Re: [Serusers] problem with compiling the new CVS release

Hi Michal, Thank you for your reply. I didn't understand what to do. Should I add the option to the command "cvs -d :pserver:anonymous@cvs.berlios.de:/cvsroot/ser checkout sip_router" ? I gues not, because it does not accept that parameter. What should I do now ? Should I download a stable version ? Where can I find a stable version with presence support ? In fact, I only need the new registrar module which includes "nat_flag" parameter. But when I just add the new registrar module to my old sources, I can't compile it. Thanks, ilker -----Original Message----- From: Michal Matyska [mailto:michal@iptel.org] Sent: Saturday, March 31, 2007 10:01 PM To: ?lker Aktuna (Koç.net) Cc: serusers(a)iptel.org Subject: Re: [Serusers] problem with compiling the new CVS release Hi, Sorry for the inconvinience... checkout the CVS code before the DB API started to change (e.g. add -D 20070328 option to the cvs co command). Until all modules will be changed and using new API, it won't compile for a while. Michal On Thu, 2007-03-29 at 11:48 +0200, Jan Janak wrote: > Folks, > > I have been working on a new database abstraction layer for SER for a > while and I'll start commiting it in CVS now. Please excuse me if > something is broken temporarily, this is a major improvement that > touches all modules that use the database. This change does not > affect Ottendorf, of course. The new DB layer will appear in 2.1 > > The main objectives for doing major changes in the DB API are: > > 1) Better support for databases that are not based on SQL, like > Berkeley db, LDAP, and fas in-memory databases. > > 2) The new code will no more store fetch all data from the database > in one step, it will be possible to get the data row by row (common > problem in usrloc) > > 3) Support for queries over multiple databases with various strategies > (i.e. usrloc database can be partitioned). > > 4) Support for database failover (i.e. if the primary database fails > that SER can start using another one, which can be even of different > type, i.e. mysql->postgres) > > I provide you with more details and documentation later once the > first version is in CVS. > > Jan. >