Re: Much leaking in lib/tie-refhash.t @7528

2000-11-03 Thread Jarkko Hietaniemi
Okay, now I see it too, -DDEBUGGING seems to be the key. -- $jhi++; # http://www.iki.fi/jhi/ # There is this special biologist word we use for 'stable'. # It is 'dead'. -- Jack Cohen

Re: README.aix

2000-11-03 Thread H . Merijn Brand
On Fri, 3 Nov 2000 07:34:16 -0600, Jarkko Hietaniemi <[EMAIL PROTECTED]> wrote: > > OK, I got some response here, and some private response, which > > indicates that the info isn't seen as useless ;-) > > I wasn't claiming uselessness, but oververboseness. > > > Jarkko, should I change the patch

Re: rsync'ed patches vs. rsync'ed source

2000-11-03 Thread H . Merijn Brand
On Fri, 3 Nov 2000 07:36:30 -0600, Jarkko Hietaniemi <[EMAIL PROTECTED]> wrote: > > I think either this doc should get a place of it's own or the info states > > should be integrated in things like perlhack, perlguts and such. > > Integrating with perlhack should do fine. Early next week. I hope

Re: Reference counts on self-tied objects

2000-11-03 Thread Alan Burlison
Steffen Beyer wrote: > Basically, after building your data structure, you save the current > ref-count (minus 1) of the object in question somewhere (in a private > attribute of that object) and set the ref-count to one. > > When the user abandons his/her last reference to this object, DESTROY()