Well … I didn’t get to quit yesterday :p – couldn’t get a chance to see my manager because of work and other stuff <vbg> So maybe today … but I did get a chance to fix the problem with Quester 🙂 I managed to figure out the problem with XML Mapper – turns out that if a set of elements are to be treated as a client dataset (so that I could have a master-detail relationship), there must be more than one of that element in the first grouping of XML elements that qualifies as a record in database terms … Not sure whether that made any sense :p but basically that was where the problem was – I finally figured it out by messing around with a sample file and changing different elements on it and loading it into the XML Mapper. Maybe that was mentioned somewhere in the docs but I have a bad habit of not wanting to wade through tons and tons of documentation <vbg>
So Quester now officially works – except for one problem. I added an option to refresh the search engine definitions in case you decided to change them or add a new definition but that doesn’t seem to work. Calling Refresh on the TClientDataSet component results in some pretty weird results – so the current option seems to be to close the Quester toolbar and re-add it to the taskbar. I guess I can live with it 🙂 I don’t think I mentioned yesterday that I heard from David – the author of DQSD, on which Quester is based … – yesterday. I’d asked him permission to release Quester publicly and he said that since the searches in Quester were contributed by a lot of people, that I should ask all of them. So I joined the DQSD mailing list and have been carrying on a conversation with some of the people there about Quester, open source and perhaps keeping the XML formats for DQSD and Quester consistent …