[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Pan-devel] 0.92 threading & GtkTreeViews
From: |
Charles Kerr |
Subject: |
Re: [Pan-devel] 0.92 threading & GtkTreeViews |
Date: |
Mon, 10 Apr 2006 21:30:43 -0500 (CDT) |
User-agent: |
SquirrelMail/1.4.6 [CVS] |
> On Mon, 10 Apr, 2006 at 23:05 +0200, Charles Kerr wrote:
>> [ threading issue ]
>> I looked at this today and narrowed the possibilities a little bit.
>> You will probably find that the References header for (2) in
>> /.pan2/groups/comp.os.linux.misc is corrupted, so the bug is somewhere
>> in data-impl -- either in the initial threading or in writing
>> References -- but probably not in the GtkTreeView.
>
> Indeed, this is the section for (2):
> <address@hidden>
> Re: VPN and VNC for Embedded Software Porting and Development in Remote
> Hardware
> b
> <address@hidden>
>
> Just as you suspected, the references were trimmed to one (the first?).
> After I removed ~/.pan2/groups/comp.os.linux.misc and retrieved all
> headers again, the threading was correct again.
Aha, this *might* be the problem, basic_string::insert(pos,Quark)
is very suspicious, though I'm not sure it's causing the problem...
--- headers.cc.bak 2006-04-10 21:24:21.000000000 -0500
+++ headers.cc 2006-04-10 21:25:56.000000000 -0500
@@ -547,9 +547,10 @@
const ArticleNode * node (h->_nodes.find(message_id)->second);
while (node->_parent != 0) {
node = node->_parent;
- references.insert (0, node->_mid);
+ const StringView ancestor_mid (node->_mid.to_view());
+ references.insert (0, ancestor_mid.str, ancestor_mid.len);
if (node->_parent)
- references.insert (0, " ");
+ references.insert (0, 1, ' ');
}
StringView refs (references);
refs.trim ();
@@ -560,7 +561,7 @@
<< author_qts(a->author) << "\n\t";
// references line *IF* the article has a References header
if (!refs.empty())
- *out << references << "\n\t";
+ *out << refs << "\n\t";
// date
*out << a->time_posted << "\n\t";
>> [ GtkTreeView padding ]
>> IMO this is too narrow. Maybe I've just gotten used to the new widget.
>> The default value is 2 and you're proposing 0; how about we split the
>> difference and make a padding of 1? ;)
>
> How about we make it a preference? :) Just kidding. Maybe I'm just not
> used to the new widget yet (I think CList did no padding). I'd be
> interested to see what other people on the list think.
You might want to post on pan-users?