No, Socrates appears as a character in Aristophanes' Clouds (and the whole point behind Attic comedies is to make fun of real people), Plato's contemporary Xenophon wrote a good deal about Socrates, and some sections of Socratic dialogues written by Aeschines of Sphettos still exist. Aristotle was born over a decade after Socrates' death, but that puts him far closer in time than us, and he never questions that Socrates existed.

A little more history of philosophy might have helped.

I'm not the anonymous author of the grandparent post, but I can answer your question: in the Symposium, there are very few such tag lines. The dialogue is embedded in a conversation between two characters (Apollodorus and one of his nameless friends), but the main content (the story of Agathon's "symposium", i.e. drinking party) is related at second hand by Apollodorus (in fact, he reports the whole thing as something someone else told to him, which makes for some interestingly complex Greek grammar).

25 lines in 2400 is about 1%. Given the uncertainty in knowing just where Kennedy's breaking the lines, that's pretty close. In fact, I took a few hundred lines from your version and moved the line breaks so that they always occured at a syllable (so that no line is over 35 characters, though many are less), and it moved the line count up about 2%. We don't know exactly how Kennedy's handling the line-break problem, so you may both be right, given your ways of counting lines. We also don't know what conventions Kennedy's following about breaking when a new character takes over (the Symposium is in a sort of embedded oratio obliqua, with two characters doing the talking, one of whom (Apollodorus) delivers most of the dialogue as a long indirect-speech report. Is Kennedy breaking where characters change, in either case?

With a 1% shorter line count, we'd expect it to begin at line 396 by your count. 396 - 377 gives 19, or about 5%, though the length variance is still 10%.

The starting-point is within 3%; if we go with 179 for the length, it's again about 10% variance between the two of you.

22 out of 800 is less than 3% variance.

20 off out of 1200 is within 1%.

53 out of 600 (9%), well above the previous variances.

This is the biggest variation in the list (13%).

So, I think it's too strong to say Kennedy has only found very approximate correlation". We need to know more about just how he's determining his lines.

Your version uses rigid 35-character lines, even when that breaks a word at a place no Greek would have (e.g. your very first line chops off a sigma from ameletêtos and puts it on the next line; a scribe would have broken at a syllabic division, surely). If the 35-character length is taken as a maximum for a line, then allowing for this will make some lines shorter than 35 characters and thus bring down the counts. Of course, you could adjust so as to get an overall average of 35. Either way, you'd need to do a lot of manual work to insert plausible breaks. I have no idea whether this would bring things more into line with Kennedy's data or whether Kennedy is allowing for it as well. It would be useful to have the details of Kennedy's algorithm.

Actually, the charge appears to have included both of these, to judge by Plato's Apology: corrupting the youth, not worshiping the gods the state worships, worshiping "strange" gods, and for good measure "making the weaker argument the stronger".

From my FreeBSD 8.0-RELEASE box:

$ file `which ldd`
/usr/bin/ldd: ELF 64-bit LSB executable, x86-64, version 1 (FreeBSD), for FreeBSD 7.2, dynamically linked (uses shared libs), FreeBSD-style, stripped

Looks like a binary to me.