Back in the day, when Ruby was almost inoperable with UTF8, you could basically just run strings through PHP's normal functions and all would work well. So, code just worked. Even with MySQL, if you had utf8 tables. The only time you needed 'mb_' IIRC was for regex or unicode character based length checking.
In hindsight I think having this just work for a lot of people is where PHP trumped perl and ruby in a way.
Hmm... I don't know what ruby and perl did, but "not barfing at 8bit characters" means something completely different for me than "unicode support". The second would know about various encodings, normalisation, sane comparisons, etc. Depends on your use case I guess.
In hindsight I think having this just work for a lot of people is where PHP trumped perl and ruby in a way.