It’s like this one time, I was using some translation software to extract, translate, and reflow text from a number of HTML files, and they were encoded in ANSI format, which, as any localization specialist will tell you, is just a bad way to go about it, as it doesn’t support non-Roman alphabets.

At any rate, the extracted text was bound for translation into Russian, which went smoothly enough (besides going overbudget, but that’s another story). But then, at the moment when I had to take the Russian translation and reflow it back into HTML format, the encoding got completely lost, and the files ended up looking like this:

??????????????????????????/?????????
??????

??
?????????????????????????????????????????????????????????????????????????????????

etc, etc.

So what could I do?

1) I tried going back to the original English files and resaving them in Unicode format, and then trying to reflow the text, but the reflow result was the same.

2) I tried resaving the original files in UTF8 format, which is the “specialty” of the localization software that I use, but to no avail.

3) In the end, I had to save the HTML file in UTF8, go into the source code itself, and find the English segments that I knew had been translated, delete them, and then use the Word document that contained the Russian translations, and cut and paste the Russian segments into the HTML source code, which, needless to say, was difficult, on account of I know precisely seven words of Russian, and I can’t spell a single one of them using the Cyrillic alphabet.

This probably has very little to do with your post, but I just wanted to share in the headache of technical difficulties that, in the end, have a simple solution (in this case, my localization software has a switch you can throw to reflow translations into HTML files and automatically save them in UTF8). And to say that cutting and pasting is a bitch.

Live long and prosper.