Re: [R-pkg-devel] Change package name
On 6/21/23 8:33 PM, Rolf Turner wrote:
On Tue, 20 Jun 2023 17:39:48 -0400
wrote:
If you really need a new package name, is there an accepted way to
update the old name to at least provide a message for a while as to
the new name to use and perhaps have the load otherwise fail?
I don't know about "accepted", but the following paradigm seems to me
to be sensible and reasonably efficacious:
* you have a package "foo" the name of which you wish to change to "bar"
* create a new package "bar" which is the same as "foo" except for its
name
* now replace foo with a skeleton package which has only the file
First.R in its R directory
* the file First.R should be something like:
.onAttach <- function(lib, pkg) {
ver <- read.dcf(file.path(lib, pkg, "DESCRIPTION"), "Version")
packageStartupMessage(paste(pkg, ver))
msg <- paste("\n This package, \"foo\" is now",
" deprecated. Users",
"\n should install and use its",
" successor \"bar\".\n")
packageStartupMessage(msg) }
* submit both "bar" and "foo" to CRAN
Years ago, I had package "RSiteSearch". I asked some R list what
might be a better name for that. "sos" was suggested, and Duncan
Murdoch contributed code for an enhancement. Then I added something
like ".onAttach" to "RSiteSearch" and kept it on CRAN for a couple of
years after introducing "sos". That allowed people to use the old
package for a couple of years without forcing them to switch immediately.
I think that worked quite well.
Hope this helps.
Spencer Graves
cheers,
Rolf Turner
__
R-package-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
Re: [R-pkg-devel] Change package name
On Tue, 20 Jun 2023 17:39:48 -0400
wrote:
> If you really need a new package name, is there an accepted way to
> update the old name to at least provide a message for a while as to
> the new name to use and perhaps have the load otherwise fail?
I don't know about "accepted", but the following paradigm seems to me
to be sensible and reasonably efficacious:
* you have a package "foo" the name of which you wish to change to "bar"
* create a new package "bar" which is the same as "foo" except for its
name
* now replace foo with a skeleton package which has only the file
First.R in its R directory
* the file First.R should be something like:
> .onAttach <- function(lib, pkg) {
> ver <- read.dcf(file.path(lib, pkg, "DESCRIPTION"), "Version")
> packageStartupMessage(paste(pkg, ver))
> msg <- paste("\n This package, \"foo\" is now",
> " deprecated. Users",
> "\n should install and use its",
> " successor \"bar\".\n")
> packageStartupMessage(msg) }
* submit both "bar" and "foo" to CRAN
cheers,
Rolf Turner
--
Honorary Research Fellow
Department of Statistics
University of Auckland
Stats. Dep't. (secretaries) phone:
+64-9-373-7599 ext. 89622
Home phone: +64-9-480-4619
__
R-package-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
Re: [R-pkg-devel] Change package name
Dear All, Thanks very much for the positive and helpful replies here. All the best, William > On 20/06/2023 19:36 CEST Uwe Ligges wrote: > > > On 20.06.2023 18:00, Dirk Eddelbuettel wrote: > > > > Hi William, > > > > On 20 June 2023 at 16:06, William Becker wrote: > > | I am the maintainer of a package which is unfortunately involved in a > > complicated dispute regarding its intellectual property (since the package > > was partly built under a contract with an organisation), but also the > > "branding" of the package, i.e. the name. > > | > > | The story is long and complicated, but suffice to say that the name of > > the package is apparently creating a misleading connection with the > > organisation, and this is causing difficulties on both sides. > > | > > | I am aware that changing the name of a package is very disruptive, and I > > am not considering it lightly. However just for information at this stage, > > I wonder if anyone could tell me whether packages have ever changed names > > on CRAN, what the rules are in these cases, and if there is any advice for > > minimising the disruption. > > | > > | To reiterate, I am not (yet) planning to do this, but exploring options > > and looking for advice. > > > > You presented the reasoning convincingly and are aware of the general "we > > would rather not do this" sentiment. Ultimately, this (as so often) is > > CRAN's > > call so you have to do that (directly and not via this list). My sense is > > that you have a case but only CRAN can tell. > > Yes, CRAN will accept this case. > > Best, > Uwe Ligges > > > > > > > Good luck, Dirk > > __ R-package-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-package-devel
