i said:
>   i'll serve my pudding on my own site...

and the first thing i will tell people about each e-text is
_exactly_how_my_version_differs_from_the_p.g._version_.

even without looking at one, one thing i can say for sure
is that these diff reports are going to make p.g look bad.

some improvements are one thing -- "legalese deleted" --
but more embarrassing is that your corpus is teeming with
obvious o.c.r. errors that are detectable with minimal effort.
(if you want a catchy name, i call this the "go clown" effect.)

a lengthy list errors, in book after book, one after another,
is -- frankly -- going to be extremely appalling to people...

you're in over your head, greg.

i was 5,000 volumes into redoing the p.g. library years ago,
and i quit, because i realized that this would be the result...
i didn't want to make p.g. suffer those terrible ramifications.
but that was out of respect and admiration for michael hart,
which is no longer a consideration.  re-think, greg.  carefully.

or maybe you're thinking you need to confront those errors.
but it won't end up to be any help, because your procedure
for remounting a corrected book is far too clumsy to enact
to repair a couple of errors that occur in hundreds of books,
and a few more across a thousand other books, and so on.
if it wasn't, you woulda fixed those errors already yourself.

it's nice to fantasize that the crowd will solve your problems.

but it ain't gonna happen...  not the way you'd _like_ it to...

and it's kind of troubling to see you do this wishful thinking,
because it's not too apparent that you realize that's what it is.

i know you want a magic bullet, greg...  we all do...

you can still count me in.  but i'm telling you it's a bad idea.

-bowerbird