[GRASS5] documentation

That's peanuts. :slight_smile:
We have to inform people about what is going on anyway.
I hope when more people help with the webpages, we get more dynamic.
(wait a second)

  THIS _is_ A CALL for HELP with DOCUMENTATION and communication.

(sorry, had to be done :slight_smile: )

Just check a project like Abiword,
which does a lot to inform people (weekly reports, bug votes...)
It is really appreciated. We need to find people to organise this for GRASS.

From a first quick evaluation of the questionnaire it follows that a
lot of people stumble over the lack of coherent documentation. I think
this should be a priority for the future and since I'm not much worth
in development, I'd be willing to work on that. I think we need to
start a general review of everything that is available, merge usable
stuff and write whatever is needed.

BTW, Markus+Helena, can we publish the english translation of the
manual ? Even if it needs polishing it might be better than what we
have now, and publishing it might attract people to work on it.

Also, on the French mailing list there is the initiative to launch a
documentation project. I think it would be better if we could
coordinate the whole thing.

Did I tell you about the _very_ conservative user base. :slight_smile:

FYI: of the ca. 60 responses I've received for the questionnaire
(i.e. something around a 10% responses rate) the overwhelming majority
uses 5.x. But that just might be because the 90% that didn't answer
constitute that conservative user base.

I'll try to finish up the evaluation of the questionnaire in the next
few days...

Moritz

On Tue, Apr 23, 2002 at 09:40:28PM +0200, Moritz Lennert wrote:

> THIS _is_ A CALL for HELP with DOCUMENTATION and communication.

> Just check a project like Abiword,
> which does a lot to inform people (weekly reports, bug votes...)
> It is really appreciated. We need to find people to organise this for GRASS.
From a first quick evaluation of the questionnaire it follows that a
lot of people stumble over the lack of coherent documentation.

Understandable.

I think this should be a priority for the future
and since I'm not much worth in development,
I'd be willing to work on that.

This would be very useful.

Also, on the French mailing list there is the initiative to launch a
documentation project. I think it would be better if we could
coordinate the whole thing.

I agree.

If you need anything in technical support, let us know. :slight_smile:

start a general review of everything that is available, merge usable stuff and
write whatever is needed.

It would be absolutely wonderful if you could review what is there - some time
ago
I suggested to Markus that just giving each document a rating (maybe *, **, ***)
in terms of how much up-to-date and usable it is would help a lot and then sort
them by this rating.
This of course means reading everything and testing those materials that are
considered
up-to-date in terms of GRASS5. And of course the rating may need updates from
time to time.
So it is quite a bit of work but it would be really useful.
As far as I know the best, most comprehensive and up-to date material about GRASS
is the users manual,
most of the rest is not really being updated.

BTW, Markus+Helena, can we publish the english translation of the manual ? Even
if it needs polishing it might be better than what we have now, and publishing
it might attract people to work on it.

The problem with the old german book is that it is old and it may cause quite a
bit of confusion rather than help.
It reads well but just take it and start doing things as described there.
For example, important (and for users often confusing) issues related to NULL
data and floating point
are there described only briefly as specialties of GRASS5, while in fact they are
now major features that
almost everybody has to deal with. Or try to process and analyze your site data -
that has changed a lot too.
We have found that we had to essentially rewrite most of the book to make it
usable for GRASS5.0 and at the same time systematically test big part of GRASS
which was a huge effort, and it is still not as complete as it should be.

So I really believe that instead of trying to fix the old book (which would be a
monumental work, as it requires much more than polishing), it would be actually
faster and better to write new on-line tutorial based on the latest GRASS5
release, rather than rely on GRASS4* materials. We have already set it up, I
think that Markus has a link to it.

I think that it should be more task oriented rather than "type of data" oriented
as is the users manual. For example it should have:
Setting-up a project, Importing Data, Integrating data from different coordinate
systems, Digitizing, Viewing maps, Working with surfaces, Computi ng the basic
statistics, etc. For some tasks we already have nice tutorials, e.g. map algebra
or nviz, so those could be updated and used directly.
Also it should be structured for on-line learning, more graphics, and interaction
than the book's plain text. (for example for our on-line erosion modeling
tutorial we have the commands linked with manual pages and maps, so you can learn
everything about the command that you are using by clicking on it and also see
how your inputs and outputs may look like:
see http://skagit.meas.ncsu.edu/~helena/gmslab/erosion/usped.html
or the later version of the whole thing at
http://skagit.meas.ncsu.edu/~helena/gmslab/reports/CerlErosionTutorial/denix/denixstart.html)

This format (Markus set it up with php instead of frames) would be also more
suitable for getting more people to work on it.
And translations should be much easier as a lot of the text could be "bulletized"
rather than writing complex sentences which are more difficult to translate.

So that is just a suggestion

Helena

On Tue, Apr 23, 2002 at 09:12:56PM -0500, Helena wrote:

> start a general review of everything that is available, merge usable stuff and
> write whatever is needed.

It would be absolutely wonderful if you could review what is there - some time
ago
I suggested to Markus that just giving each document a rating (maybe *, **, ***)
in terms of how much up-to-date and usable it is would help a lot and then sort
them by this rating.
This of course means reading everything and testing those materials that are
considered
up-to-date in terms of GRASS5. And of course the rating may need updates from
time to time.
So it is quite a bit of work but it would be really useful.

I will look into that, but as you say, it will be a lot of work.

So I really believe that instead of trying to fix the old book (which would be a
monumental work, as it requires much more than polishing), it would be actually
faster and better to write new on-line tutorial based on the latest GRASS5
release, rather than rely on GRASS4* materials. We have already set it up, I
think that Markus has a link to it.

There already is a new tutorial ?

I think that it should be more task oriented rather than "type of data" oriented
as is the users manual. For example it should have:
Setting-up a project, Importing Data, Integrating data from different coordinate
systems, Digitizing, Viewing maps, Working with surfaces, Computi ng the basic
statistics, etc. For some tasks we already have nice tutorials, e.g. map algebra
or nviz, so those could be updated and used directly.
Also it should be structured for on-line learning, more graphics, and interaction
than the book's plain text. (for example for our on-line erosion modeling
tutorial we have the commands linked with manual pages and maps, so you can learn
everything about the command that you are using by clicking on it and also see
how your inputs and outputs may look like:
see http://skagit.meas.ncsu.edu/~helena/gmslab/erosion/usped.html
or the later version of the whole thing at
http://skagit.meas.ncsu.edu/~helena/gmslab/reports/CerlErosionTutorial/denix/denixstart.html)

This format (Markus set it up with php instead of frames) would be also more
suitable for getting more people to work on it.
And translations should be much easier as a lot of the text could be "bulletized"
rather than writing complex sentences which are more difficult to translate.

So that is just a suggestion

But a very goog one :wink:

I have been thinking along the same lines and I would propose somethin
like this:

1) A man page of a manual should contain links to parts of the
manual/examples where it is used. A while ago I sent a message to the
list, asking whether it would be possible to include a keywords
section the man pages which would then be used to link to tutorials
(i.e. on the d.rast page have a link to "displaying maps", "using
raster data", etc..).

2) If we use the keywords carefully, we can automatically create a
page that lists all the keywords with links to the relevant man pages
and more detailed tutorials if the latter exist. This could be the
beginning of a tutorial and people could add and maintain small
tutorials on one of these aspects. If there is no tutorial, at least
the user has all the relevant modules listed.

3) I think that seen the limited amount of people, this formula allows
   incremental growth of documentation, while being usable almost
   immediately.

So, if someone could confirm that 1)+2) are possible and maybe point
me into the right directions on how to do it, I am willing to begin
with that. Once the keyword page (i.e. the table of contents of a
task-oriented tutorial) is ready, I (and others) can weed through the
existing jungle and select those documents that should be linked to
from that page. In a third phase we then have to write whatever is missing.

Moritz

On Wed, Apr 24, 2002 at 10:28:52AM +0200, Moritz Lennert wrote:

On Tue, Apr 23, 2002 at 09:12:56PM -0500, Helena wrote:
> > start a general review of everything that is available, merge usable
> > stuff and write whatever is needed.

[...]

> So I really believe that instead of trying to fix the old book (which
> would be a monumental work, as it requires much more than polishing), it
> would be actually faster and better to write new on-line tutorial based
> on the latest GRASS5 release, rather than rely on GRASS4* materials. We
> have already set it up, I think that Markus has a link to it.

There already is a new tutorial ?

You can find an outlined version here:
http://grass.itc.it/gdp/grass5tutor/

This must be filled with contents now.
Eventually we should use "docbook" or something
similar (http://www.docbook.org/)

I have been thinking along the same lines and I would propose somethin
like this:

1) A man page of a manual should contain links to parts of the
manual/examples where it is used. A while ago I sent a message to the
list, asking whether it would be possible to include a keywords
section the man pages which would then be used to link to tutorials
(i.e. on the d.rast page have a link to "displaying maps", "using
raster data", etc..).

... and at least I agreed :slight_smile: In fact I already contacted colleagues
here to generated keywords semiautomated (as it is a huge task for
nearly 400 pages). But up to now I do not have a good suggestion
for software to use.

2) If we use the keywords carefully, we can automatically create a
page that lists all the keywords with links to the relevant man pages
and more detailed tutorials if the latter exist. This could be the
beginning of a tutorial and people could add and maintain small
tutorials on one of these aspects. If there is no tutorial, at least
the user has all the relevant modules listed.

Such an index would be very welcome.

3) I think that seen the limited amount of people, this formula allows
   incremental growth of documentation, while being usable almost
   immediately.

But I don't think that a tutorial can grow from an index. I stick
more with Helena's idea of writing task oriented. However, the
tutorial should not miss such an index.

So, if someone could confirm that 1)+2) are possible and maybe point
me into the right directions on how to do it, I am willing to begin
with that. Once the keyword page (i.e. the table of contents of a
task-oriented tutorial) is ready, I (and others) can weed through the
existing jungle and select those documents that should be linked to
from that page. In a third phase we then have to write whatever is missing.

Moritz

Markus

On Wed, Apr 24, 2002 at 11:43:33AM +0200, Markus Neteler wrote:

On Wed, Apr 24, 2002 at 10:28:52AM +0200, Moritz Lennert wrote:
> On Tue, Apr 23, 2002 at 09:12:56PM -0500, Helena wrote:
> > > start a general review of everything that is available, merge usable
> > > stuff and write whatever is needed.
[...]
> > So I really believe that instead of trying to fix the old book (which
> > would be a monumental work, as it requires much more than polishing), it
> > would be actually faster and better to write new on-line tutorial based
> > on the latest GRASS5 release, rather than rely on GRASS4* materials. We
> > have already set it up, I think that Markus has a link to it.
>
> There already is a new tutorial ?

You can find an outlined version here:
http://grass.itc.it/gdp/grass5tutor/

This must be filled with contents now.
Eventually we should use "docbook" or something
similar (http://www.docbook.org/)

It looks good. however, it is data oriented, not task oriented as
Helena suggested.

> I have been thinking along the same lines and I would propose somethin
> like this:
>
> 1) A man page of a manual should contain links to parts of the
> manual/examples where it is used. A while ago I sent a message to the
> list, asking whether it would be possible to include a keywords
> section the man pages which would then be used to link to tutorials
> (i.e. on the d.rast page have a link to "displaying maps", "using
> raster data", etc..).

... and at least I agreed :slight_smile: In fact I already contacted colleagues
here to generated keywords semiautomated (as it is a huge task for
nearly 400 pages). But up to now I do not have a good suggestion
for software to use.

No idea there, but I believe that if a few people are willing to spend
a bit of time at this, we should be able to do this fairly quickly,
especially if we have pre-defined keywords.

> 2) If we use the keywords carefully, we can automatically create a
> page that lists all the keywords with links to the relevant man pages
> and more detailed tutorials if the latter exist. This could be the
> beginning of a tutorial and people could add and maintain small
> tutorials on one of these aspects. If there is no tutorial, at least
> the user has all the relevant modules listed.

Such an index would be very welcome.

> 3) I think that seen the limited amount of people, this formula allows
> incremental growth of documentation, while being usable almost
> immediately.

But I don't think that a tutorial can grow from an index. I stick
more with Helena's idea of writing task oriented. However, the
tutorial should not miss such an index.

I guess I was not clear enough. The keywords I was thinking about
would be task-oriented. So we might use the outline you've already
made, or make another, task-oriented one. This would give us an idea
about the keywords to use (corresponding to sections or subsections of
the tutorial (i.e. data_import, data_display, erosion_modelling,
image_rectification, etc, etc). So, if we have the outline and man
pages which contain the respective keywords, it should be quite easy
to automatically create the links to the relevant man pages in every
(sub)section of the tutorial. This would already allow a new user to
go to the section "displaying data in Grass" and, even if there is no
tutorial text, yet, she/he would at least find links to the relevant
man pages. Once we have written a tutorial on how to display data,
there will also be a link to this tutorial. Thus, the tutorial would
be usable almost immediately, and filled up little by little.
(http://grass.itc.it/gdp/grassmanuals/grass50_commands_sorted.html
already contains the same idea, but without links and with each module in
only one category.)

Any developer contributing a new man pages would be asked to include
the keywords she/he thinks relevant. And I and others would make a
deliberate effort to write section tutorials as quickly as possible.

Does this sound reasonable ?

Moritz

Moritz Lennert wrote:

On Wed, Apr 24, 2002 at 11:43:33AM +0200, Markus Neteler wrote:
> On Wed, Apr 24, 2002 at 10:28:52AM +0200, Moritz Lennert wrote:
> > On Tue, Apr 23, 2002 at 09:12:56PM -0500, Helena wrote:
> > > > start a general review of everything that is available, merge usable
> > > > stuff and write whatever is needed.
> [...]
> > > So I really believe that instead of trying to fix the old book (which
> > > would be a monumental work, as it requires much more than polishing), it
> > > would be actually faster and better to write new on-line tutorial based
> > > on the latest GRASS5 release, rather than rely on GRASS4* materials. We
> > > have already set it up, I think that Markus has a link to it.
> >
> > There already is a new tutorial ?
>
> You can find an outlined version here:
> http://grass.itc.it/gdp/grass5tutor/
>
> This must be filled with contents now.
> Eventually we should use "docbook" or something
> similar (http://www.docbook.org/)

It looks good. however, it is data oriented, not task oriented as
Helena suggested.

Moritz I am changing the outline right now to task oriented. I will submit the
change
to CVS and Markus can set-up automated update of the web document. New items or
themes
can always be added but it would be good to have a good basic structure.
Are you able to use CVS? If yes, you could get the manual after I fix the outline

and you can start updating the first sections. You can also use these first
sections
to test the coordination with the french effort .

Markus I think that we should change the copyright there on GRASS documentation
team
or something like that.

Helena

On Wed, Apr 24, 2002 at 01:12:25PM -0400, Helena Mitasova wrote:

Moritz Lennert wrote:

> On Wed, Apr 24, 2002 at 11:43:33AM +0200, Markus Neteler wrote:
> > On Wed, Apr 24, 2002 at 10:28:52AM +0200, Moritz Lennert wrote:
> > > On Tue, Apr 23, 2002 at 09:12:56PM -0500, Helena wrote:
> > > > > start a general review of everything that is available, merge usable
> > > > > stuff and write whatever is needed.
> > [...]
> > > > So I really believe that instead of trying to fix the old book (which
> > > > would be a monumental work, as it requires much more than polishing), it
> > > > would be actually faster and better to write new on-line tutorial based
> > > > on the latest GRASS5 release, rather than rely on GRASS4* materials. We
> > > > have already set it up, I think that Markus has a link to it.
> > >
> > > There already is a new tutorial ?
> >
> > You can find an outlined version here:
> > http://grass.itc.it/gdp/grass5tutor/
> >
> > This must be filled with contents now.
> > Eventually we should use "docbook" or something
> > similar (http://www.docbook.org/)
>
> It looks good. however, it is data oriented, not task oriented as
> Helena suggested.

Moritz I am changing the outline right now to task oriented. I will submit the
change
to CVS and Markus can set-up automated update of the web document. New items or
themes
can always be added but it would be good to have a good basic structure.
Are you able to use CVS? If yes, you could get the manual after I fix the outline

and you can start updating the first sections. You can also use these first
sections
to test the coordination with the french effort .

Markus I think that we should change the copyright there on GRASS documentation
team or something like that.

yes - feel free to change.

Markus

On Wed, Apr 24, 2002 at 01:12:25PM -0400, Helena Mitasova wrote:

Moritz I am changing the outline right now to task oriented. I will submit the
change
to CVS and Markus can set-up automated update of the web document. New items or
themes
can always be added but it would be good to have a good basic structure.

I agree. We should try to rapidly have a fixed list of keywords (while
staying flexible obviously).

Are you able to use CVS? If yes, you could get the manual after I fix the outline

Yes, CVS is ok. I just need to get write access once we start working
on it.

and you can start updating the first sections. You can also use these first
sections
to test the coordination with the french effort .

Ok, I'm still waiting for feedback from their side.

Moritz