
The Rcpp Core Team is once again thrilled, pleased, and chuffed (am I
doing this right for LinkedIn?) to announce a new release (now at
1.0.14) of the
Rcpp package. It
arrived on
CRAN earlier today,
and has since been uploaded to
Debian. Windows and macOS builds
should appear at CRAN in the next few days, as will builds in different
Linux distribution and of course
r2u should catch up
tomorrow too. The release was only uploaded yesterday, and as always get
flagged because of the grandfathered
.Call(symbol) as well
as for the url to the
Rcpp book (which
has remained unchanged for years) failing . My email reply was promptly
dealt with under European morning hours and by the time I got up the
submission was in state waiting over a single reverse-dependency
failure which is also spurious, appears on some systems and not
others, and also not new. Imagine that: nearly 3000 reverse dependencies
and only one (spurious) change to worse. Solid testing seems to help. My
thanks as always to the
CRAN
for responding promptly.
This release continues with the six-months January-July cycle started
with
release
1.0.5 in July 2020. This time we also need a one-off
hotfix
release 1.0.13-1: we had (accidentally) conditioned an upcoming R
change on 4.5.0, but it already came with 4.4.2 so we needed to adjust
our code. As a reminder, we do of course make interim snapshot dev or
rc releases available via the
Rcpp drat repo as well as
the
r-universe page and
repo and strongly encourage their use and testing I run my systems
with these versions which tend to work just as well, and are also fully
tested against all reverse-dependencies.
Rcpp has long established itself
as
the most popular way of enhancing
R with C or C++ code. Right now,
2977 packages on
CRAN depend on
Rcpp for making analytical code go
faster and further. On CRAN, 13.6% of all packages depend (directly) on
Rcpp, and 60.8% of all compiled
packages do. From the cloud mirror of CRAN (which is but a subset of all
CRAN downloads),
Rcpp has been
downloaded 93.7 million times. The two published papers (also included
in the package as preprint vignettes) have, respectively, 1947 (
JSS, 2011) and 354 (
TAS, 2018)
citations, while the the
book (Springer useR!,
2013) has another 676.
This release is primarily incremental as usual, generally preserving
existing capabilities faithfully while smoothing our corners and / or
extending slightly, sometimes in response to changing and tightened
demands from
CRAN or
R standards. The move towards a
more standardized approach for the C API of R once again to a few
changes; Kevin did once again did most of these PRs. Other contributed
PRs include G bor permitting builds on yet another BSD variant, Simon
Guest correcting
sourceCpp() to work on read-only files,
Marco Colombo correcting a (surprisingly large) number of vignette
typos, I aki rebuilding some documentation files that tickled (false)
alerts, and I took care of a number of other maintenance items along the
way.
The full list below details all changes, their respective PRs and, if
applicable, issue tickets. Big thanks from all of us to all
contributors!
Changes in
Rcpp release version 1.0.14 (2025-01-11)
- Changes in Rcpp API:
- Support for user-defined databases has been removed (Kevin in #1314 fixing #1313)
- The
SET_TYPEOF function and macro is no longer used
(Kevin in #1315
fixing #1312)
- An errorneous cast to
int affecting large return
object has been removed (Dirk in #1335 fixing #1334)
- Compilation on DragonFlyBSD is now supported (G bor Cs rdi in #1338)
- Use read-only
VECTOR_PTR and STRING_PTR
only with with R 4.5.0 or later (Kevin in #1342 fixing #1341)
- Changes in Rcpp Attributes:
- The
sourceCpp() function can now handle input files
with read-only modes (Simon Guest in #1346 fixing #1345)
- Changes in Rcpp Deployment:
- One unit tests for arm64 macOS has been adjusted; a macOS
continuous integration runner was added (Dirk in #1324)
- Authors@R is now used in DESCRIPTION as mandated by CRAN, the
Rcpp.package.skeleton() function also creates it (Dirk in
#1325 and #1327)
- A single datetime format test has been adjusted to match a change
in R-devel (Dirk in #1348 fixing #1347)
- Changes in Rcpp Documentation:
- The Rcpp Modules vignette was extended slightly following #1322
(Dirk)
- Pdf vignettes have been regenerated under Ghostscript 10.03.1 to
avoid a false positive by a Windows virus scanner (I aki in #1331)
- A (large) number of (old) typos have been corrected in the
vignettes (Marco Colombo in #1344)
Thanks to my
CRANberries, you
can also look at a
diff
to the previous release Questions, comments etc should go to the
rcpp-devel
mailing list off the
R-Forge page.
Bugs reports are welcome at the
GitHub issue tracker as
well (where one can also search among open or closed issues).
This post by Dirk
Eddelbuettel originated on his Thinking inside the box
blog. If you like this or other open-source work I do, you can sponsor me at
GitHub.