-
Notifications
You must be signed in to change notification settings - Fork 7
Release 1.1.60 #81
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Release 1.1.60 #81
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, thanks a lot!
And still compiles and passes the checks on my Mac.
Yes I have of course also been running So ok to pack it up and ship it to CRAN ? |
I think so, looks good to me! |
I need one more check. Eye-balling the DESCRIPTION file I just noticed that it e.g. dropped Jeroen as 'ctb'. I must have flattened an earlier commit away which is moderately worrisome. One sec. |
Ok, I appear to have 'lost' #79 but that was just a single commit which I cherry-picked back and pushed. I did eyeball all the changes of the other PRs and they are there. So my bad for bumbling one micro-change, and lucky me I caught it... |
And we fell into a Kurt Hornik-shaped hole:
@kurthornik I know of course it is Sunday but the above does not lead to any downloadable logs. It ends for me in
and I don't have 00install.out. The (now enlarged) team of five of us has been banging at his for a few month now and I do not think we have seen a single build failure. I will fire up a Debian container now, on Ubuntu (23.10, 22.04 at CI and for others) it "just works" ... |
Installs like a champ in
|
Which ends with
They seem to be using llvm=17 with flang-new-17. I don't know anything about fortran but it looks like fdate is a GNU extension? |
Hm, where is that linked from / how did I miss that after all those years and being familiar with the mail format? Bug agreed. Line before the block you quoted end has |
I think the |
Thanks, @jeroen. I tried to make sense of things in a container, current to Debian testing just like the Vienna setup, but to no avail. @kurthornik If you can find a few minutes, please give the package a good look. It would us all ahead a good step from the current practice relying on binaries. If and when you keep the default Fortran compiler in Debian testing (or Ubuntu LTS or newer) it 'just works'. Would we nice if we could opt into your standard compiler as opposed to the bleeding edge one. |
Added workaround here: #82 |
Also re-adds Date which I find helpful in preparing releases as it provides and 'at source' timestamp.
1d9bc38
to
f5e5a65
Compare
Merging now as we crossed the finish line:
Big big thanks to everybody. This package is now in much better shape, and we solved a buildability puzzle or two as well. |
This PR prepares 1.1.60.
I made a final change simplifying the one tests (which we only need to run once) but also adding a reference output of 'success' so that R will fail a check if this fails to match. It uses the old-old trick of storing the test output as a file ending in
.Rout.save
.And I made a quick attempt at backfilling the ChangeLog. Improvements to that are of course welcome too.
I won't upload until a get a 'thumbs up' approval, and I suggest to not merge until this is on CRAN allowing for any last minute change the CRAN team may require. That pattern of 'holding' the new release PR has worked well for me in many other package contexts.