problems with installing maptools package (due to Foreign package)

[giohappy]

I'm configuring my new debian lenny, and as I was installing R and the
related spatial packages I've faced some problems with maptools
package setup.
I've installed R via apt, from the debian official repositories. The R
versione installed is 2.5.1.
Then I've run the usual commands:

install.packages(c("sp", "maptools", "rgdal"), dependencies=TRUE).
(Are these packages still enough for gstat?)

Everything seems to have gone ok for sp and rgdal (having previously
"apt-get-installed" libgdal1-dev and the other 27 Mb of
dependencies!!!)
The problem has come with maptools setup which har raised an error
when trying to load "foreign" package for lazy loading. The exit error
said

"This is R 2.5.1, package 'foreing' needs >= 2.6.0"

What should I do? It seems there's an inconcistency between the
packaged versions, but I can I upgrade to 2.6 *without* compiling it?

Thanks in advance,
Giovanni

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[Daniel McInerney]

Hi Giovanni,
The current version of foreign on CRAN is 0.8-23,
which depends on R >= 2.6.0
You could try and install a previous version of
foreign either (0.8-22 or 0.8-21) available under
the Archived Packages on the cran.r-project website.

If you don't want to compile R 2.6, you could add
the sid repository to your apt-sources. You should be
able to safely build for lenny using the sid sources.

Hope this helps.
regards,

Daniel.

G. Allegri wrote:

> I'm configuring my new debian lenny, and as I was installing R and the
> related spatial packages I've faced some problems with maptools
> package setup.
> I've installed R via apt, from the debian official repositories. The R
> versione installed is 2.5.1.
> Then I've run the usual commands:
>
> install.packages(c("sp", "maptools", "rgdal"), dependencies=TRUE).
> (Are these packages still enough for gstat?)
>
> Everything seems to have gone ok for sp and rgdal (having previously
> "apt-get-installed" libgdal1-dev and the other 27 Mb of
> dependencies!!!)
> The problem has come with maptools setup which har raised an error
> when trying to load "foreign" package for lazy loading. The exit error
> said
>
> "This is R 2.5.1, package 'foreing' needs >= 2.6.0"
>
> What should I do? It seems there's an inconcistency between the
> packaged versions, but I can I upgrade to 2.6 *without* compiling it?
>
> Thanks in advance,
> Giovanni
>
> _______________________________________________
> statsgrass mailing list
> [hidden email]
> http://grass.itc.it/mailman/listinfo/statsgrass
>

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[giohappy]

Thanks Daniel,
I've momentaneously inserted the sid repositories to install r-base.
Everything seems to have gone right. I've gstat and spgrass6 back
again! :-)
It has given me troubles with some recomended packages (exit errors
for "tkrplot", "rgl" and "gam"), but things seem to work ok anyway...
I'll solve later whit these.

Bye,
Giovanni

2007/10/31, Daniel McInerney <[hidden email]>:

> Hi Giovanni,
> The current version of foreign on CRAN is 0.8-23,
> which depends on R >= 2.6.0
> You could try and install a previous version of
> foreign either (0.8-22 or 0.8-21) available under
> the Archived Packages on the cran.r-project website.
>
> If you don't want to compile R 2.6, you could add
> the sid repository to your apt-sources. You should be
> able to safely build for lenny using the sid sources.
>
> Hope this helps.
> regards,
>
> Daniel.
>
> G. Allegri wrote:
> > I'm configuring my new debian lenny, and as I was installing R and the
> > related spatial packages I've faced some problems with maptools
> > package setup.
> > I've installed R via apt, from the debian official repositories. The R
> > versione installed is 2.5.1.
> > Then I've run the usual commands:
> >
> > install.packages(c("sp", "maptools", "rgdal"), dependencies=TRUE).
> > (Are these packages still enough for gstat?)
> >
> > Everything seems to have gone ok for sp and rgdal (having previously
> > "apt-get-installed" libgdal1-dev and the other 27 Mb of
> > dependencies!!!)
> > The problem has come with maptools setup which har raised an error
> > when trying to load "foreign" package for lazy loading. The exit error
> > said
> >
> > "This is R 2.5.1, package 'foreing' needs >= 2.6.0"
> >
> > What should I do? It seems there's an inconcistency between the
> > packaged versions, but I can I upgrade to 2.6 *without* compiling it?
> >
> > Thanks in advance,
> > Giovanni
> >
> > _______________________________________________
> > statsgrass mailing list
> > [hidden email]
> > http://grass.itc.it/mailman/listinfo/statsgrass
> >
>

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[Edzer Pebesma-2]

Giovani, package gstat is formally only dependent on sp, according to
http://cran.r-project.org/src/contrib/Descriptions/gstat.html
(rgdal and fields are only needed to run the regression tests)

Of course you may need maptools and rgdal to do all kinds of stuff...
--
Edzer

G. Allegri wrote:

> I'm configuring my new debian lenny, and as I was installing R and the
> related spatial packages I've faced some problems with maptools
> package setup.
> I've installed R via apt, from the debian official repositories. The R
> versione installed is 2.5.1.
> Then I've run the usual commands:
>
> install.packages(c("sp", "maptools", "rgdal"), dependencies=TRUE).
> (Are these packages still enough for gstat?)
>
> Everything seems to have gone ok for sp and rgdal (having previously
> "apt-get-installed" libgdal1-dev and the other 27 Mb of
> dependencies!!!)
> The problem has come with maptools setup which har raised an error
> when trying to load "foreign" package for lazy loading. The exit error
> said
>
> "This is R 2.5.1, package 'foreing' needs >= 2.6.0"
>
> What should I do? It seems there's an inconcistency between the
> packaged versions, but I can I upgrade to 2.6 *without* compiling it?
>
> Thanks in advance,
> Giovanni
>
> _______________________________________________
> statsgrass mailing list
> [hidden email]
> http://grass.itc.it/mailman/listinfo/statsgrass
>  

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[giohappy]

Thanks Edzer,
I've prefered to install all these packages just in case... :-)

2007/10/31, Edzer J. Pebesma <[hidden email]>:

> Giovani, package gstat is formally only dependent on sp, according to
> http://cran.r-project.org/src/contrib/Descriptions/gstat.html
> (rgdal and fields are only needed to run the regression tests)
>
> Of course you may need maptools and rgdal to do all kinds of stuff...
> --
> Edzer
>
> G. Allegri wrote:
> > I'm configuring my new debian lenny, and as I was installing R and the
> > related spatial packages I've faced some problems with maptools
> > package setup.
> > I've installed R via apt, from the debian official repositories. The R
> > versione installed is 2.5.1.
> > Then I've run the usual commands:
> >
> > install.packages(c("sp", "maptools", "rgdal"), dependencies=TRUE).
> > (Are these packages still enough for gstat?)
> >
> > Everything seems to have gone ok for sp and rgdal (having previously
> > "apt-get-installed" libgdal1-dev and the other 27 Mb of
> > dependencies!!!)
> > The problem has come with maptools setup which har raised an error
> > when trying to load "foreign" package for lazy loading. The exit error
> > said
> >
> > "This is R 2.5.1, package 'foreing' needs >= 2.6.0"
> >
> > What should I do? It seems there's an inconcistency between the
> > packaged versions, but I can I upgrade to 2.6 *without* compiling it?
> >
> > Thanks in advance,
> > Giovanni
> >
> > _______________________________________________
> > statsgrass mailing list
> > [hidden email]
> > http://grass.itc.it/mailman/listinfo/statsgrass
> >
>
>

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[HamishB]

> G. Allegri wrote:
> > The problem has come with maptools setup which har raised an error
> > when trying to load "foreign" package for lazy loading. The exit error
> > said
> >
> > "This is R 2.5.1, package 'foreing' needs >= 2.6.0"

Daniel McInerney wrote:
> The current version of foreign on CRAN is 0.8-23,
> which depends on R >= 2.6.0
> You could try and install a previous version of
> foreign either (0.8-22 or 0.8-21) available under
> the Archived Packages on the cran.r-project website.
>
> If you don't want to compile R 2.6, you could add
> the sid repository to your apt-sources. You should be
> able to safely build for lenny using the sid sources.


R 2.6.0 should arrive in testing in the next day or few,
  http://packages.qa.debian.org/r/r-base.html

I hope there is a good reason for the version dep (which there well may be), as
it is very annoying when things are not backwards compatible for trival
reasons. Not everyone is running or wants to run the latest Gentoo.


Hamish


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com 

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass

[Roger Bivand]

On Wed, 31 Oct 2007, Hamish wrote:

>> G. Allegri wrote:
>>> The problem has come with maptools setup which har raised an error
>>> when trying to load "foreign" package for lazy loading. The exit error
>>> said
>>>
>>> "This is R 2.5.1, package 'foreing' needs >= 2.6.0"
>
> Daniel McInerney wrote:
>> The current version of foreign on CRAN is 0.8-23,
>> which depends on R >= 2.6.0
>> You could try and install a previous version of
>> foreign either (0.8-22 or 0.8-21) available under
>> the Archived Packages on the cran.r-project website.
>>
>> If you don't want to compile R 2.6, you could add
>> the sid repository to your apt-sources. You should be
>> able to safely build for lenny using the sid sources.
>
>
> R 2.6.0 should arrive in testing in the next day or few,
>  http://packages.qa.debian.org/r/r-base.html
>
> I hope there is a good reason for the version dep (which there well may be), as
> it is very annoying when things are not backwards compatible for trival
> reasons. Not everyone is running or wants to run the latest Gentoo.
>

The foreign package is "recommended", so is distributed as part of the
2.6.0 release, and should not be present in a 2.5.1 install. The maptools
package (which spgrass6 depends on) depends on foreign, because the
shapelib dbf read/write functions are (now) in foreign. The current
foreign 0.8-23 depends on R >= 2.6.0, so running update.packages() in R
2.5.1 will *not* update it. "Recommended" packages are typically freshened
at a release, so as to utilise improvements in "base" packages and in the
R engine itself.

This means that the real question is how a 2.5.1 system got polluted with
a version of foreign that is too "new", where update.packages() could not
do it. It looks either like packaging slackness, or user intervention to
force an update. Looking at:

Package: r-cran-foreign
Priority: optional
Section: math
Installed-Size: 652
Maintainer: Dirk Eddelbuettel <[hidden email]>
Architecture: i386
Source: foreign
Version: 0.8.22-1~etchcran.2.6.0
Replaces: r-recommended (<< 1.9.0)
Depends: libc6 (>= 2.3.6-6), r-base-core (>> 2.5.1)
Filename: etch/r-cran-foreign_0.8.22-1~etchcran.2.6.0_i386.deb
Size: 147498
MD5sum: 8961360323b9cbff2e1a701e31057640
SHA1: 749e3fe450a5419f6fe30191b413ea310bc5ffde
SHA256: 244857df424a8b4394466cfcda27c8d7cfec09bba3702776f88bb40e5b4ab1f6
Description: GNU R package to read/write data from other stat. systems
  This package provides functions for reading and writing data stored by
  statistical packages such as Minitab, S, SAS, SPSS, Stata, ...
  .
  This package is part of the set of packages that are 'recommended'
  by R Core and shipped with upstream source releases of R itself.

on http://cran.r-project.org/bin/linux/debian/etch/Packages

the condition on this 0.8-22 build is R strictly greater than 2.5.1, which
looks OK, and will prevent updating (same for stable).

Roger

>
> Hamish
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> _______________________________________________
> statsgrass mailing list
> [hidden email]
> http://grass.itc.it/mailman/listinfo/statsgrass
>

--
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: [phone]; fax [phone]
e-mail: [hidden email]

_______________________________________________
statsgrass mailing list
[hidden email]
http://grass.itc.it/mailman/listinfo/statsgrass