CRAN submission planned for early 2026

[dankelley]

We've had lots of new things since the last CRAN submission (see below), and I like having a new update for the new school year. Given that it can take a week or so for all the CRAN remote builds (even if there are no problems), this suggests we ought to release in the first week or two of August.

I don't think there are any easily-solved and generally-important issues open at the moment. ("Easily-solved" means an issue that might be solved with a day of coding. The map-projection problems are not in that category. "Generally-Important" means an issue that is likely to affect many users, e.g. a problem with SBE or similar CTD files. The detailed AD2CP issues are not in that category.)

How do @richardsc and @clayton33 feel about a release? Is there an issue (or two or three) that I ought to work on before doing that? Note that I'm not asking for a vote at this moment, but more for comments about the issues that might block a release. (I may be able to address several of them, perhaps with help on new datasets.)

Below is the changelog since the last CRAN release:

Details

oce 1.8.4

• Add bodcNames2oceNames() to rename NERC/BODC variables (issue should we try to decode CIOOS-style variable names in NetCDF files? #2235).
• Add D4902337_219.nc built-in Argo profile data file (issue package should include a sample Argo file #2272). (issue
  add cubeHelix colourscale #2331).
• Add netcdfTOC() to overview NetCDF files (issue consider adding way to scan netcdf contents #2232).
• Add oceColorsCubeHelix(), a perceptual colour scheme used in Astronomy
• Add oceRename() to rename variables, flags and units (issue maybe as.oce() should get a 'renamer' parameter #2238).
• Add read.xbt.edf2() to read a tab-separated file type.
• Add read.xbt.noaa2() to handle UBT format (issue read.xbt() should handle UBT format #2289).
• Add Tidal Analysis vignette (issue Add a Tidal Analysis vignette #2243).
• Change [[ to handle "cabbeling", for objects holding sufficient
  information for this computation (issue [["cabbeling"]] should return as from gsw::cabbeling() #2338).
• Change adpRdiFileTrim() to reorder parameters, and to accept indices
  parameter (issue adpRdiFileTrim() should get argument to select data chunks #2330)
• Change amsr[[ to handle "metadata" parameter (issue amsr[["metadata"]] does not work #2229).
• Change argo2ctd() (issue as.ctd on an argo object should create a metadata item that identifies the float ID and profile number? #2270).
• Change as.unit() to handle more units (issue as.unit() not handling "S m-1" properly #2240).
• Change as.ctd() to handle Argo data better (issue as.ctd on an argo object should create a metadata item that identifies the float ID and profile number? #2270).
• Change binApply1D() to
  • handle the ... argument (issue binApply1D is not passing '...' to 'FUN' #2265) and
  • handle time zones better (issue binApply1D() odd results if x and xbreaks are times with different zones (and if the interval includes a daylight-savings event) #2266).
• Change concatenate() to handle lists properly (issue concatenate() should accept a list as its first parameter #2278).
• Change download.amsr() to handle realtime data files (issue download.amsr() should handle realtime data files #2231).
• Change imagep() to handle decimate as set in ~/.Rprofile (issue Make "decimate" for imagep() an option that can be set #2263).
• Change lonlat2map() to handle a [coastline-class] object as first parameter
  (issue lonlat2map() ought to accept a coastline object as first parameter #2284).
• Change magneticField() to handle the latest (generation 14) IGRF
  formulation (issue consider adding new world-magnetic model #2274).
• Change mapScalebar() to
  • handle x="top" (issue mapScalebar() should permit x="top" etc #2283) and
  • handle length under 1km (issue mapScalebar() should use decimal numbers for tiny scales #2287).
• Change mapPlot() to handle las as a parameter (issue mapPlot() should offer more control of axis labels #1707).
• Change oceRename() to default to a built-in IOOS dictionary (issue rename() default dict should be "ioos", not "ioos.csv" #2311).
• Change oce.plot.ts() to
  • handle grid better (issue oce.plot.ts() does not put grid at right spots #2267),
  • handle simplify better (issue oce.plot.ts() fails with "need finite 'ylim' values" even though plot() works #2277) and
  • handle xlab better (issue oce.plot.ts() ignores 'xlab' parameter #2285).
• Change plot.section() to handle eos="gsw" (issue plot,section-method() should be able to plot GSW variables #2256).
• Change pwelch() to fix error if nfft and spec are provided (issue
  pwelch with a spec= argument given fails in an argument dependency loop #2299).
• Change read.amsr() to handle realtime data files (issue perhaps read.amsr() should be able to read realtime data #2230).
• Change read.adp.ad2cp() to
  • handle TOC, datasets and plans better (issue read.adp.ad2cp() fails on a user's file #2303 and related),
  • read temperature correctly (issue ad2cp temperatures look to be wrong (possible signed vs unsigned int issue) #2324),
  • save elements of AST (etc) individually, not within a list (issue read.adp.ad2cp() should not use lists for altimeter, AST, altimeterRaw and AHRS #2318)
    and
  • scale altimeterRawSamples by 1/2^15 (issue read.adp.ad2cp() should multiply altimeterRaw data by 2^15 #2327).
  • set data@distance only if NC (number of cells) is nonzero (issue read.adp.ad2cp() defining data$distance oddly (in some cases) #2319).
  • use number of raw altimeter samples from header, if it disagrees with value
    in record (issue ad2cp dataType="averageAltimeter" not being read correctly #2326).
• Change read.adp.rdi(..., which="??") to return a data frame that includes
  ensemble size (issue read.adp.rdi(file, which="??") should return also the chunk size #2329).
• Change read.ctd.sbe() to
  • handle files that lack salinity and conductivity (issues read.ctd.sbe() fails if no conductivity or salinity data present #2279 and Should read.ctd.sbe() have a parameter to select warning or error for missing S/C? #2282)
    and
  • handle time-related header elements better (issues read.ctd.sbe() problem with '* sample rate = 1 scan every 0.5 seconds *' #2280 and read.ctd.sbe() warns for a time given as e.g. 107 #2281).
  • handle "new" format CNV files, at least provisionally (issue Issue uploading cnv files to OCE #2328).
• Change read.ctd.aml() to handle "type 3" format (issue Unable to import CTD files from an AML Base.X2 downloaded with Sailfish 1.4.8.0 #2247).
• Change read.coastline() to handle file in "shapefile" format (issue
  mapPlot() should handle coastlines in SFG polygon format #1850).
• Change read.met() to handle flags better (issue problem with read.met() flags or concatenate() actions #2298).
• Change read.netcdf() to
  • document flag handling better (issue read.netcdf() docs should discuss flags #2236),
  • handle units better and
  • permit renaming data (issue should we try to decode CIOOS-style variable names in NetCDF files? #2235).
• Change read.rsk() to
  • solve a column renaming problem (issue read.rsk() misidentifying some data columns #2291),
  • handle files that have empty (but not missing) geodata tables (issue
    rsk files with empty geodata tables #2336), and
  • handle more unit formats (issue rsk units 'counts' and 'µmol/L' not recognized #2337).
• Change sectionSmooth() to disallow method="kriging", because it relied on
  the automap package, which was removed from CRAN on 2025-06-30 (issue
  problem with automap #2332).
• Change subset,amsr-method() to handle "pass" parameter.
• Change subset,ctd-method() to handle items of unexpected length (issue
  subset,ctd-method() inserts NA values for short elements #2250).
• Change tidem() to stop if there are any non-finite time values (issue
  tidem: uninformative error if data have NAs #2269).
• Change vectorShow() to obey "digits" parameter for list arguments (issue
  vectorShow() ignores argument 'n' for list objects #2313).

[dankelley]

It looks like a fairly high probability of a labour disruption at Dalhousie. Unfortunately, oce names my Dalhousie email address, and that will not work during a disruption. There is a way to switch emails for a package, but I think that requires emails from both the old (Dal) and new (gmail, probably) addresses to be examined by members of the CRAN team. I don't think we have time for that and I don't want things to be stuck on CRAN.

Therefore I'll revisit the idea in a week or two, when I think we'll know if there will be a labour disruption. (This also avoids the week or 10-day holiday that I think CRAN members take in August.)

September seems like a better plan. But we'll see.

[dankelley]

I changed the title because we are past "early". Rather than say "September", I'm now just saying "autumn". I cannot submit a new version until my dal.ca email account works again, after the university ends the lock-out of professors.

[dankelley]

I moved the target time from the original.