On Mon, 2012-02-06 at 13:27 +0100, Kees Bos wrote: > On Mon, 2012-02-06 at 12:06 +0000, lkcl luke wrote: > > On Mon, Feb 6, 2012 at 10:41 AM, Kees Bos <[email protected]> wrote: > > > On Mon, 2012-02-06 at 10:46 +0100, Kees Bos wrote: > > >> On Mon, 2012-02-06 at 09:29 +0100, Kees Bos wrote: > > >> > On Sun, 2012-02-05 at 20:30 +0000, lkcl luke wrote: > > >> > > mainly this goes to kees: i had to revert a small change that you > > >> > > made, bug #521 - it stopped LibTest from operating correctly on > > >> > > firefox (ok, iceweasel, v 3.5.9). > > >> > > > >> > OK I'll have a look. > > >> > > > >> I can reproduce this (looks like a symptom of another problem). BTW you > > >> haven't pushed your revert yet. > > >> > > > > > > Got it. Was a bug in dir(). > > > Fixed in commit 67926637770597ca00a51b05b0e5008fa816f1e5 > > > > > > Luke, could you take care not to push the commit with the revert for the > > > fix of issue #521? (It wouldn't be a problem if you had, but since you > > > didn't ... :-) ) > > > > :) > > > > well it appears i haven't done anything at all, which is bizarre. i > > think i did the bugreport first then just plain forgot to patch the > > darn thing. huh. > > :-) > > > > > let me know ok? > > The attribute error was also fixed with the fix for dir() (dir() doesn't > report the default javascript 'prototype' attribute anymore, so there > isn't a getattr lookup for the non-existent CompileTest().prototype) >
Now I read this, you might ask: why is it non-existent now? Well, the mapping is foo.$$prototype, which doesn't exists. The previous getattr, however, would return the (false) foo.prototype.
