Ok, that's what I had thought may have been happening (sem-random sort
order). I don't know if the order *needs* to be exactly correct when
the dcdate or timestamp are the same, but at least keeping the feed
order consistent is a must.
This change should be made and will be put in the 0.2 Release Candidate.
Thanks for your help
Andy
On 12/12/05, Kevin <ke...@dr...> wrote:
>
> > Ok - I've noticed this over the past couple of days - but only
> > verified it this morning.
> >
> > When I'm reading feeds, I"ll choose "flag all up to this item". If I
> > click on, say, the fourth feed, and then click "mark as read", I
> > notice that what was supposed to be the next newest item is still down
> > on the list (originally I thought it marked it as read), and items
> > that were lower (or not there?) are now at the top, chronilogically.
> >
> > Perhaps this is something to do with the timestamp for ordering? I'm
> > still investigating, but I wanted to put it out there for you devs to
> > also keep an eye out for. Before you click "Mark as Read" on a number
> > of flagged entries, see what the next item (that is unflagged) is and
> > see if it's at the top of your items when you click "Read".
> >
>
> For identical values in the sort column (currently dcdate), it is not
> guaranteed the item order will be consistent.
>
> I see this often for feeds which don't include a pubDate or dcdate. The
> fallback is to use the cache timestamp.
>
> We could change this behavior easily by using SQL and specifying 2 sort
> columns.
>
> "SELECT ... ORDER BY dcdate, id"
>
> --
> Kevin
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems? Stop! Download the new AJAX search engine that makes
> searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
> http://ads.osdn.com/?ad_idv37&alloc_id=16865&opclick
> _______________________________________________
> Fofredux-devel mailing list
> Fof...@li...
> https://lists.sourceforge.net/lists/listinfo/fofredux-devel
>
--
Andrew Turner
ajt...@hi... 42.4266N x 83.4931W
http://highearthorbit.com Northville, Michigan, USA
|
