Author pje
Recipients barry, brett.cannon, eric.snow, inada.naoki, isoschiz, kristjan.jonsson, ncoghlan, pconnell, pje
Date 2013-04-14.14:34:58
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <CALeMXf7iZb94v-C9hK3bWXVfWboK++FTERAucnHHs+bjX8fk2Q@mail.gmail.com>
In-reply-to <1365925912.95.0.799491999036.issue17636@psf.upfronthosting.co.za>
Content
On Sun, Apr 14, 2013 at 3:51 AM, Nick Coghlan <report@bugs.python.org> wrote:
> Your analysis is one of the pieces that was missing,

Unfortunately, I just noticed it's actually incorrect in a pretty
important part  In my original example, I said, "because of the
circularity, this will *also* happen if you import 'a' first."  This
statement is actually false.  Importing 'a' first in that example will
result in a.util == b.util:util, not a.util=b.util.  I made the
mistake because I was for some reason thinking that 'a' was going to
execute its import while being imported from b.util, and in that
scenario it will not.

That means there *is* an ordering dependency, and an ambiguity like
this one can lie dormant until long after you've introduced the
circularity.  :-(
History
Date User Action Args
2013-04-14 14:34:58pjesetrecipients: + pje, barry, brett.cannon, ncoghlan, kristjan.jonsson, inada.naoki, eric.snow, pconnell, isoschiz
2013-04-14 14:34:58pjelinkissue17636 messages
2013-04-14 14:34:58pjecreate