In article <[EMAIL PROTECTED]>, Stefan `Sec` Zehl <[EMAIL PROTECTED]> writes: > On Sat, Jun 08, 2002 at 12:05:55PM +0100, Stephen Turner wrote: >> But why didn't the referees allow it? There's nothing in the rules about it, >> is there? And surely it wasn't just the problems with submitting it, because >> other non-printable characters are as bad. Did they give a reason? > > As far as I found out, there are three characters that can't be > submitted. Binary \0 (which trunctates the file then and there), Binary > \r which gets mangled to a \n (My first compressed solution failed on > this), and from the comments somewhere in the post mortem \xff.
No, \xff works fine, but it's easy for the judges to mangle it with cut and paste. (probably the judges mail should have an attachment (too?) for the hard cases) > > I would very much like to see PGAS allow _all_ characters, and the %.. > method seems okay to me. I think the combined post-mortem page should > show them all in that form - perhaps with a download link to download a > single solution in binary correct form? > The link idea is great, since pasting solutions from the page doesn't work too well anyways due to spaces getting weird (at least with netscape). I think I'd prefer replacing most lower chars with ^control-char code (so \x0 would become ^@) code (this doesn't work in string literals), or \octal(or \hex) (so \x0 would display as \0), (which doesn't work for the special variables). Maybe a semi-intelligent one: use \octal unless directly after a $, then use ^letter. That should all cases except a single quoted construct containing $ foillowed by a low char. And do an HTML::Entities::encode for the HTML display of the high ones (e.g. \xff would then become ÿ in the html)
