Re: balsa 0.6.0
- From: pavlov netscape com (Stuart Parmenter)
- To: "Michal Jaegermann" <michal ellpspace math ualberta ca>
- Cc: <balsa-list gnome org>
- Subject: Re: balsa 0.6.0
- Date: Fri, 24 Dec 1999 13:18:58 -0500
Hi Michal,
I just checked in things for the address book that should fix the int
compairison problems. Have you tried removing your ~/.balsarc? My guess is
that you have an old config and that 0.6.0 is looking for things that are
not there.
Thanks,
Stuart
>I tried your new release on my Alpha. As far as the program itself goes
>then results are not very appealing. Either it crashes immediately in
>add_mailbox() function from local-mailbox trying to dereference null
>pointers in order to check various (non-existing yet) mailbox types or,
>with this fixed, opens a huge window (two screens high and one screen
>wide - I think) with a title "Configure Balsa", an inscription "Welcome
>to Balsa!" on a navy-blue top and left margins and nothing else. It
>just sits there, making hard to get to anything else, and leaving
>as the only option a kill. No any reaction whatsoever on a keyboard
>or mouse. On stdout I am getting:
>'*** One of inbox/outbox/sentbox/draftbox/trash is NULL'
>which is indeed the case but hardly helpful.
>
>balsa-0.4.9.5, which I tried before, at least starts on my machine
>and opens some mail window. Not that I really attempted to use it to
>send and receive mail but this is at least a start.
>
>There is a usual spate of complaints from gcc about conversions between
>pointers and ints of different sizes. Sigh! Comparing pointers in
>ab_delete_compare():address-book.c after casting them effectively to
>'int' will definitely not fly on any 64-bit platform. Sometimes it may
>even work, purely by an accident, but do not count on it. Besides
>casting them to signed types may actually reverse returned values. I do
>not think that you are much worse comparing straight pointers anyway;
>even if C does not guarantee that results will be valid there are also
>no guarantees that a cast helps in any manner. Other nasties may simply
>reflect design deficiencies in gtk; I do not know at this moment.
>
> Regards,
> Michal
> michal@harddata.com
>
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]