Closed IndrajeetPatil closed 4 months ago
Sorry, I don't see why we should increase the minimum version dependency unless we actually use something from that version?
If some user can't upgrade {withr}, why should we prevent them from using {lintr} unless there's something specific in the new version that we "need" for {lintr}?
I must be missing something...
If some user can't upgrade {withr}, why should we prevent them from using {lintr}
But it's not a runtime dependency of {lintr}. Why would it prevent them from using {lintr}?
At any rate, this is not critical. withr 3.0.0 was quite an extensive rewrite, and I would have liked for us to see some of its benefits in our dev workflow (performance, better messages, etc.).
Maybe working on https://github.com/r-lib/lintr/issues/2539 will help motivate this further.
But it's not a runtime dependency of {lintr}. Why would it prevent them from using {lintr}?
True, but all the more reason not to enforce a minimum version IMO.
If there's some actual code changes you want to make that use updated functionality, I am happy to proceed.
To benefit from performance and other enhancements in this release: https://www.tidyverse.org/blog/2024/01/withr-3-0-0/
We have already fixed issues stemming from this version: https://github.com/r-lib/lintr/pull/2519
We didn't bump the version because it was quite new at the time, but we can do so now.