[GRASS-dev] WinGRASS-6.3.0RC5 Self Installer

GRASS GRASS developer
21

Marco Pasetti ha scritto:

Glynn,

In fact, I'm quite worried that the main consequence of an easy-to-use

installer could be a large number of useless bug reports which don't provide
any useful information and just end up distracting (and possibly
de-motivating) the developers who are working on the Windows version.

Sorry Glinn, I do not see your point here.

Anyone who wants a robust version of GRASS for Windows should be using the

Cygwin version.

I would like to say that, from a normal user's point of view, the cygwin
solution is simply not feasible.
I wish to thank Marco for his great and productive effort. His installer
is a major breakthrough for the spreading out of GRASS (something we all
like, isn't it?).
All the best.
pc
--
Paolo Cavallini, see: http://www.faunalia.it/pc

22

On Sat, 15 Mar 2008, Marco Pasetti wrote:

[...]

- No help pages from main help menu (no message displayed)

As I think I mentioned in an earlier e-mail, I suspect this is due to the problem with the Windows _spawn* functions and arguments with spaces in them. I managed to find the earlier patch to g.parser I proposed, but which I am still not very sure about:
http://lists.osgeo.org/pipermail/grass-dev/2007-June/031931.html

Maybe we could test if this removes the problem?

Paul

23

Paul,

As I said in previous e-mails, in next weeks I'll have enough time to
compile and test grass_trunk, even if recent comments let me reflect on how
useful is all this work...

Marco

-----Messaggio originale-----
Da: Paul Kelly [mailto:paul-grass@stjohnspoint.co.uk]
Inviato: sabato 15 marzo 2008 12.30
A: Marco Pasetti
Cc: 'Moritz Lennert'; 'Glynn Clements'; 'Michael Barton';
grass-dev@lists.osgeo.org
Oggetto: Re: [GRASS-dev] WinGRASS-6.3.0RC5 Self Installer

On Sat, 15 Mar 2008, Marco Pasetti wrote:

[...]

- No help pages from main help menu (no message displayed)

As I think I mentioned in an earlier e-mail, I suspect this is due to the
problem with the Windows _spawn* functions and arguments with spaces in
them. I managed to find the earlier patch to g.parser I proposed, but which
I am still not very sure about:
http://lists.osgeo.org/pipermail/grass-dev/2007-June/031931.html

Maybe we could test if this removes the problem?

Paul

24

On 15/03/08 11:48, Marco Pasetti wrote:

Glynn,

In fact, I'm quite worried that the main consequence of an easy-to-use

installer could be a large number of useless bug reports which don't provide
any useful information and just end up distracting (and possibly
de-motivating) the developers who are working on the Windows version.

Anyone who wants a robust version of GRASS for Windows should be using the

Cygwin version.

Thanks a lot for de-motivating ME on working this way!
Actually, I really could have spent all those hours and energies to do
something else, ...my thesis, for example.
If you think that my work is not only useless for WinGRASS, but even
counterproductive for it, I can stop it right now. It really would be a
great saving of wasted time!

I don't think that Glynn meant that your work is useless and wasted time ! On the contrary, I think your installer will allow many more people to test winGRASS, as it takes away the hurdle of having to install manually.

I think what Glynn means is that we have to be very honest about what we are selling here. winGRASS works well in many aspects, but there are still quite a lot of problems. You might not see them the first time around, but when you actually start using it for production work, you will sooner or later stumble upon them. So, if we now go out and say: "Here's winGRASS, it's perfect unless you install it in a path with spaces", then we risk creating more dissatisfied "I told you so" users, because that's not true.

You say:

I decided to force a "working installation" for the following
reason: the aim of all this work is, obviously, to spread GRASS as most as possible, because there are a lot of Windows users out there... so, to convince them to use GRASS, we need to be sure that it will definetely work on their machines, because Windows users are not used to *test* software and report bugs...

At this stage, winGRASS cannot "definitely work on their machines", as there are still many winGRASS specific bugs which go far beyond the spaces in paths problem (BTW we need to clean up the wiki wingrass status page). We should probably clearly state (and if possible at the beginning of the installation procedure) that this is alpha-stage software (I don't know what the threshold is for calling it beta ?) and that they are quite probably going to encounter problems sooner or later, but that we would be very grateful if they would still try and tell us about their experience.

AFAIK, many MS Windows software providers release at least beta-versions for testing, and people do test. This is what we have been doing, and your installer greately facilitates this process.

Let's just be frank about the state of affairs.

Moritz

25

Let's go with number one, if you don't mind. It encourages people to help in debugging.

Michael
____________________
C. Michael Barton, Professor of Anthropology
Director of Graduate Studies
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University

Phone: 480-965-6262
Fax: 480-965-7671
www: <www.public.asu.edu/~cmbarton>

On Mar 15, 2008, at 3:50 AM, Marco Pasetti wrote:

Sorry, confused: what?

This

"WARNING: you are about to install GRASS into a directory that has spaces in
either its name or the path of directories leading up to it.
Some functionalities of GRASS might be hampered by this. We would highly
appreciate if you tried and reported any problems, so that we can fix them.
However, if you want to avoid any such issues, we recommended that you
choose a simple installation path without spaces such as: C:\GRASS."

Or this?

"WARNING: you are about to install GRASS into a directory that has spaces in
either its name or the path of directories leading up to it. You may go
ahead, but without customization of some GRASS modules, you may encounter
unexpected behaviour or failure of some functions. It is highly recommended
to choose a simple installation path without spaces such as: C:\GRASS."

Marco

-----Messaggio originale-----
Da: Michael Barton [mailto:c.michael.barton@gmail.com] Per conto di Michael
Barton
Inviato: venerdì 14 marzo 2008 16.24
A: Moritz Lennert
Cc: marco.pasetti@alice.it; Glynn Clements; grass-dev@lists.osgeo.org
Oggetto: Re: R: WinGRASS-6.3.0RC5 Self Installer

I like this.

Michael
____________________
C. Michael Barton, Professor of Anthropology Director of Graduate Studies
School of Human Evolution & Social Change Center for Social Dynamics &
Complexity Arizona State University

Phone: 480-965-6262
Fax: 480-965-7671
www: <www.public.asu.edu/~cmbarton>

On Mar 14, 2008, at 5:24 AM, Moritz Lennert wrote:

On 14/03/08 12:08, marco.pasetti@alice.it wrote:

Hi,
Please, suggest me also the message to display.
Benjamin suggested as follows:
"WARNING: you are about to install GRASS into a directory that has
spaces in either its name or the path of directories leading up to
it. You may go ahead, but expect trouble and weird behaviour from
GRASS modules. It is highly recommended to choose a simple
installation path without spaces such as: C:\GRASS."

I find this a bit too strong. How about

"WARNING: you are about to install GRASS into a directory that has
spaces in either its name or the path of directories leading up to it.
Some functionalities of GRASS might be hampered by this. We would
highly appreciate if you tried and reported any problems, so that we
can fix them. However, if you want to avoid any such issues, we
recommended that you choose a simple installation path without spaces
such as: C:\GRASS."

Moritz

Marco
---------------------------------------------------------------------
---
*Da:* Moritz Lennert [mailto:mlennert@club.worldonline.be]
*Inviato:* ven 14/03/2008 10.38
*A:* marco.pasetti@alice.it
*Cc:* Glynn Clements; Michael Barton; grass-dev@lists.osgeo.org
*Oggetto:* Re: WinGRASS-6.3.0RC5 Self Installer On 14/03/08 10:25,
marco.pasetti@alice.it wrote:

Hi all,

sorry for the late, I'm really drowning in workflows...

I don't think this is an appropriate answer.

I'm really afraid, I didn't want to give a "bad" answer! I

apologise if > I did something wrong!
No problem.

This said, if you prefer to not have a *forced* installation,

I'll > prepare a *only warnings* one; just tell me what I have to
do...
+1 for the "only warnings" version.
Moritz

26

Anyone who wants a robust version of GRASS for Windows should be using

the

Cygwin version.

In a corporate environment, getting permission from IT branch
to install a native windows application is a lot easier than
getting permission to install a new environment (Cygwin).

nick

***********************************************************************
WARNING: This e-mail (including any attachments) may contain legally
privileged, confidential or private information and may be protected by
copyright. You may only use it if you are the person(s) it was intended
to be sent to and if you use it in an authorised way. No one is
allowed to use, review, alter, transmit, disclose, distribute, print
or copy this e-mail without appropriate authority.

If this e-mail was not intended for you and was sent to you by mistake,
please telephone or e-mail me immediately, destroy any hardcopies of
this e-mail and delete it and any copies of it from your computer
system. Any right which the sender may have under copyright law, and
any legal privilege and confidentiality attached to this e-mail is not
waived or destroyed by that mistake.

It is your responsibility to ensure that this e-mail does not contain
and is not affected by computer viruses, defects or interference by
third parties or replication problems (including incompatibility with
your computer system).

Opinions contained in this e-mail do not necessarily reflect the
opinions of the Queensland Department of Main Roads, Queensland
Transport or Maritime Safety Queensland, or endorsed organisations
utilising the same infrastructure.
***********************************************************************

27

Moritz Lennert wrote:
...

I don't think this is an appropriate answer. Glynn did not necessarily
mean that you are the one that has to host these files. In fact they
should be on the osgeo GRASS site next to the installer, but to just
shrug it off this way is not the way to go. Making sources available is
a fundamental element of free software.

...

So, to shorten a long discussion, I just used this simple command line
to download all the source packages mentioned in your guide:

for i in `grep "source code" BuildFromSource.html | awk -F'=' '{print
$3}' | awk -F'"' '{print $2}' | sort`; do wget $i; done

And I downloaded the msys, flex and bison sources manually.

Everything is available at
http://geog-pc40.ulb.ac.be/grass/wingrass/wingrass_sources/ including
the version of the GRASS sources you used (don't think this is
necessary, but just to be complete). All packages are available
individually, and there also is a wingrass_sources.tar which contains
them all.

Maybe Markus can just get the tar and put it on the download site next
to the installer ?

you mean:
wingrass_sources.tar 13-Mar-2008 11:48 45M
?

I am not sure if we should really store that on the OSGeo
server. Then we have to do the same also for all the other
binaries and fill up the server space with overhead.
If you insist, I suggest to discuss this on the GRASS-PSC list.

We could store the hard-to-get sources, but why host PROJ4,
GDAL and such?

Markus

...

[1]

- MSYS (1.0.11)
- Flex (2.5.4a-1)
- Bison (2.1)
- Zlib (1.2.3)
- Libpng (1.2.24)
- Libtiff (3.8.2)
- Xdr (4.0)
- Freetype (2.3.5)
- FFTW (3.1.2)
- PDCurses (3.3)
- PROJ.4 (4.6.0)
- GEOS (2.2.3)
- PostgreSQL (8.2.6)
- SQLite (3.5.6)
- GDAL (1.5.0) *
- Tcl/Tk (8.5.1)

[2] http://www.webalice.it/marco.pasetti/grass/BuildFromSource.html

28

On 25/03/08 10:13, Markus Neteler wrote:

Moritz Lennert wrote: ...

I don't think this is an appropriate answer. Glynn did not necessarily mean that you are the one that has to host these files.
In fact they should be on the osgeo GRASS site next to the installer, but to just shrug it off this way is not the way to go.
Making sources available is a fundamental element of free
software.

...

So, to shorten a long discussion, I just used this simple command line to download all the source packages mentioned in your guide:

for i in `grep "source code" BuildFromSource.html | awk -F'=' '{print $3}' | awk -F'"' '{print $2}' | sort`; do wget $i; done

And I downloaded the msys, flex and bison sources manually.

Everything is available at http://geog-pc40.ulb.ac.be/grass/wingrass/wingrass_sources/ including the version of the GRASS sources you used (don't think this is necessary, but just to be complete). All packages are available individually, and there also is a wingrass_sources.tar which contains them all.

Maybe Markus can just get the tar and put it on the download site next to the installer ?

you mean: wingrass_sources.tar 13-Mar-2008 11:48 45M ?

I am not sure if we should really store that on the OSGeo server. Then we have to do the same also for all the other binaries and fill
up the server space with overhead. If you insist, I suggest to discuss this on the GRASS-PSC list.

What do you mean by "all the other binaries" ?

We could store the hard-to-get sources, but why host PROJ4, GDAL and
such?

Here is what Glynn wrote a bit earlier in this thread:

Please note that, if you provide binaries which are covered by the GPL, you must provide the corresponding source code for download *from the same place*. It isn't sufficient to point to the source on a different site.

Alternatively, you can provide a written offer to supply the source code upon request, but that means that you have to keep those exact versions of the source code handy for the next 3 years (the website where you obtained it may cease to provide it when a new version is released).

Moritz