As a native speaker of the dialect spoken in the south east of England I
find "before" much clearer. Before Sunday means not later than Saturday
to me. I find "by" less clear, by Sunday could mean before Monday.
Michael
On 12/12/2025 15:40, Uwe Ligges wrote:
On 12.12.2025 16:33, Ben Bolker wrote:
I hesitate to prolong this part of the conversation, but ...
... when CRAN says "issues need fixing before [DATE]" (e.g. see
https://hadley.github.io/cran-deadlines/), is that technically "on or
before" or "strictly before"? Without worrying about time zones at
the moment, if my deadline is 2026-01-01 and I upload a new version of
my package sometime on 2026-01-01, am I OK? Or do I need to have
uploaded a version on the previous day?
"Before" is meant as "<".
In the past we had communication issues as is was not obvious for
everybody that British English speakers use "by" in the sense of "<".
I believed "before" was clear enough, but maybe we have to use a
mathematical expression?
Best,
Uwe Ligges
cheers
Ben Bolker
On 2025-12-11 6:38 p.m., Kenny, Christopher wrote:
Thank you both for your responses.
The deadline had not passed at the time the package was removed. It
was before the deadline at every single point on Earth. He removed
the package early. It’s an error, plain and simple.
I'm happy to share the email with the deadline if that would help.
Thank you also for reviewing the update to geomander.
(On the point of manual review, the package goes to newbies, but
based on the speed of the response, it may not take an entire human
review in the same way as other packages.)
Best,
Chris
________________________________________
From: Uwe Ligges <[email protected]>
Sent: Thursday, December 11, 2025 6:24 PM
To: Kenny, Christopher <[email protected]>; R Package
Development <[email protected]>
Subject: Re: [R-pkg-devel] Resubmitting Archived Packages
On 11.12.2025 20:07, Kenny, Christopher wrote:
Hello all,
I was looking for advice on best practices for resubmitting packages
that were archived.
This morning, Brian Ripley erroneously archived one of my packages
(geomander) prior to the deadline to fix a broken example. He then
archived its reverse dependencies a few minutes later.
Why erroneously? Your deadline passed.
A new deadline was set for the reverse dependencies, but as mail to the
maintainer bounced, that new deadline was not needed as the reverse
dependencies got archived for themselves as they are unmaintained.
Given that CRAN is closing for the winter soon, I would like to fix
this issue quickly. I have already resubmitted geomander this
morning, as I had originally planned. However, as the packages are
archived, each now must undergo a manual review, which takes more time.
Is it acceptable to submit the other downstream packages while the
first is awaiting manual checks, as long as they are submitted in
order of the dependency tree? As the original package had only a
broken example and the reverse dependencies were actively passing
checks, the review should be simple.
Please submit geomander and get it accepted first. The others will be
auto-archived otherwise.
Best,
Uwe Ligges
I would appreciate any advice or wisdom on how to reupload them
expeditiously.
Best,
Chris
christophertkenny.com - [email protected]
Christopher T. Kenny, PhD
Postdoctoral Research Associate, Data-Driven Social Science,
Princeton University
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
--
Michael Dewey
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
