From: | Daniele Varrazzo <daniele(dot)varrazzo(at)gmail(dot)com> |
---|---|
To: | Benjamin <ikipou(at)gmail(dot)com> |
Cc: | psycopg(at)postgresql(dot)org |
Subject: | Re: Conversion of empty arrays |
Date: | 2011-02-20 11:32:28 |
Message-ID: | AANLkTikh6qved9Mo9nZRyKUWP_Kx=q6UXnQdGM=Hfor+@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | psycopg |
2011/2/20 Benjamin <ikipou(at)gmail(dot)com>:
> Hello,
> The commit 89e4d4c7bb9534711c03df054f90fcad4787d209 caused some havoc on my
> code because comparisons involving empty arrays were failing due to
> incorrect types.
> I quickly hacked something to fix the regression, the patch is attached to
> this mail. I would like this regression to be fixed upstream, it the mailing
> list the right way to submit patches?
Hello,
the regresison has been reported (ticket #42) and is already fixed in
the current dev (but your fix seems more exhaustive, I'll see if there
is something interesting to take, thank you).
I'm afraid the problem is made worse from the fact that easy_install
installs the beta by default: even if the beta is not listed on PyPI,
setuptools, pip and all the happy brothers prefer to screenscrape the
project homepage and install the beta. I've reported the problem but
they seem to think that's a feature.
For this reason it's probably better if we release the 2.4 soon.
Federico, I have a few commits in my dev: take a look if you want to
merge everything and then I think we can go. I'm also up-to-date with
Jason so there should be no problem on Windows either.
-- Daniele
From | Date | Subject | |
---|---|---|---|
Next Message | Brian Sutherland | 2011-02-20 16:47:00 | NULL dereference when memory is tight |
Previous Message | Benjamin | 2011-02-20 01:49:05 | Conversion of empty arrays |