>
> It is defined as a constant:
>
> *bgcolor = \"beige";
>
> This is the way to do constants which I read about in the Programming
> Perl book...
Never saw this before, but it really defines a constant. The problem is from
within Perl I don't see any chance to catch the difference between such a
constant and a normal scalar. So the only possiblity at the moment is to
ignore the error message. (Embperl handles the cleanup in XS, there I can
check if it is a read only value and don't clean it up)
> if this messes with Embperl in some way then is there
> another way to do it?
The normal way to define a constant I know, is to say
use constant bgcolor => 'beige' ;
In your code your refer to it as bgcolor and _not_ as $bgcolor.
> But the thing is, I have defined a whole bunch of
> these, yet only the one error pops up. Strange.
If I remeber right I saw another one in the logfile you send me, but I am
not sure...
>
> Incidentally, here's another possible clue: The errors I get (in the
> browser) when I use the latest CVS release are all very similar, if I
> have the logging turned off - I don't get this with EMBPERL_LOG not set.
> It's a long list, and they all look like this, but referring to
> different methods:
>
> [1175]ERR: 24: Line 3: Error in Perl code: Recursive inheritance
> detected while looking for method 'get_tour' in package
> 'HTML::Embperl::DOC::_3' at
> /www/crazyguyonabike/com/htdocs/journal/contents/init.html line 13.
>
> I think this is something to do with the caching or ISA thing,
yes, the @ISA is corrupted. Maybe it's the same problem as Angus and Gavin
reported, but they reported it with the old epmain.c
> because
> when I copy the epmain.c which you sent me a few weeks ago over the one
> in the latest CVS and rebuild Embperl, it all works again. So it should
> be relatively straightforward to see what's different between the two
> versions of epmain.c - I did a diff and there are only a few
> differences. I sent you the working epmain.c yesterday, Gerald.
>
Yes, I did a diff too and the difference is as I expected. There was a
conditional which tests for absolut paths and it was wrong, so I added a NOT
operator. The effect was that with the old epmain.c it could happen, that
the same file was compiled into two different packages, while the new one
makes sure that the same file is compiled into the same package.
I try to track this further down with the log you send me
Gerald
-------------------------------------------------------------
Gerald Richter ecos electronic communication services gmbh
Internetconnect * Webserver/-design/-datenbanken * Consulting
Post: Tulpenstrasse 5 D-55276 Dienheim b. Mainz
E-Mail: [EMAIL PROTECTED] Voice: +49 6133 925131
WWW: http://www.ecos.de Fax: +49 6133 925152
-------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
