Kirill Müller
Kirill Müller
We have a workaround here. What's the proposed course of action? How much can the C core do to reduce the number of calls?
To bring this up with Posit/RStudio, a reproducible example would be nice.
Can we, as a rule, offload all computation to another thread?
Oh, that's a neat idea. @cboettig: also something for duckdbfs?
Thanks. I don't think it's a one-time action, at least in the medium term. If cpp11 gets nice new features, we want to use them.
Let's wait for https://github.com/r-lib/cpp11/pull/353. Also, the upside of no vendoring means that there's less maintenance effort at the (fairly low) risk of breakages from upstream.
Thanks. The underlying function is `vctrs::vec_rbind()`, this behavior is consistent with vctrs principles.
Thanks. Is the following helpful: https://tibble.tidyverse.org/articles/numbers.html (linked from https://pillar.r-lib.org/articles/numbers.html) ? Is the behavior different from the documentation? Also: ``` r round(100.5) #> [1] 100 ``` Created on 2024-10-17 with [reprex...
Have you seen #1296? See https://github.com/igraph/rigraph/blame/6eeb6e4a575a8c0f718f0606cfc8805932b9e604/tools/update-cigraph.sh#L58 for where the line in question was added. True, this should have been a comment in the code.
Thanks. So, now we can compile, but tests fail? Can we identify which, and then perhaps create a checklist?