Advertisement ยท 728 ร— 90
#
Hashtag
#r2u
Advertisement ยท 728 ร— 90
Dual-figure chart with #r2u cumulative downloads in million on the left, hitting 72 millions, and average weekly download (on a log scale) beating 100lk/week a first time as we hit 4.29 million in the month of March.

Dual-figure chart with #r2u cumulative downloads in million on the left, hitting 72 millions, and average weekly download (on a log scale) beating 100lk/week a first time as we hit 4.29 million in the month of March.

#r2u had its first 'four million packages delivered' month in March (beating even the two-day Sep 2023 outlier) bringing the total to over 72 million #CRAN #Rstats packages shipped as #Ubuntu .deb binaries.

#r2u. Fast. Easy. Reliable. Pick All Three.

16 2 3 0

Finally making the move to Linux Mint*. Steam to play #GW2. Bottles to run the necessary Windows apps. Raspberry Pi NAS to replace OneDrive. #r2u** to help with packages in Positron.

Looking hopeful.

* rather than just playing with it
** shout out to @eddelbuettel.com

6 3 1 0
Dual-figure chart with #r2u cumulative donwload in mullion on the left, hitting 68 millions, and average weekly download (on a log scale) coming up on 100lk/week as we hit 3.49 million in the short month of February.

Dual-figure chart with #r2u cumulative donwload in mullion on the left, hitting 68 millions, and average weekly download (on a log scale) coming up on 100lk/week as we hit 3.49 million in the short month of February.

#r2u had its first 'three point five million packages delivered' month in Februaru (outside that two-day Sep 2023 outlier) bringing the total to over 68 million #CRAN #Rstats packages shipped as #Ubuntu .deb binaries.

#r2u. Fast. Easy. Reliable. Pick All Three.

18 2 0 0

It is six of one versus half a dozen of another but I think I would just have stuck the functionality into your `b64` package. But hey another test of the improved `rustc` and `cargo` builds in #r2u so all good. (BTW you have a 'rustc (>= 1.65)' there and I suspect the condition is too weak)

1 0 0 0

#r2u PSA: I added a 'rustup' invocation to the build containers so (current) `cargo` and `rustc` are present. If your Rust-using #Rstats package is on CRAN, it will be in #r2u as a (completely dependency-resolved) .deb binary that installs in seconds.

#r2u. Fast. Easy. Reliable. Pick All Three.

7 1 0 0

Something that never get's old: today's edition also updates `quarto` and `rstudio` (from my 'personal' PPA) along with 40+ CRAN packages for Ubuntu thanks to #r2u: Fast. Easy. Reliable. Binaries with full and complete dependencies.

2 0 0 1
Screenshot of running 'wajig new' which compares the most recent `apt` package index with the previous one, and thereby identifies new packages.  As the machine was update yesterday, the change is just the new CRAN packages provided by r2u.

Screenshot of running 'wajig new' which compares the most recent `apt` package index with the previous one, and thereby identifies new packages. As the machine was update yesterday, the change is just the new CRAN packages provided by r2u.

Something that never gets old: running `wajig new` to see what `apt` get-able packages have been added since the previous `apt` run -- this time once more only #r2u packages bring #Rstats CRAN repo as fully-formed binaries with _all_ dependencies to Ubuntu.

5 0 0 1

I have not yet looked at `rv` but we cover in the FAQ for #r2u what works / does not work with e.g. `renv`. In short, the way the apt repo is _currently_ built does not support dated snapshot but that is _not_ an apt limitation. With sufficient resources (I am a one-man team here) this could work.

0 0 0 0

Have you seen #r2u? I have written a few posts (some with video) showing e.g.

docker run --rm -ti rocker/r2u

followed by

install.packages(c("tidyverse", "devtools"))

(to take your example) takes care of all system packages along with the R packages via the apt integration. Try it!
#rstats

5 0 1 0

Have you considered having each of the (2 x N) reverse-dependency runs be based on #r2u to skip the (possibly long-tailed) dependency installation?

0 0 1 0
Screenshot of the simple log counter showing a quick 60k downloads in a good hour today.  Count now at 85k, or about the daily average, at 10:00h.

Screenshot of the simple log counter showing a quick 60k downloads in a good hour today. Count now at 85k, or about the daily average, at 10:00h.

Always nice to see someone put the petal to the metal for #r2u: a quick 10x in use pulling over 60k #rstats CRAN binaries for Ubuntu in about an hour.

#r2u: Fast. Easy. Reliable. Pick All Three.

8 0 1 0

Use `ccache` as the duckdb docs suggests (also via `ninja`). I wrote about that on my blog too.

Even faster: don't compile. On #Ubuntu with #r2u:

> system.time(install.packages("duckdb"))
[...]
user system elapsed
3.506 0.664 5.532
>

as five and a half seconds is hard to beat.

2 0 0 0

Love to hear it! Gratyfying that a crazy 'let me try this real quick' weekend idea 3 1/2 years ago is helping people!

And this extends to wherever one runs 'cloud' or lambda tasks, r2u gives a fully-featured R + CRAN instance it seconds. Raw power!

#r2u: Fast. Easy. Reliable. Pick all three.

4 0 0 0
Screenshot of blog post at https://dirk.eddelbuettel.com/blog/2025/11/19#055_more_frequent_r2u describing post #055 in the r^4 series about more frequent and automated r2u builds

Screenshot of blog post at https://dirk.eddelbuettel.com/blog/2025/11/19#055_more_frequent_r2u describing post #055 in the r^4 series about more frequent and automated r2u builds

r^4 post #055: More Frequent r2u Updates
r2u builds fully at GitHub Actions, multiple runs per day
dirk.eddelbuettel.com/blog/2025/11...
#rstats #r2u

Internal updates allowing for more frequent builds of CRAN binaries for Ubuntu

6 0 0 0

#r2u for #Rstats now also includes #BioConductor on arm64 and not only on standard amd64. Following the upgrade to #BioConductor 3.22 we added this hardware platform.

`zlibbioc` was removed so a few packages we have for amd64 could not be built. Should I backfill it (and them)?

2 0 0 0
CRAN as Ubuntu Binaries - r2u Easy, fast, reliable -- pick all three!

Are you compiling from source ie on Unix? If it is Ubuntu, consider #r2u which has all of CRAN (and good chunks of BioConductor) as binaries for #rstats so something like `install.packages("tidyverse")` installs everything in a few seconds -- and all dependencies.

eddelbuettel.github.io/r2u

1 0 1 0
CRAN as Ubuntu Binaries - r2u Easy, fast, reliable -- pick all three!

#r2u now contains BioConductor 3.22 released yesterday for Ubuntu 24.04 and 22.04 (for the 500-some packages we cover, details at the site). #Rstats 4.5.2 has been built, we will update the containers in a few days as well.

#r2u. Fast. Easy. Reliable. Pick All Three.

eddelbuettel.github.io/r2u

9 2 0 0

#r2u is back with 162 (jammy) and 202 (noble, for arm64 and amd64) package "kittens" (well: .deb files) catching up for time lost since the last update Friday. Big thanks to the CRAN admins and the @posit.co folks for unplugging their piping, and to @glin.github.io for the heads-up. #rstats

9 0 0 0
Photo of a _very_ sad puppy waiting for new CRAN packages as Ubuntu binaries.

Photo of a _very_ sad puppy waiting for new CRAN packages as Ubuntu binaries.

No p3m.dev updates, so no #r2u updates as we follow its pulse. Archive sync issue between CRAN and posit said to be at the root of this. Lots of sad puppies to go around. Whoever is looking into it (@posit.co is not listing a contact person, maybe they should): thank you. We will be back. #rstats

6 0 2 1
Dual chart with #r2u cumulative downloads in millions on the left, and average weekly downloads (on a log scale) on the right.

Dual chart with #r2u cumulative downloads in millions on the left, and average weekly downloads (on a log scale) on the right.

#r2u had its first 'three million packages delivered' month in September (outside that two-day Sep 2023 outlier) bringing the total to over 54 million #CRAN #Rstats packages shipped as #Ubuntu .deb binaries.

#r2u. Fast. Easy. Reliable. Pick All Three.

14 1 0 0
Dual-graphic of (on the left) cumulative downloads now in excess of 50 million and (on the right) average weekly download (shown on a log scale) continuing to grow.

Dual-graphic of (on the left) cumulative downloads now in excess of 50 million and (on the right) average weekly download (shown on a log scale) continuing to grow.

#r2u update: Now 50 million .deb packages of #rstats CRAN packages served as of end of August.

#r2u: Fast. Easy. Reliable. Pick All Three.

More at eddelbuettel.github.io/r2u

8 0 0 0
Video

#r2u: Fast. Easy. Reliable. Pick All Three.

New (quick) demo (on new machine with fast(er) disk): Install the whole of the `tidyverse` in _nine seconds_ including all dependencies thanks to #r2u. #rstats

More as usual at eddelbuettel.github.io/r2u

19 4 0 0

Yup. If you have a reproducible example we can surely cook up a demo as posted over the years as gif/mp4 files. Eg 'tidyverse in 18 secs' or 'tidyverse + sf + rstan in 30-some secs'. Have a new (faster) machine so tidyverse is now 7s for me ๐Ÿคฉ.

#r2u: Fast. Easy. Reliable. Pick All Three.

#rstats

6 1 0 0

I second this recommendation! Added bonus of magically summoning @eddelbuettel.com by using #r2u

4 0 1 0

There are always alternatives. On a yuge (48 cores, 0.5tb ram) server for students the root/sudo account uses bspm + r2u for common packages, reliably + fast + easily. Students use non-bspm installation below $HOME by default. Best of both worlds.

Also, should your team use Docker? #r2u #rstats

3 0 0 0

Everything has upsides and downsides. Manylinux works more broadly but has the virtual env problem of littering your machine with library copies that go stale. When you can integrate with the system (as r2u does on Ubuntu) nothing beats that approach.

#r2u: Fast. Easy. Reliable. Pick all three.

3 0 2 0

I remember now about #r2u. That looks good indeed.

Is there a way to make it work with `renv` or another package management system? (I looked at a couple Github issues and I think the answer is no, but maybe it's changed)

0 0 1 0