Release v0.9.26

[mcol]

We want to make a small release to fix known regressions that are present in 0.9.25. As of now, the dev_0.9.26 branch already contains these commits from 0.9.25:

1da0100 Remove CRAN notes for good
8310844 Update .Rbuildignore
be82a86 Fix doi.org link in CITATION.
1aad6fa Remove spurious newline.
c55357d Post-release version bump.
ee736cb Merge pull request #240 from R-Lum/post-release-version-bump
0a4e193 Don't use the bbmle::profile() try_harder option in calc_MinDose().
6ff6a6f Skip test on windows CI.
82a5ada Remove the windows-2019 (oldrel) instance from the CI actions.
6509111 Add the expect_snapshot_RLum() test helper and start using it.
d920de9 Merge pull request #244 from R-Lum/snapshots_01
f06aadf plot_RLum.Data.Spectrum() + ad support for light arguments + fx cause of index warning + ad NEWS + up docu
833fe99 analyse_SAR.CWOSL(): + fx regression test that would cause an error because if different objects
54d62db Regression: the new internal functions for .throw_warning() and throw_error() were flushing the terminal in case of do.call() + fx calls + ad tests
57c549a Regression: the new internal functions for .throw_warning() and throw_error() were flushing the terminal in case of do.call() + fx calls + ad tests
be1602b Ådd NEWS to cherry pick and build news and manual
555e2b8 Regression: the new internal functions for .throw_warning() and throw_error() were flushing the terminal in case of do.call() + fx calls + ad tests
4a693e4 analyse_pIRIR_Sequence(): + ad better support for plotting (or not plotting) on too small devices + ad tests
fcc0115 Add more tests for internals
7b8f2c8 analyse_pIRIRSequende() + fx a few more plot issues in edge case scenarios
ea73669 Regression: the new internal functions for .throw_warning() and throw_error() were flushing the terminal in case of do.call() + fx calls + ad tests
dafcb73 analyse_pIRIR_Sequence(): + ad better support for plotting (or not plotting) on too small devices + ad tests
42e8ba7 Add more tests for internals
446b655 Add NEWS and update build number
41e15cd get_RLum() + fx behaviour of null.rm = TRUE; it was not removing empty list elements + ad tests + ad NEWS
3ab4064 Just a little bit houskeeping in get_RLum()
aa1d263 Merge branch 'dev_1.0.0'
21c56bb (origin/dev_0.9.26) write_BIN2R(): + fx non-ASCII character bug + fx open connection bug after crash + rf code polish + ad tests + ad NEWS + up manual

@RLumSK Is there anything else that you think should be included?

[mcol]

Perhaps #373?

[mcol]

I have moved the dev_0.9.26 branch up to 3b3e2bc, so that we we get fixes for #147, #245, #252, #254 and #258. The full list of commits added is the following:

68219e9 Correct the rejection.criteria data frame created by analyse_SAR.TL().
d70d78d get_RLum() subsetting (RLum.Analysis-method): + fx poor subsetting attempt + ad NEWS + ad tests + up manual
7860e6c Update analyse_SAR.TL.R
a3f1d08 Merge pull request #246 from R-Lum/issue_245
852d417 Code cleanup in analyse_SAR.TL().
cedfec1 Avoid computing an infinite ylim.
8499237 Don't nest the call to calc_TLLxTxRatio() inside one to get_RLum().
ff15efb Check that dose.points is compatible with the number of signals available.
a530df8 Handle the case of plot_GrowthCurve() returning NULL instead of an error.
00723c3 Rework the creation of RejectionCriteria in analyse_SAR.TL().
149b5e6 Update NEWS.
a714635 Increase the tolerance in a test that fails in CI.
a77eb40 Merge pull request #247 from R-Lum/issue_147
9336417 Silence notes raised by R CMD check.
053a1c5 Merge pull request #249 from R-Lum/issue_248
dc62ee7 Minor code reformating so that it shows in the outline in RStudio
8d435dd write_BIN2R(): + fx non-ASCII character bug + fx open connection bug after crash + rf code polish + ad tests + ad NEWS + up manual
dee4f4f Use add = TRUE with on.exit().
c9834f9 Merge pull request #255 from R-Lum/on.exit
fc99178 Add the infrastructure to keep a stack of function names.
bb061f4 Add calls to .set_function_name() to the functions that use the throwing functions.
906e2cc Use add = TRUE with on.exit().
6120173 Replace the implementation of .throw_error() and .throw_warning().
0950f02 Merge pull request #256 from R-Lum/issue_254
4d3c35f Set calc_Kars2008() as defunct.
15c8952 Merge pull request #257 from R-Lum/issue_252
12699c3 Rebuild NEWS and bump build number
27e2cb7 Minor typo corrections analyse_FadingMeasurement()
3b3e2bc calc_Huntely2006() cauterise #258 by reducing default n.MC until other solution is found.

I've also cherry-picked the following:

06d9cd0 Don't stop if the device size is too small when plot.single = FALSE.
de06d06 Increase coverage in analyse_pIRIRSequence.R.
4a75bb4 analyse_pIRIRSequence(): + fx minor regression issue with the plot warning + ad tests
a4b4fb1 Fix warning message from analyse_pIRIRSequence().
02d37ad analyse_pIRIRSequence(): + ad warning catcher
d18cc729f Fix #373 + fx the single that caused the issue + ad a dedicated test + ad NEWS + up NEWS o no manual update - regression

I've then brought NEWS up to date and reworded them a bit. I've also resurrected the 0.9.25 NEWS so that users that check an update get all recent ones, not just the last. It's otherwise very hard to see what has changed before the current release, only somebody with good github expertise could do it.

R CMD check passes.

[mcol]

Further steps taken:

• run make after setting DEVEL := 0 (see this change: f35b199)
• run RLumBuild::module_check_ReverseDependencies()
• run RLumBuild::module_write_codemetar()
• uploaded the package to the win-builder

I've been getting these messages during the check pass, but the links seem to be working in a browser, so I'm not sure if it's something to worry about:

   Found the following (possibly) invalid URLs:
     URL: https://gepris.dfg.de/gepris/projekt/250974974
       From: README.md
       Status: Error
       Message: libcurl error code 60:
         	SSL certificate problem: unable to get local issuer certificate
         	(Status without verification: OK)
     URL: https://gepris.dfg.de/gepris/projekt/46526743
       From: README.md
       Status: Error
       Message: libcurl error code 60:
         	SSL certificate problem: unable to get local issuer certificate
         	(Status without verification: OK)
     URL: https://gepris.dfg.de/gepris/projekt/505822867
       From: README.md
       Status: Error
       Message: libcurl error code 60:
         	SSL certificate problem: unable to get local issuer certificate
         	(Status without verification: OK)
     URL: https://gepris.dfg.de/gepris/projekt/528704761?language=en
       From: README.md
       Status: Error
       Message: libcurl error code 60:
         	SSL certificate problem: unable to get local issuer certificate
         	(Status without verification: OK)

Things still to do:

• update the date in DESCRIPTION
• commit all the changes produced as full rebuild
• submit to CRAN
• add the v0.9.26 tag
• add the website tag to the 0.9.26 branch to trigger the pkgdown.yaml workflow to build the package site
• add release to github
• update website with NEWS entry
• Brief e-mail to developers (at) r-luminescence.org

[RLumSK]

If the win builder does not flag it; it does not. We should not worry about it because this kind of certificate messages eventually go away.

[mcol]

All is done! It went quite smoothly overall, considering it was my first time handling the release!

[RLumSK]

@mcol Thanks a lot!