DT’s indexing finds searched documents quickly, but if a document is 400K, for example, it takes 5 seconds or more to find the search-term location within the document – after I click on the document I’m interested in.
Does this mean that DT is doing a linear, pattern-matching search within the document?
Is there any way to speed this up. When scanning for information, 5+ seconds starts getting into the “slow enough to notice” category.
DT can get sluggish after lots of activity, especially if the database hasn’t recently been optimized using the Backup & Optimize tool. (I did a Backup & Optimize last night, but have been using DT a good deal today.)
I’m running on a TiBook 500 Mhz with 1 GB RAM. I just did a test search for the term “development” that returned 2,712 hits in 0.171 second. When I selected a 976 KB item among the hit list (a plain text capture of a large PDF file), it took less than 2 seconds to open the file in the right window and scroll down to the first occurence of the search term, highlighted. That’s not bad for a relatively slow and old computer.
This was a simple one-term “All Words” search on the database. Other searches, especially Phrase searches, may take longer – the more terms, probably the longer the search time.
I see slowdowns once in a while, too. When that gets irritating, I take a break and run Backup & Optimize. If DT still feels a bit sluggish, I might close the application and relaunch it, which sometimes speeds things up.
Sometimes slowdowns occur when Virtual Memory is being used, which requires operations to be paged in and out of hard drive caches. A computer restart may help when things get to the point of very frequent VM use.
Future versions of DT will use a different database structure that will considerably reduce memory requirements and greatly speed up searches.