[GRASS-dev] Re: [GRASS GIS] #160: WinGRASS: v.report> incompatability issue

GRASS GRASS developer
1

Thanks very much Marco.

Also, I don't think there is any reason that binaries (and especially needed
dependencies like coreutils) cannot be updated between GRASS version
releases.

Michael

On 5/13/08 9:00 AM, "grass-dev-request@lists.osgeo.org"
<grass-dev-request@lists.osgeo.org> wrote:

Message: 1
Date: Mon, 12 May 2008 16:00:16 -0000
From: "GRASS GIS" <trac@osgeo.org>
Subject: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS: v.report
incompatability issue
To: undisclosed-recipients:;
Message-ID: <053.1d439124360a83589c8bc8bd20682ef1@osgeo.org>
Content-Type: text/plain; charset="utf-8"

#160: WinGRASS: v.report incompatability issue
-------------------------+--------------------------------------------------
  Reporter: isaacullah | Owner: grass-dev@lists.osgeo.org
      Type: defect | Status: closed
  Priority: major | Milestone: 6.4.0
Component: default | Version: 6.3.0
Resolution: fixed | Keywords: v.report paste WinGRASS
-------------------------+--------------------------------------------------
Changes (by 4everskiff):

  * status: new => closed
  * resolution: => fixed

Comment:

I already did the job on my local machine. All the MSYS coreutils will be
added to the next WinGRASS release.

This done, I checked the v.report module using the North-Carolina sample
database: it worked fine for me.

Until the next release will be published, do as follows:

Download the coreutils package from here:
MinGW - Minimalist GNU for Windows - Browse Files at SourceForge.net
snapshot.tar.bz2

Unpack it to a temporary folder, then copy all the content of the
coreutils-5.97 folder to %your GRASS installation path%\msys

Marco

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

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

2

Michael,

Yes. The new installer is already in my GRASS folder.
Just tell me if upload it or not (and with what name, 6.3.0-1?)

Marco

-----Messaggio originale-----
Da: grass-dev-bounces@lists.osgeo.org
[mailto:grass-dev-bounces@lists.osgeo.org] Per conto di Michael Barton
Inviato: martedì 13 maggio 2008 18.40
A: grass-dev@lists.osgeo.org
Oggetto: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS:
v.report>incompatability issue

Thanks very much Marco.

Also, I don't think there is any reason that binaries (and especially needed
dependencies like coreutils) cannot be updated between GRASS version
releases.

Michael

On 5/13/08 9:00 AM, "grass-dev-request@lists.osgeo.org"
<grass-dev-request@lists.osgeo.org> wrote:

Message: 1
Date: Mon, 12 May 2008 16:00:16 -0000
From: "GRASS GIS" <trac@osgeo.org>
Subject: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS: v.report
incompatability issue
To: undisclosed-recipients:;
Message-ID: <053.1d439124360a83589c8bc8bd20682ef1@osgeo.org>
Content-Type: text/plain; charset="utf-8"

#160: WinGRASS: v.report incompatability issue
-------------------------+--------------------------------------------
-------------------------+------
  Reporter: isaacullah | Owner: grass-dev@lists.osgeo.org
      Type: defect | Status: closed
  Priority: major | Milestone: 6.4.0
Component: default | Version: 6.3.0
Resolution: fixed | Keywords: v.report paste WinGRASS
-------------------------+--------------------------------------------
-------------------------+------
Changes (by 4everskiff):

  * status: new => closed
  * resolution: => fixed

Comment:

I already did the job on my local machine. All the MSYS coreutils
will be added to the next WinGRASS release.

This done, I checked the v.report module using the North-Carolina
sample
database: it worked fine for me.

Until the next release will be published, do as follows:

Download the coreutils package from here:
http://downloads.sourceforge.net/mingw/coreutils-5.97-MSYS-1.0.11-
snapshot.tar.bz2

Unpack it to a temporary folder, then copy all the content of the
coreutils-5.97 folder to %your GRASS installation path%\msys

Marco

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

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

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

3

Great.

I don't think you need to version your installer. GRASS is the same version
that it was previously, you have just updated the GRASS installer. So I'd
just put a date on it so that people know it's newer than the old installer.
In fact, I would not change the version because that implies a different
version of GRASS.

Michael

On 5/14/08 1:56 PM, "Marco Pasetti" <marco.pasetti@alice.it> wrote:

Michael,

Yes. The new installer is already in my GRASS folder.
Just tell me if upload it or not (and with what name, 6.3.0-1?)

Marco

-----Messaggio originale-----
Da: grass-dev-bounces@lists.osgeo.org
[mailto:grass-dev-bounces@lists.osgeo.org] Per conto di Michael Barton
Inviato: martedì 13 maggio 2008 18.40
A: grass-dev@lists.osgeo.org
Oggetto: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS:
v.report>incompatability issue

Thanks very much Marco.

Also, I don't think there is any reason that binaries (and especially needed
dependencies like coreutils) cannot be updated between GRASS version
releases.

Michael

On 5/13/08 9:00 AM, "grass-dev-request@lists.osgeo.org"
<grass-dev-request@lists.osgeo.org> wrote:

Message: 1
Date: Mon, 12 May 2008 16:00:16 -0000
From: "GRASS GIS" <trac@osgeo.org>
Subject: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS: v.report
incompatability issue
To: undisclosed-recipients:;
Message-ID: <053.1d439124360a83589c8bc8bd20682ef1@osgeo.org>
Content-Type: text/plain; charset="utf-8"

#160: WinGRASS: v.report incompatability issue
-------------------------+--------------------------------------------
-------------------------+------
  Reporter: isaacullah | Owner: grass-dev@lists.osgeo.org
      Type: defect | Status: closed
  Priority: major | Milestone: 6.4.0
Component: default | Version: 6.3.0
Resolution: fixed | Keywords: v.report paste WinGRASS
-------------------------+--------------------------------------------
-------------------------+------
Changes (by 4everskiff):

  * status: new => closed
  * resolution: => fixed

Comment:

I already did the job on my local machine. All the MSYS coreutils
will be added to the next WinGRASS release.

This done, I checked the v.report module using the North-Carolina
sample
database: it worked fine for me.

Until the next release will be published, do as follows:

Download the coreutils package from here:
MinGW - Minimalist GNU for Windows - Browse Files at SourceForge.net
snapshot.tar.bz2

Unpack it to a temporary folder, then copy all the content of the
coreutils-5.97 folder to %your GRASS installation path%\msys

Marco

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

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
grass-dev Info Page

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

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

4

Marco:

> Yes. The new installer is already in my GRASS folder.
> Just tell me if upload it or not (and with what name,
> 6.3.0-1?)

or -2.

I don't think you need to version your installer. GRASS
is the same version
that it was previously, you have just updated the GRASS
installer. So I'd
just put a date on it so that people know it's newer
than the old installer.
In fact, I would not change the version because that
implies a different version of GRASS.

the typical (linux) convention is to label by source code version then "-" package revision. so 6.3.0-2 would seem appropriate here. For source code changes and extra "." is used, like "6.3.0.1". I would hesitate to put the date in there unless it is a SVN snapshot or the line of package revisions is someway broken. If date is used, please use a sortable way like YYYYMMDD.

Hamish

5

-2 is correct for me, since I already submitted a patch for the installer script and updated the 6.3.0 installer some days ago.
Now I need to figure out how to introduce some changes in the GRASS regkey layout to let the installer recognize if the current installed GRASS (if present) is the latest release or not, and thus decide if update the installed copy.
A the moment, the GRASS release is deteminated by the SVN revision number; it represents binaries built upon different source code distributions.
I think that I should introduce a new code number, that represents releases with differences in binaries, libraries, install script patches… that is all but not differences in GRASS source code: I could call it BNR (Binary Release Code) and start it with 0 every time the SVN rev number has changed in the binary release.

What do you think about?

Marco

Da: Hamish [mailto:hamish_b@yahoo.com]
Inviato: gio 15/05/2008 4.33
A: marco.pasetti@alice.it; grass-dev@lists.osgeo.org; Michael Barton
Oggetto: Re: R: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS: v.report>incompatability issue

Marco:

Yes. The new installer is already in my GRASS folder.
Just tell me if upload it or not (and with what name,
6.3.0-1?)

or -2.

I don’t think you need to version your installer. GRASS
is the same version
that it was previously, you have just updated the GRASS
installer. So I’d
just put a date on it so that people know it’s newer
than the old installer.
In fact, I would not change the version because that
implies a different version of GRASS.

the typical (linux) convention is to label by source code version then “-” package revision. so 6.3.0-2 would seem appropriate here. For source code changes and extra “.” is used, like “6.3.0.1”. I would hesitate to put the date in there unless it is a SVN snapshot or the line of package revisions is someway broken. If date is used, please use a sortable way like YYYYMMDD.

Hamish

6

A the moment, the GRASS release is deteminated by the SVN
revision number; it represents binaries built upon
different source code distributions.
I think that I should introduce a new code number, that
represents releases with differences in binaries,
libraries, install script patches... that is all but not
differences in GRASS source code: I could call it BNR
(Binary Release Code) and start it with 0 every time the
SVN rev number has changed in the binary release.

What do you think about?

FWIW this is how debian expresses it:
http://www.debian.org/doc/debian-policy/ch-controlfields.html#s-f-Version

not that they should dictate what we do, but fyi. (their "BNR" resets to "-1" upon new code releases; "-0" is sometimes used for a release candidate or preparatory packages)

the version before the hyphen is the source code release's domain, after is the packager's domain.

Hamish

7

Hamish,

Thanks. I'm working on the new installer, that should automatically
recognize if and what version of GRASS is installed, and, if newer, asks the
user if he wants to upgrade to the newer version.

Marco

-----Messaggio originale-----
Da: Hamish [mailto:hamish_b@yahoo.com]
Inviato: giovedì 15 maggio 2008 11.33
A: grass-dev@lists.osgeo.org; Michael Barton; marco.pasetti@alice.it
Oggetto: Re: R: R: [GRASS-dev] Re: [GRASS GIS] #160: WinGRASS:
v.report>incompatability issue

A the moment, the GRASS release is deteminated by the SVN revision
number; it represents binaries built upon different source code
distributions.
I think that I should introduce a new code number, that represents
releases with differences in binaries, libraries, install script
patches... that is all but not differences in GRASS source code: I
could call it BNR (Binary Release Code) and start it with 0 every time
the SVN rev number has changed in the binary release.

What do you think about?

FWIW this is how debian expresses it:
http://www.debian.org/doc/debian-policy/ch-controlfields.html#s-f-Version

not that they should dictate what we do, but fyi. (their "BNR" resets to
"-1" upon new code releases; "-0" is sometimes used for a release candidate
or preparatory packages)

the version before the hyphen is the source code release's domain, after is
the packager's domain.

Hamish

8

Hi all,

I finished my work on the installer: now it is capable to recognize if and what version of GRASS is already installed on the system and tells the user if he wants to upgrade to the newer version.
Now we need to decide the installer name. At the moment it is:

WinGRASS-SourceCodeRelese-BinaryRelease.exe

where the BinaryRelease number reflects only changes in the binaries (but not in the GRASS sources used to build them) and/or the installer, starting from 1 every time we have a new SourceCodeRelese.

Since 6.3.0 has been released, I updated the installer once time, to fix the bug #143 [1], so now we should be at the BinaryRelease N° 3 (1: the first release, 2: the update to fix #143), but I always released the previous intallers as 6.3.0, and not 6.3.0-1 and 6.3.0-2 (sorry, my fault). That means that releasing 6.3.0-3 now might be confusing…

what do you think about?

Marco

[1] http://trac.osgeo.org/grass/ticket/143

9

On Mon, May 19, 2008 at 11:27 AM, <marco.pasetti@alice.it> wrote:
...

Since 6.3.0 has been released, I updated the installer once time, to fix the
bug #143 [1], so now we should be at the BinaryRelease N° 3 (1: the first
release, 2: the update to fix #143), but I always released the previous
intallers as 6.3.0, and not 6.3.0-1 and 6.3.0-2 (sorry, my fault). That
means that releasing 6.3.0-3 now might be confusing...

what do you think about?

To me 6.3.0-3 sounds good.

Markus

10

I think it’s OK to release 6.3.0-3. I guess I don’t think people will wonder where 1 and 2 went. You could put a little message in the README if you want.

Michael

On 5/19/08 2:27 AM, “marco.pasetti@alice.it” marco.pasetti@alice.it wrote:

Hi all,

I finished my work on the installer: now it is capable to recognize if and what version of GRASS is already installed on the system and tells the user if he wants to upgrade to the newer version.
Now we need to decide the installer name. At the moment it is:

WinGRASS-SourceCodeRelese-BinaryRelease.exe

where the BinaryRelease number reflects only changes in the binaries (but not in the GRASS sources used to build them) and/or the installer, starting from 1 every time we have a new SourceCodeRelese.

Since 6.3.0 has been released, I updated the installer once time, to fix the bug #143 [1], so now we should be at the BinaryRelease N° 3 (1: the first release, 2: the update to fix #143), but I always released the previous intallers as 6.3.0, and not 6.3.0-1 and 6.3.0-2 (sorry, my fault). That means that releasing 6.3.0-3 now might be confusing…

what do you think about?

Marco

[1] http://trac.osgeo.org/grass/ticket/143

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

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton