be possible to
change meters to m for the scale in future releases -
it looks funny written in full given that kilometers are km (and it also makes scale too long).
Given that mi should be used for miles
(see wiki - the National Institute of Standards and Technology uses and recommends mi),
there shouldn't be any confusion that m is meters.

Helena Mitasova
Associate Professor
Department of Marine, Earth, and Atmospheric Sciences
2800 Faucette Drive, Rm. 1125 Jordan Hall
North Carolina State University
Raleigh, NC 27695-8208
hmitaso@ncsu.edu

"All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.”

On Jul 22, 2012, at 8:32 PM, Hamish wrote:

Hi,

we're finalizing the next version of the OSGeo live sampler DVD,
USB stick, and Virtual Machine release. It's a live boot system
with all the latest goodies, including GRASS 6.4.2 and almost
all of the OSGeo family of desktop and server apps.
( http://live.osgeo.org )

It would be really helpful if others with fresh eyes could have
a look over the grass pages and double check that they're all
still relevant for v6.4.2, no dumb mistakes, you like what you
see, etc.

http://adhoc.osgeo.osuosl.org/livedvd/docs/en/overview/grass_overview.html

http://adhoc.osgeo.osuosl.org/livedvd/docs/en/quickstart/grass_quickstart.html

The finalized English version will go out for translation at the
end of this week. (conspicuously there are no Italian translations
yet even though there is a strong Italian FOSS4G presence, hint
hint :slight_smile:

There's a copy of the quickstart in the GRASS wiki. Note that
when I wrote it there was a GRASS workshop at the targeted FOSS4G
conference which was going to focus on vector processing, so I
didn't bother to put much of that in the quickstart. Contributions
demonstrating some fancy vector usage that makes GRASS special,
or a common vector task to show how GRASS does it easily, would
be appreciated. Just note that the quickstarts are intended to be
quick- i.e. 5-10 minutes worth, so we've got to keep it concise.

In addition, on the disc the QGIS-GRASS bridge is broken because
of this bug:
http://hub.qgis.org/issues/2947
https://trac.osgeo.org/osgeo/ticket/868

(QGIS doesn't take into account GRASS's mapset ownership rules,
so the plugin fails in a multi-user environment)

thanks for any assistance,
Hamish
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

Helena wrote:

the quickstart looks very nice and I did not find any major
problems (but other may have a look too)

I just have couple comments:

- given that OSgeo live will be mostly used by new users
should you even mention the tcltk interface? It sounds a little
bit confusing for people who don't know what TclTk is
and makes GRASS look overly complex.

the doc was written when the wxPython had just hit stable and
when it was not yet the default. I've not removed the reference
to the tcl/tk gui.

Instead, it would be useful to put here a link to the wxPython
GUI Help page so that people know where to find explanation of
what you are talking about later on (especially what each
button on the layer manager means - the verbal description may
not be clear enough for some users).

ok, done.

Also, the note about the netbook use probably cannot be
avoided at this time but in future releases should we just
redesign the big image on top of startup screen into something
smaller to avoid this problem?

I think there's a ticket about it. My idea in the past had been
to query the OS for the root window size, then, in order,
selectively not show the "world leading ..." text, the banner
image, and finally the "Welcome to $version" text until it was
small enough.

that paragraph moved to a tip box so that it's not as much in
the way.

Note that for shaded relief (and watersheds ?) there is a
button in the GUI that is turned on and will add the resulting
map to the layer manager automatically so you don't need to add
it, although I am not sure this is set as a default on the
live DVD version.

it is, the docs were just from an earlier version where you
had to do it manually. now fixed.

As for the legend and scale - we often have trouble moving
around the legend with mouse after moving around the scale -
even if I click on the legend the mouse still picks up the
scale - there is a way around it but it can be frustrating for
the first time users. It may be safer to position the legend
first and the scale second but even that may be tricky. This
may be the case only for mac users (or perhaps windows) so it
may work properly on Live DVD under linux.

it's on all platforms, the trick (workaround) is to think of
the d.legend and d.barscale renderings on the wx map canvase
as printed on a plate of glass, which is of finite size. you
can slide those two pieces of glass around, but if you go to
the left of the map decoration you be beyond the left side of
the piece of glass so dragging on nothing. likewise if you
drag off the top or the bottom of the pane of glass it won't
move it either (which top/bottom side that is different for
the two decorations). And finally if the legend's plate of
glass is on top of the barscale's one, you'll only be able to
drag the top bit of glass around, so to move the one below you
have to move the top one out of the way, then move the bottom
one, then move the top one back again. A well developed sense
of spatial relations helps your imagination get it right..
The current mechanism needs to be replaced, I'm pretty sure
there's a ticket open for this one too.

For the vector material, I don't have anything for
spearfish, but you can find some examples that produce
nice maps for nc data set using d.vect.thematic

note that most of the functionality of d.vect.thematic has
been moved into d.vect now, so d.vect.thematic can become
a whole lot simpler and just focus on the legendy things now.

at the end og this page (including the GUI instrcutions)
http://courses.ncsu.edu/mea582/common/GIS_anal_grass/GIS_Anal_grvisual.html

and then some database stuff here (not really cleaned up)
http://courses.ncsu.edu/mea792/common/Assign_GISamodel/Vect_GISamodel.html

any volunteers? :slight_smile:

P.S. We have had discussion already, but talking about scale
would it be possible to change meters to m for the scale in
future releases -

:-/

it looks funny written in full given that kilometers are km

but you don't see both at the same time.

(and it also makes scale too long).

you're talking about d.barscale, yes?

Given that mi should be used for miles
(see wiki - the National Institute of Standards and
Technology uses and recommends mi),
there shouldn't be any confusion that m is meters.

Given any chance of confusion I'd rather spell out the full
word. The IHO board which sets out the rendering rules for
global nautical charts also uses "M" (upper case) for nautical
miles. And merchant marine people trained in that method I've
come across have a complete mental block of thinking of it in
any other meaning. Then the SI-purists get in to objections
of "nm" for nautical miles regardless of obvious non-nanometer
context, and it all just does my head in.. it's an infinite
debate with no right answer. So I just spell everything out
whenever possible -- to me the most important thing to pursue is
clarity of meaning, not strict adherence to conflicting
abbreviation standards. (my 2c on that subject)

two small issues I noticed along the way:

* the wx barscale overlay did not automatically tick the "show"
box the first time I ran it on a fresh install.

* g.mkfontcap will write to e.g. ~/.gfontcap (as suggested by the
man page), but that is not able to be used by d.font or anything
else? Only work-around is to make the system's $GISBASE/etc
/fontcap file globally writiable or have the admin pre-seed it
somehow? (not so good on a managed system)

thanks,
Hamish

Hamish wrote:

the doc was written when the wxPython had just hit stable
and when it was not yet the default. I've not removed the
reference to the tcl/tk gui.

typo --> I've _now_ removed the reference to the tcl/tk gui.

H

On Tue, Jul 24, 2012 at 4:31 AM, Hamish <hamish_b@yahoo.com> wrote:

Helena wrote:

the quickstart looks very nice and I did not find any major
problems (but other may have a look too)

I just have couple comments:

As for the legend and scale - we often have trouble moving
around the legend with mouse after moving around the scale -
even if I click on the legend the mouse still picks up the
scale - there is a way around it but it can be frustrating for
the first time users. It may be safer to position the legend
first and the scale second but even that may be tricky. This
may be the case only for mac users (or perhaps windows) so it
may work properly on Live DVD under linux.

it's on all platforms, the trick (workaround) is to think of
the d.legend and d.barscale renderings on the wx map canvase
as printed on a plate of glass, which is of finite size. you
can slide those two pieces of glass around, but if you go to
the left of the map decoration you be beyond the left side of
the piece of glass so dragging on nothing. likewise if you
drag off the top or the bottom of the pane of glass it won't
move it either (which top/bottom side that is different for
the two decorations). And finally if the legend's plate of
glass is on top of the barscale's one, you'll only be able to
drag the top bit of glass around, so to move the one below you
have to move the top one out of the way, then move the bottom
one, then move the top one back again. A well developed sense
of spatial relations helps your imagination get it right..
The current mechanism needs to be replaced, I'm pretty sure
there's a ticket open for this one too.

I think the only way to make it work is to create some new flag in the
d.barscale and d.legend modules which prints the size of the image.
Otherwise the GUI has no information about the size, especially for
d.barscale. Something like ps.map -b flag is needed, like Hamish did
before. Would it be possible, Hamish?

Regards,
Anna

Anna wrote:

I think the only way to make it work is to create some new
flag in the d.barscale and d.legend modules which prints the
size of the image.
Otherwise the GUI has no information about the size,
especially for d.barscale. Something like ps.map -b flag is
needed, like Hamish did before. Would it be possible, Hamish?

mmm, everything is possible, but I don't really like adding new
options too much to work around problems, I'd rather try to fix
the problem itself. otherwise after a while all it gets out of
hand and the modules get too complicated to use. that is not
meant to be a ban on new options, just some pushback to make
sure the programmers have exhausted all other ideas first.

If cropping the image is the only way, I wonder if we could
write a g.pnmcrop the same way that was done for g.pnmcomp,
mimicking (or cloning) the NetPBM tool for that purpose. It
could output the final size to stdout as an optional flag.

But I think there must be a better way if we completely rethink
the approach to how it is implemented right now. What that is,
I'm not sure, but I feel it's out there somewhere waiting to be
found.

thanks,
Hamish

On Mon, Jul 23, 2012 at 2:32 AM, Hamish <hamish_b@yahoo.com> wrote:

There's a copy of the quickstart in the GRASS wiki. Note that
when I wrote it there was a GRASS workshop at the targeted FOSS4G
conference which was going to focus on vector processing, so I
didn't bother to put much of that in the quickstart. Contributions
demonstrating some fancy vector usage that makes GRASS special,
or a common vector task to show how GRASS does it easily, would
be appreciated. Just note that the quickstarts are intended to be
quick- i.e. 5-10 minutes worth, so we've got to keep it concise.

An example for vector processing using the basins raster created in
the previous step:

1) convert the basins to vector areas with r.to.vect -v feature=area
2) load average elevation for each basin to the attribute table with
v.rast.stats
3) assign the color table of the raster elevation using average
elevation values to the basins vector with v.colors column=elev_mean
raster=elevation.10m
4) display the vector using the colors in GRASSRGB

BTW, there is no raster map named 'elevation' in the spearfish
dataset, but e.g. 'elevation.10m' as seen in one of the screenshots in
the quickstart. This should be corrected throughout the quickstart.
It seems that the basins in 'Watersheds and streams' were created with
the minimum size of the exterior watershed basin threshold set to
100000 cells, not 10000 cells (missed a zero).

In addition, on the disc the QGIS-GRASS bridge is broken because
of this bug:
  http://hub.qgis.org/issues/2947
  https://trac.osgeo.org/osgeo/ticket/868

The live dvd could use sextante for QGIS to access GRASS
functionality. An added bonus would be that sextante has bindings to
several GIS engines, one of them being GRASS, but others should also
be already available on the live dvd.

Markus M

Markus Metz wrote:

An example for vector processing using the basins raster
created in the previous step:

1) convert the basins to vector areas with r.to.vect -v feature=area
2) load average elevation for each basin to the attribute table with
v.rast.stats
3) assign the color table of the raster elevation using average
elevation values to the basins vector with v.colors column=elev_mean
raster=elevation.10m
4) display the vector using the colors in GRASSRGB

thanks, that leads through nicely. Now added, with some short notes
about the attribute table manager, cartographic composer, and graphical
modeler under "other things to try".

http://grass.osgeo.org/wiki/Quick_wxGUI_tutorial#Vector_modules
http://adhoc.osgeo.osuosl.org/livedvd/docs/en/quickstart/grass_quickstart.html

BTW, there is no raster map named 'elevation' in the spearfish
dataset, but e.g. 'elevation.10m' as seen in one of the screenshots in
the quickstart. This should be corrected throughout the quickstart.

I know, I was trying to make the instructions generic enough to work
with either of the two datasets. I figured those using Spearfish would
see the .dem and .10m and figure out that they should pick one, but
those running in NC would be hunting around for an exact name which
didn't exist.

Due to severe space limitations on the disc I'm continually concerned
about having to remove the NC dataset, or cut it down to Helena's 40mb
set. We already ship geotiffs and shapefiles of the same maps as part of
the OSGeo Educational dataset for use by all of the 40 or so projects
contributing to the disc & their workshops, tutorials, etc. If the
instructions are vague enough the quickstart docs might survive an
upheaval in the available data. :slight_smile:

It seems that the basins in 'Watersheds and streams' were
created with the minimum size of the exterior watershed basin threshold
set to 100000 cells, not 10000 cells (missed a zero).

It's not a typo, I just ran it for elevation.dem not elevation.10m. Using
elevation.10m with a threshold of 10000 cells doesn't match the screenshot
but is not so bad, so just try to pick something in the middle and stay
loose.

> In addition, on the disc the QGIS-GRASS bridge is broken because
> of this bug:
> http://hub.qgis.org/issues/2947
> https://trac.osgeo.org/osgeo/ticket/868

The live dvd could use sextante for QGIS to access GRASS functionality.
An added bonus would be that sextante has bindings to several GIS
engines, one of them being GRASS, but others should also be already
available on the live dvd.

I believe Alex is working on that right now. Ideally we'll have lots of
ways to flow between GRASS and QGIS and R and PostGIS all setup and
working. helpers & testers are most welcome! IRC: OSGeoLive on Freenode
or the live-demo@o.o mailing list.

Hamish

Due to severe space limitations on the disc I'm continually concerned
about having to remove the NC dataset, or cut it down to Helena's 40mb
set.

Hamish - I highly recommend to use this smaller dat set - it is much easier for users
to understand (as I tried to avoid various obscure names there)
and you can still do a lot with it and the simple names also allow for more generic
quickstart docs (e.g. if somebody wants to use elevation or streams in other countries)
I started to use it as a baseline data set and then expand it with various specialized mapsets
for more advanced courses.
here it is again (it should be the same as on the grass website)
http://courses.ncsu.edu/mea792/common/Assign_GISamodel/Assignall.html

Helena

Hamish:

> Due to severe space limitations on the disc I'm continually concerned
> about having to remove the NC dataset, or cut it down to Helena's 40mb
> set.

Helena:

Hamish - I highly recommend to use this smaller dat set - it
is much easier for users to understand (as I tried to avoid various
obscure names there) and you can still do a lot with it and the simple
names also allow for more generic quickstart docs (e.g. if somebody wants
to use elevation or streams in other countries)
I started to use it as a baseline data set and then expand
it with various specialized mapsets for more advanced courses.
here it is again (it should be the same as on the grass website)
http://courses.ncsu.edu/mea792/common/Assign_GISamodel/Assignall.html

ok, done.

I note some left over Mac ._files and some empty .tmp/ cruft, and also
the SEARCH_PATH in user1 lists some mapsets not shipped with the smaller
dataset. harmless, but..

thanks,
Hamish

2012/7/23 Hamish <hamish_b@yahoo.com>:

Hi,

hi hamish

The finalized English version will go out for translation at the
end of this week. (conspicuously there are no Italian translations
yet even though there is a strong Italian FOSS4G presence, hint
hint :slight_smile:

How can we (Italian) start to work on translation?

thanks for any assistance,
Hamish

--
ciao
Luca

http://gis.cri.fmach.it/delucchi/
www.lucadelu.org

On Wed, Jul 25, 2012 at 5:59 AM, Hamish <hamish_b@yahoo.com> wrote:

Markus Metz wrote:

An example for vector processing using the basins raster
created in the previous step:

1) convert the basins to vector areas with r.to.vect -v feature=area
2) load average elevation for each basin to the attribute table with
v.rast.stats
3) assign the color table of the raster elevation using average
elevation values to the basins vector with v.colors column=elev_mean
raster=elevation.10m
4) display the vector using the colors in GRASSRGB

thanks, that leads through nicely. Now added, with some short notes
about the attribute table manager, cartographic composer, and graphical
modeler under "other things to try".

http://grass.osgeo.org/wiki/Quick_wxGUI_tutorial#Vector_modules
http://adhoc.osgeo.osuosl.org/livedvd/docs/en/quickstart/grass_quickstart.html

BTW, there is no raster map named 'elevation' in the spearfish
dataset, but e.g. 'elevation.10m' as seen in one of the screenshots in
the quickstart. This should be corrected throughout the quickstart.

I know, I was trying to make the instructions generic enough to work
with either of the two datasets. I figured those using Spearfish would
see the .dem and .10m and figure out that they should pick one, but
those running in NC would be hunting around for an exact name which
didn't exist.

So it is better to use the spearfish location but a raster name from
the NC dataset and let users hunt around for an exact name that does
not exist? Sorry for the provocation, but that is the status quo.
IMHO, if there are screenshots in the quickstart, the instructions
should be such that the screenshots can be reproduced, otherwise users
might be confused that their results do not look like the provided
screenshots (easy reproducibility of results is not such a bad idea).

You could add a short half-sentence to the very first paragraph, e.g.
"... some slight adjustments may be needed for the NC dataset, e.g. by
using elevation@PERMANENT instead of elevation.10m@PERMANENT."

It seems that the basins in 'Watersheds and streams' were
created with the minimum size of the exterior watershed basin threshold
set to 100000 cells, not 10000 cells (missed a zero).

It's not a typo, I just ran it for elevation.dem not elevation.10m. Using
elevation.10m with a threshold of 10000 cells doesn't match the screenshot
but is not so bad, so just try to pick something in the middle and stay
loose.

If the quickstart is supposed to target beginners, it would IMHO be
better to have matching examples and screenshots to avoid confusion.

Markus

Hamish wrote:

> The finalized English version will go out for translation at the
> end of this week. (conspicuously there are no Italian translations
> yet even though there is a strong Italian FOSS4G presence, hint
> hint :slight_smile:

Luca:

How can we (Italian) start to work on translation?

see the links here:
   http://adhoc.osgeo.osuosl.org/livedvd/lang_stats.html
   http://wiki.osgeo.org/wiki/Live_GIS_Translate

thanks,
Hamish

Hamish:

> I know, I was trying to make the instructions generic enough to work
> with either of the two datasets. I figured those using Spearfish would
> see the .dem and .10m and figure out that they should pick one, but
> those running in NC would be hunting around for an exact name which
> didn't exist.

MarkusM:

So it is better to use the spearfish location but a raster name from
the NC dataset and let users hunt around for an exact name that does
not exist?

Yes! I'd like to encourage a little exploring and making the student
think instead of exactly telling every move to make click by click. But
no so abstract that they get lost. If you are told to pick the "elevation"
map but are only presented with "elevation.dem" and "elevation.10m", both
of which work fine, I'd argue that it's healthy to make them make a
personal choice, they get more personal reward from it that way (sort
of like telling them to select their home-town from a global dataset).
[ok, I'm reaching out on the limb :)]

Sorry for the provocation, but that is the status quo.

No need to apologize, we get a better end result from frank but courteous
discussion.

I do take your point, and I've now reworded it slightly to give the
exact map names at the start and make it more like "click on your elevation
map" instead of "click on the `elevation` map".

IMHO, if there are screenshots in the quickstart, the instructions
should be such that the screenshots can be reproduced, otherwise users
might be confused that their results do not look like the provided
screenshots (easy reproducibility of results is not such a bad idea).

it is positive feedback indeed, and that's critical (especially in the
first 5 minutes of the tutorial when they are very unsure of what they're
doing), but again I wouldn't mind a little deviations and choose your
own adventure to it. (both ways visually look fine, and fwiw I'd say the
smaller threshold size has a better looking v.colors result)
Or you can read that as the payoff:effort ratio for me to cut a new
screenshot has not yet exceeded 1.0.

I'm much more concerned about the time they'll spend in the menus and
tabs trying to find first the r.watershed module and then the threshold
option. I know it well and where it is, and still I have to hunt for it.
Lots of screenshots with menu entries circled in red with arrows and things
would be great (Markus N did one like that for the startup screen help
page), but is beyond the scope of this quickstart which is already a bit
on the large side for what it's supposed to be.

You could add a short half-sentence to the very first
paragraph, e.g. "... some slight adjustments may be needed for the NC
dataset, e.g. by using elevation@PERMANENT instead of
elevation.10m@PERMANENT."

Done, but I left off the @PERMANENT part, those are the only maps to
choose from at that point so you can't go wrong & it just clutters the
text.

cheers,
Hamish

On Wed, Jul 25, 2012 at 1:32 PM, Hamish <hamish_b@yahoo.com> wrote:

Hamish:

> I know, I was trying to make the instructions generic enough to work
> with either of the two datasets. I figured those using Spearfish would
> see the .dem and .10m and figure out that they should pick one, but
> those running in NC would be hunting around for an exact name which
> didn't exist.

MarkusM:

So it is better to use the spearfish location but a raster name from
the NC dataset and let users hunt around for an exact name that does
not exist?

Yes! I'd like to encourage a little exploring and making the student
think instead of exactly telling every move to make click by click. But
no so abstract that they get lost.

... well, with the today's click-and-try mentality this means: click-and-throw.

Examples *must* work out of the box from my point of view.
Also, there is not much advertisement coming from an outdated dataset
of the '80th (Spearfish) while having for many years a way more
recent data set available, published, and meanwhile used in most
GRASS manual pages.

...

No need to apologize, we get a better end result from frank but courteous
discussion.

Yes: so please let's forget about the "Spearfish" data set.

My 0.02 cent,
markusN

ok, done.

I note some left over Mac ._files and some empty .tmp/ cruft, and also
the SEARCH_PATH in user1 lists some mapsets not shipped with the smaller
dataset. harmless, but..

Hamish - thanks for pointing it out - I will fix it and post a new one ASAP, Helena

thanks,
Hamish
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

Hamish,

thanks for pointing it out - there was a lot of stuff that I did not know was there,
it should be cleaned up now, please let me know if you still find something
or if there are any problems. Also, I can export the data into some external formats - perhaps shape files
and geotiff? in case somebody would like to use it with other software.

http://courses.ncsu.edu/mea792/common/media/gisdemo.zip

Helena

On Jul 25, 2012, at 1:20 AM, Hamish wrote:

Hamish:

Due to severe space limitations on the disc I'm continually concerned
about having to remove the NC dataset, or cut it down to Helena's 40mb
set.

Helena:

Hamish - I highly recommend to use this smaller dat set - it
is much easier for users to understand (as I tried to avoid various
obscure names there) and you can still do a lot with it and the simple
names also allow for more generic quickstart docs (e.g. if somebody wants
to use elevation or streams in other countries)
I started to use it as a baseline data set and then expand
it with various specialized mapsets for more advanced courses.
here it is again (it should be the same as on the grass website)
http://courses.ncsu.edu/mea792/common/Assign_GISamodel/Assignall.html

ok, done.

I note some left over Mac ._files and some empty .tmp/ cruft, and also
the SEARCH_PATH in user1 lists some mapsets not shipped with the smaller
dataset. harmless, but..

thanks,
Hamish
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

Helena wrote:

thanks for pointing it out - there was a lot of stuff that I
did not know was there, it should be cleaned up now, please
let me know if you still find something or if there are any
problems.

...

http://courses.ncsu.edu/mea792/common/media/gisdemo.zip

ok, thanks. should we copy that into
   http://grass.osgeo.org/sampledata/north_carolina/

as .tar.gz too? keeping the same filename?

Also, I can export the data into some external formats -
perhaps shape files and geotiff? in case somebody would
like to use it with other software.

We are already doing that for NC shape, rast_geotiff, and kml
tarballs at the above sampledata/ URL. The cloning of the data
was one reason I felt pretty shaky about shipping the full NC
grass Location in addition to the shp/gtiff files.

Also along with selected Natural Earth data layers, an
OpenStreetMap data extract, and links to a bunch of other
random shapefiles and geotiffs that other projects supply
as part of their quickstart docs, all in a ~/data/ directory.
Finally, there's an icon on the Desktop which leads to a OSGeo
website listing links to geodata & courseware downloads needed
for various workshops and tutorials that were too big to fit on
the disc or not ready at the time the discs went to press.

Hamish

Markus N wrote:

Examples *must* work out of the box from my point of view.

(they do, afaik)

Also, there is not much advertisement coming from an
outdated dataset of the '80th (Spearfish) while having for
many years a way more recent data set available, published,
and meanwhile used in most GRASS manual pages.

due to the tight space limitations it was the only one that was
safe not to be cut. with the nc_basic one it's better now, but
I've only had that in place since yesterday.

does geodata age? in defense of the Spearfish dataset, it is
both small and comprehensive, even if it doesn't show off modern
geo-layers as much and is a bit crusty. but point taken.

Hamish