- From: Adam Dickmeiss <adam@indexdata.dk>
- Date: Mon, 17 Sep 2001 14:07:20 +0200
- To: Mike Taylor <mike@tecc.co.uk>
- Cc: ian.ibbotson@k-int.com, www-zig@w3.org
On Mon, Sep 17, 2001 at 12:21:04PM +0100, Mike Taylor wrote: > > Date: Sun, 16 Sep 2001 13:48:51 +0100 > > From: Ian Ibbotson <ian.ibbotson@k-int.com> > > > > The real need for sort to be close to search-request is, as the doc > > says, for targets that need to be able to optimise the way they > > work. I have a few SQL backends, and it's really inconvienient to > > execute what can be a heavy SQL statement, only to have to > > immediately re-evaluate it because a sort-request has just come in > > and now we need to tag some order by clause on the end. > > This is purely an implementation issue, isn't it? Your server can > implement this stuff however it likes. For example, when you get a > search request, do nothing but remember what the request was, and > return a "success" indicator. You need only actually submit the SQL > query when the user either (A) sorts the result set (in which case you > can include the ORDER BY clause) or (B) presents some records, in > which case that's a great moment to actually do the search. Well, you need to return a resultCount immediately. And that require some work to be carried out.. -- Adam > > ``Mmmm ... laziness. Is there _anything_ it can't do?'' > -- Homer Simpson, paraphrased. > > _/|_ _______________________________________________________________ > /o ) \/ Mike Taylor <mike@miketaylor.org.uk> www.miketaylor.org.uk > )_v__/\ "Good luck, everyone" -- Bob the Angry Flower after the WTC > disaster. See http://angryflower.com/septem.gif -- Adam Dickmeiss mailto:adam@indexdata.dk http://www.indexdata.dk Index Data T: +45 33410100 Mob.: 212 212 66
Received on Monday, 17 September 2001 08:08:29 UTC