Debian-only R-devel fail in CRAN incoming checks

[mattdowle]

Dear maintainer,

package data.table_1.11.8.tar.gz does not pass the incoming checks automatically, please see the following pre-tests:
Windows: https://win-builder.r-project.org/incoming_pretest/data.table_1.11.8_20180927_133414/Windows/00check.log
Status: OK
Debian: https://win-builder.r-project.org/incoming_pretest/data.table_1.11.8_20180927_133414/Debian/00check.log
Status: 2 ERRORs, 1 WARNING

Last released version's CRAN status: NOTE: 1, OK: 11
See: https://CRAN.R-project.org/web/checks/check_results_data.table.html

CRAN Web: https://cran.r-project.org/package=data.table

Please fix all problems and resubmit a fixed version via the webform.
If you are not sure how to fix the problems shown, please ask for help on the R-package-devel mailing list:
https://stat.ethz.ch/mailman/listinfo/r-package-devel
If you are fairly certain the rejection is a false positive, please reply-all to this message and explain.

More details are given in the directory:
https://win-builder.r-project.org/incoming_pretest/data.table_1.11.8_20180927_133414/
The files will be removed after roughly 7 days.

*** Strong rev. depends ***: Ac3net ActisoftR AdhereR AF AFM AGread aLFQ alpaca antaresProcessing antaresRead antaresViz APSIM ARTP2 asciiSetupReader assertable BatchExperiments BatchJobs batchtools bea.R bedr behavr benford.analysis bigreadr biomartr bold bomrang BootstrapQTL brainGraph brandwatchR bridger2 BTYDplus bupaR BuyseTest c3 caretEnsemble casebase CaseBasedReasoning CatEncoders Causata ccafs cdparcoord Census2016 cffdrs ChannelAttributionApp chicane chromer classifierplots cleanEHR clickstream CoDiNA CompareCausalNetworks constellation CorporaCoCo corpustools corregp csvy ctmm cvAUC D3partitionR damr dartR dat data360r DataCombine datadr DataExplorer dataPreparation dataview dblr DCD detectRUNS dfmeta DGM dgo Diderot difconet DiffNet doremi drgee droptest dtplyr Eagle easycsv eclust ecoengine edarf edeaR edgeRun EdSurvey eeptools eiCompare electionsBR emil EmissV Epi eplusr EthSEQ etm EurosarcBayes EventStudy evoper expss extracat FAOSTAT fastDummies fastLink fbRads FedDat
a finch flippant freqweights fulltext funchir futureheatwaves gasfluxes gbp gcite gems genBart genderizeR genomic.autocorr GenomicTools GenomicTools.fileHandler GeoMongo getCRUCLdata getlandsat getProxy getTBinR GFE gfer ggdistribute ggdmc ggetho GGIR ggpval ggTimeSeries githubinstall gofastr grattan greport gridsample gunsales haploReconstruct happybiRthday healthcareai heatwaveR heims hesim hisse Hmisc hoardeR HRM httk hurricaneexposure hutils hyfo hypervolume IDmining ie2misc iemisc iml inbreedR IncDTW Information InformationValue IntClust inTrees ipeaData IPtoCountry iRF isdparser ivtools joinXL jsTree jubilee JuniperKernel JWileymisc kangar00 koRpus LabourMarketAreas lar ldamatch lexicon LexisNexisTools lidR live livechatR lllcrc logiBin LogisticDx lookupTable LSPFP Luminescence maditr mapr margins mau MetaComp MetamapsDB metaMix mgcViz microdemic miLineage miscset mknapsack MlBayesOpt mlr mlrMBO mltools mmpf ModelMetrics MODIStsp mregions mrMLM mrMLM.GUI mschart msmtools mtcon
nectR multdyn multicastR MultiFit multistateutils musica mvQuad natserv ndjson net.security NetLogoR networkR ngstk nneo NNS noaastormevents nodbi noncompliance ompr openEBGM OpenML openSTARS optiRum optiSel orgR origami originr otvPlots PakPMICS2014Ch PakPMICS2014HH PakPMICS2014HL PakPMICS2014Wm panelaggregation partialAR partialCI partools patternator pcadapt pcrsim PeakSegJoint penaltyLearning PGRdup PhenotypeSimulator pinnacle.API pivottabler pkggraph pkgnet PKPDmisc plotly pointdensityP poisFErobust politeness polmineR popEpi PopGenReport precrec prediction PReMiuM prioritizr prisonbrief processmapR ProFound prrd PSF psidR psychmeta PTXQC Publish pxweb qdapTools QTL.gCIMapping QTL.gCIMapping.GUI quanteda QuantTools quickPlot R2DT r2glmm RAM rAmCharts randomForestExplainer randomNames rangeMapper RAPIDR ratios rBayesianOptimization rbi Rbitcoin RBMRB rbraries rbtt RcextTools rcites RcppGreedySetCover Rcrawler Rdice RDML rdpla rdryad readtext RecordLinkage REDCapR REIDS Reinforce
mentLearning reinsureR rENA REndo repmis reproducible rerddap rfacebookstat rfisheries rFTRLProximal rgbif rGoodData rif Rilostat rio rIsing riskRegression ritis rlas rlist RLumShiny RndTexExams Rnets RNewsflow Rnightlights rnpn robCompositions Robocoap rODE rolypoly rorcid rPackedBar rprev rpubchem rqdatatable RSauceLabs rsnps rsunlight rtimes RWDataPlyr RWildbook RxnSim RZooRoH SafeQuant sbpiper scidb scopr scorecard scraEP sdcMicro sdcTable SDLfilter SEA segregation SELF sentimentr sentometrics SetRank SGP SGPdata ShinyItemAnalysis simPH simPop simstudy SIRItoGTFS sitree skm skynet slim SmartEDA socialmixr solarius sommer spacyr SpaDES.core SpaDES.tools sparseFLMM SpATS spatsoc speaq spectralAnalysis spellcheckr splitstackshape spocc sport squid sstModel staRdom starmie statar StatMeasures stdReg stm stranger strataG strvalidator stUPscales subscreen surveyplanning survMisc SWMPr tableMatrix takos taskscheduleR tatoo taxize taxonomizr tcpl tcR teachingApps TeXCheckR text2vec text
clean textrank textreadr textshape textTinyR TideCurves TideTables timeseriesdb tosca totalcensus toxboot TrafficBDE traits tree.bins treemap tRophicPosition tsbox TSstudio tstools TSTr ttwa twl TwoRegression uavRmp udpipe ukbtools UKgrid uptasticsearch utilsIPEA validateRS vardpoor vein VIM vortexR vosonSML votesys wally WebGestaltR wikitaxa wilson windfarmGA winRatioAnalysis word.alignment worrms wTO xesreadR xgboost xlm xROI zeitgebr zFactor

Best regards,
CRAN teams' auto-check service

[mattdowle]

Hm. Seems to be Debian-only R-devel. I've just installed and built yesterday's R-devel and I can't reproduce.

R Under development (unstable) (2018-09-26 r75369)

That's the exact same commit as both CRAN's Windows and Debian. Why Debian is different beats me.

[mattdowle]

I replied :

---

Hi CRAN team,

I am baffled by this. It appears to be a Debian-only R-devel problem ?
I clicked the links in this email and the versions of R-devel are :
Windows: using R Under development (unstable) (2018-09-26 r75369)
Debian: using R Under development (unstable) (2018-09-26 r75369)

This seems to be precisely the same commit. Why would there be a difference in method registration between Debian and Windows? I have indeed added the S3Method deferred registration as Kurt requested, so there has been a change in data.table in this area at your request. But why Debian is different beats me as that's just an R level change and it passes fine on Windows R-devel.

I have refetched daily R-devel snapshot from yesterday, rebuilt fresh and retested. I can't reproduce here ... passes fully OK. I'm Linux which is similar to Debian, so it doesn't seem like an OS difference to me. All I can think is that it's something to do with the packages available in the library on the Debian machine. Scrolling all the way to end of the log file from Debian, it shows package rmarkdown is not available. Could it be that data.table didn't fully install properly on that machine, because it found rmarkdown was missing towards the end and the installation stopped early? And now that S3Method is deferred, this somehow makes a difference?

Here's my local R-devel matching the commit you used, and it passes fine here.

> sessionInfo()
R Under development (unstable) (2018-09-26 r75369)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.1 LTS

Matrix products: default
BLAS: /home/mdowle/build/R-devel/lib/libRblas.so
LAPACK: /home/mdowle/build/R-devel/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
[1] compiler_3.6.0
> 

Best, Matt

[mattdowle]

I followed up:

---

I took a guess that maybe reshape2 isn't installed on the Debian machine. So I remove.packages("reshape2") locally and tried Rdevel CMD check again. It still passes, just with a single note about reshape2 not being available to check. I'm out of ideas.

[mattdowle]

Maybe it's time to just to remove reshape2 suggests. It's the last one outstanding in #2740 which we wanted to do anyway. Jan more strongly than me at the time. His instinct was justified then it seems. Would have saved this hassle now had I agreed then.