Hi Rene and rooters, I saw another message on the roottalk about TTree modification. Rene, is "one write - many read" policy your final answer? I still do not understand why you do not want to provide (optional) insert/delete/modify entry mechanism if you told once it is simple from technical point of view. Here is a simple example of an application where ROOT fits data warehousing needs but can not be efficiently used without TTree modifocation. Trade and quote real time data from Reuters/Bloomberg and other providers are quite similar to experimental on-line data. Every day you get a plane stream of thousands of records ordered in time. Thus you can write them in a root tree and beat any other RDBMS with the speed when you analyse this data warehouse. One small detail breaks down this idyllic picture: after some time you can get a signal saying that a record #xxx was not correct and the next record is corrected #xxx. The same happens when you want to mark certain events in the warehouse as outliers. Etc. etc. I would say that in general this feature prevents ROOT from spreading out in the other than hep communities. One of such applications I have mentioned above. Should we discuss this tree modification topic within ROOT community and formulate the final decision? Regards, Anton http://www.smartquant.com
This archive was generated by hypermail 2b29 : Tue Jan 01 2002 - 17:50:37 MET