On Fri, Sep 20, 2002 at 10:19:16AM +0200, Lars Gullik Bjønnes wrote: > >> Development-wise, the advantage is that validate gets field testing, > > > | Development-wise, it adds about two hundred lines of code which > | "sometimes guesses right". I did not really want to use _this_ argument > | again, but as you started with development issues: Here we go. > > _why_ does it only "sometimes guesses right", I would it suspect that > it "sometimes guesses wrong". But why does that happen, but for a bug > in the validate functions?
And remember we cannot (really) get rid of validate. There are many other things that we want to include 'as-needed'. So, we'll have to get the bugs out, like it or not. And there is always the escape hatch of doing 'when in doubt, require'. Whis *is* done in some places now -- as I suspect, to cover up validate bugs. Validate is a great thing! It is worth getting working well. > -- > Lgb > Martin
msg45036/pgp00000.pgp
Description: PGP signature
