err, I humbly and abjectly withdraw my comments below.  (I should have
checked the error message in the error_log when it failed after manually
including the second file's contents in the eruby file.)

-pate  <-- note the penitent set of the signature

On Tue, 5 Feb 2002, Pat Eyler wrote:

> On Tue, 5 Feb 2002, Sean Chittenden wrote:
>
> > You need to declare cgi in your included file... that or you need to
> > include your file, then pass cgi as an argument.
>
> hmm, it's not actually a seperate function or object.   I made the
> #include-like assumption that if I were writing something procedural (I
> know, I'm a heretic) that the code from the 'require'd file would just be
> set in place and executed in the normal flow of traffic.
>
> > Something tells me your background is in PHP, yes?  ;~)
>
> not really, I've done a bit of perl, a bit of php, and a (tiny) little
> bit of C, but mostly shell stuff.
>
> > Ruby/eRuby is much more like a
> > normal programming language in that its got sane scopping and
> > namespace pragma.  -sc
>
> I don't understand how scoping would affect this.  If I just add the code
> into the eruby file it would look like (by the way this doesn't work on
> the production machine either):
> ============
> require 'cgi'
>
> cgi = CGI.new("html3")
>
> if (cgi['process'])
>  ## begin 'require'd file
>
>  date = cgi['date']  # this doesn't work when it lives in a
>                      # different file or in the eruby file
>  # process the date to make ensure it fits our defined format
>  # write the date and other POSTed data into a file
>
>  ## end 'require'd file
>
> else
>
>  # create a form to gather the required data
>
> end
> ============
>
> Just for clarity, I built ruby and eruby from the same tarballs for both
> machines.  I'd be happy to dig more information out as needed.
>
> thanks,
> -pate
>
> >
> >
>
>
>