Hi out there …

I just started using Tamino and i’ve got the following problem:
i have a list of about 30.000 articles. I’ve put them into Tamino with the perl taminload tool. It took about 20 minutes (a bit long if i compare it to MS SQL Server).
To select articles by the “id” attribute i found out that the “~=” statement works quite faster (what’s the difference to a simple “=” statement, that takes lot more time to load ?).
No i want to select articles by their timestamp (set in milliseconds) with a “<” statement. But this takes a lot of time (min. 1 minute) … and i think that’s not acceptable for a database. Is there a faster way to select the items from the database (I’m sure i’m doing something wrong)?
The id and the timestamp field are defined as “search-type=text” in the doctype as it is described in the documentation (you remember ? … the thing with the birthdate of a patient ?)

Hope to hear from you …
Michael

search-type = “text” means you wish to perform text type searches, e.g. to find an element with the word ‘Hello’ you could say x/y[z~=“hello”]. For search-type = “standard” (which is what you need) allows standard type queries (=, >, < etc) to be issued, e.g x[timestamp>2323]. If the appropiate index is not built Tamino will still attempt to resolve the query by reading every document and this make take time. Of course if you want the best of both indexes you can set the search-type = “standard text” - this will require more database space and will take a little longer to store documents. :rolleyes: