Locale settings

[fanli]

Hi all,

What locale(s) are recommended for lab servers? We primarily work in applied genomics (think off-the-shelf programs, scripting, R).

I just ran into the whole locale issue when sorting a pair of FASTQ files by read ID. Apparently my default "en_US.UTF-8" locale ignores colons and other punctuation. So I switch to "C", and if so, will this have any unintended consequences?

Thanks!

[maubp]

I believe GATK and samtools sort read names differently anyway - but that's a good point about the locale having an input.

[fanli]

Yeah, most programs (should) do a byte-by-byte comparison but unix sort is affected by locale:

*** WARNING *** The locale specified by the environment affects sort order. Set LC_ALL=C to get the traditional sort order that uses native byte values.

[dpryan]

I would stick to LC_ALL=C for the reasons mentioned by fanli. Half the time that people report problems with deepTools to us it's because of this.