I think i must explain some things about 2.0 version... The Database exists in 2 forms : * "standard db" : the DB coming from the 1.2 version * "MARC-DB" : the MARC one. A parameter tables maps marc fields and standard-db ones. The new 2.0 guarantees that both DB are always synchronised. That's why it was long to write, and to debug. But, now we have a big advantage, as the user can decide to use MARC or non-MARC interface. In both cases, the 2 databases are living. If you choose the MARC one, the parameter tables provides the following features : * support any MARC flavour : just define what 100$a, 200$a... mean * support what you want in your MARC flavour : 100$a => used 100$b => used, but less important, 100$c => ignored The manual MARC cataloguing tool is ready in 1.3.3, and used by a french library in France. My question was not about marc/non marc, my question was to have advices for a deepest separation between acquisition and cataloguing. I repeat that here, in France, there are almost no librarians that enters manually their biblios : universities have a common pool, the National Library (BNF) provides a cdrom with 4 millions of UNIMARC notices, and Electre provide a commercial one with recent books. So most ils separate acquisition and cataloguing, as librarians "spend not more than 10 mn a week on cataloguing" as mailed me a french koha-addict ! Let's continue your answers/ideas/... on this topic, i'll conclude and propose something (on the wiki) next week. you'll be informed of my proposal when ready, and get the opportunity to comment it too ! Note : i plan to do something SIMPLE (from a developping time point of vue) for the 2.0 release which should occur in a month (at least for a RC one) In 2.0.1, I expect to improve it (i WILL improve it if it's funded by someone !) -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)