make.contigs sometimes puts all reads in scrap

Running make.contigs with fastq data from MiSeq, I sometimes find that it has put all reads into *.scrap.contigs.fasta.
Re-running the same command can result in the opposite: everything in *.trim.contigs.fasta

This is mothur v.1.28.0, running on Mac OSX 10.6.8.
I ran make.contigs for a bunch of files in batch mode, then went back in interactive mode to re-run samples for which everything had been put in scrap.

I’m not using any of the optional commands. i.e.,
make.contigs(ffastq=PAL2_E8_CAGAGAGG-GTAAGGAG_L001_R1_001.fastq, rfastq=PAL2_E8_CAGAGAGG-GTAAGGAG_L001_R2_001.fastq, processors=2)

Thanks for bringing this to our attention. It is a bug caused by the tdiffs parameter not being initialized properly. It will be fixed in our next release. In the meantime, if you give tdiffs a value it should work properly.

make.contigs(ffastq=PAL2_E8_CAGAGAGG-GTAAGGAG_L001_R1_001.fastq, rfastq=PAL2_E8_CAGAGAGG-GTAAGGAG_L001_R2_001.fastq, processors=2, tdiffs=0)