Sort lines

Because it's comparing text character by character, and the first character of 10 is 1, which comes before 2. So 1, 10, 100, 2, 20, 3 is correct alphabetical order, and it's almost never what anyone wants for numbers.

The same reason explains why item2 lands after item10.

Working around it

Pad the numbers. 01, 02, 10 sorts correctly as text, because every entry is the same width. If you're generating the list, padding at the source fixes it permanently.

Put the number last in the line, if the text before it is identical. Sorting then compares the text first and the numbers only where the text matches, which still hits the same problem but on a smaller scale.

Sort numerically elsewhere. A spreadsheet or a script that understands numbers is the right tool for a list that's genuinely numeric.

Natural sorting, which reads runs of digits as numbers, is what file managers do when they show file2 before file10. It's the intuitive behaviour and it's not what plain alphabetical sorting does.

Capitals

In the raw character order that computers use, every capital letter comes before every lowercase one. So Zebra sorts before apple, which looks broken and is technically correct.

Case-insensitive sorting treats apple and Apple as equal and gives the order people expect. That's the sensible default for anything a person will read.

Case-sensitive sorting still matters when the list is code, filenames on a case-sensitive system, or anything where ID and id are genuinely different things.

Accents and other alphabets

Sorting by raw character value puts accented characters after all the unaccented ones, so zebra comes before Ångström, and Cyrillic or Greek text lands in its own block regardless of meaning.

Proper alphabetical order is language-specific and genuinely disagrees between languages. In Swedish, å, ä and ö come after z. In German they sort with a, a and o. In Spanish, ñ has its own place after n. There is no single correct order, only the correct order for a particular language.

For a list going in front of readers in one language, sorting with that language's rules is what you want, and that's a job for a spreadsheet or a locale-aware tool rather than a general text sorter.

Blank lines and whitespace

Blank lines sort to the top, because an empty string comes before everything. Removing them first usually gives a cleaner result.

Leading spaces are compared like any other character, so a line beginning with a space sorts above one that doesn't. Text pasted from a document often carries them invisibly, which is why a list can sort into two apparent groups for no visible reason.

Reversing, and what it isn't

Reversing sorts descending. It's not the same as reversing the order of the lines as they were, which keeps the original sequence and flips it.

Both are useful and they're different operations. If you want the last line first, you want the second one.

After sorting

Sorting brings duplicates next to each other, which is what makes removing them reliable. Doing it in that order is the usual reason to sort a list at all.

Everything happens as you type, in your browser. Nothing is uploaded, which matters when the list is names, emails or anything else you'd rather not paste into a server.