Bob Rogers <[email protected]> writes: > That's pretty much what I would do. Though I don't understand why you > need the "eval" in: > > eval('&Net::FTPSSL::IMP_CRYPT') > > Shouldn't Net::FTPSSL::IMP_CRYPT always be defined when you need it, and > never referenced when you don't? >
Looks like the eval is unnecessary. I guess I blundered into this with only IMP_CRYPT and got a bareword complaint. Then I thought for some reason that eval would solve that (oops), but then found I needed to qualify it, for the same reason indirected I suppose, and didn't take out the unnecessary eval. I guess I was confused thinking a level of indirection would be needed to prevent the compilation seeing my use of the Net::FTPSSL::IMP_CRYPT symbol before seeing the one pulled in from the require Net::FTPSSL statement and deciding it was a bareword/string instead of a sub/constant. Or I wasn't thinking clearly at all maybe. Thanks. I'm not unhappy with it now. -- Mike Small [email protected] _______________________________________________ Boston-pm mailing list [email protected] http://mail.pm.org/mailman/listinfo/boston-pm
