Ciao Andri, To keep the results of a query consistent one better starts a transaction. A database can be updated by one thread at a time, but read by many. Suppose you need a list of all employees of a department. If you start to search one at a time a lot of changes can happen. An employee can be moved to another department, fired or a new one entered. Another reason is speed. All (or most if it is a huge department and/or a tiny cache) employees are in cache so you're sure you have the requested list showing the situation on the moment of creation.
Andrii Motsok wrote Thu, 6 Oct 2016 11:43:10: >My understanding is that any reading from database automatically starts read >transaction. >The question: why does sqlite need to do it on read-only database? Kind regards | Vriendelijke groeten | Cordiali saluti, Klaas `Z4us` van Buiten V, Experienced Freelance ICT-Guy https://www.linkedin.com/in/klaas-van-buiten-0325b2102 _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
