Back when Perl was the language of choice for bioinformatics, I found a huge performance boost pre-processing large (~1Tb) text files using built in unix tools like sed and awk with regex. So while it might take me a full hour to peck out the correct incantation, the task would then run in an hour, compared to four hours or more for the same task using Perl.
So many pipes…
Keep sharing the tapes.