We get .csv files from companies we acquire to bring in their customers, their contacts and their software licenses (custom record). It may come from SalesForce, manually managed .xls files, Quicken, et al. Without fail, the import process will puke after selecting the file to import due to some number of lines (sometimes hundreds) that have characters it doesn't like for some reason. Often these look like normal American alphabet characters that I can actually fix by simply re-typing (copy/paste won't work). Other times, it's from the extended-ASCII character set. Spanish characters, German, French, other. It doesn't like those. They'll look totally "normal" in Notepad or Excel, but NS pukes. Takes forever to clean this, which usually has to involve changing something like n to n, which is an entirely different character.
I've tried different encodings, saving to Notepad as text then back to CSV. What should take maybe 20 minutes to prepare has become several hours and results in inaccurate data.
Currently, Netsuite Support has no reliable solution. They even seem surprised we're having the trouble. I've had these issues since first using NS 5 years ago. At my previous company, it was small files with maybe 500 lines. Here, I'm importing sometimes 75,000 lines. Other software seems to have no trouble.