On Sat, Jul 21, 2001 at 07:12:30AM -0400, Jeff 'japhy/Marillion' Pinyan wrote: > On Jul 21, Michael G Schwern said: > > >caller() appears to be confused by the & prototype and reports the > >line where the function call *ended* rather than where it starts. > > This also happens when you send sub { ... } as an argument to a > function. Someone has already posted a bug about it. (He was willing to > pay the porter who fixed it!) That's exactly what's happening here. The existing bug id is 20010622.008. Robert Spier pointed it out to me when I fixed another line numbering problem recently, but this case is slightly different. In the bug I fixed the correct information was in the optree, and just had to be found. In this case the line numbers you see are those which are in the optree. I was planning on investigating, but haven't found time. If you fix it (Michael) it might reduce your exposure when the tests are complete ;-) -- Paul Johnson - paul@pjcj.net http://www.pjcj.netThread Previous