[Opensync-commits] r966 - trunk

[<svn...@op...>]

Author: bellmich
Date: Thu Mar  5 16:56:31 2009
New Revision: 966
URL: http://libsyncml.opensync.org/changeset/966

Log:
committed ChangeLog for libsyncml 0.5.2

Modified:
   trunk/ChangeLog

Modified: trunk/ChangeLog
==============================================================================
--- trunk/ChangeLog	Tue Mar  3 18:07:33 2009	(r965)
+++ trunk/ChangeLog	Thu Mar  5 16:56:31 2009	(r966)
@@ -0,0 +1,71 @@
+2009-03-09  Michael Bell <mic...@we...>
+  * Released 0.5.2
+  * Used SourceForge.net FRS for the first time
+  * Discovered that xmlInitParser and especially xmlCleanupParser must
+    be handled and called by the library user.
+  * Added support for automatic Bluetooth channel detection
+  * Added support for generic content-types
+  * Added automatic base64 conversion for binary data types
+    which are prefixed with application, audio, image, message and video
+    (ticket #185)
+  * Added IRDA support to OBEX client transport (ticket #212).
+    The patch was supplied by Peter Collingbourne.
+  * Added a test for the high level DS API
+  * Added automatic UTF-16 to UTF-8 conversion for wrongly embedded SMS
+    which is used bei Nokia S60 platform (ticket #217)
+  * Added OMA DS 1.2 SAN support for content-types without an OMNA WSP
+    number (ticket #217)
+  * Added support for OMA DS 1.2 SAN without any data store. Mobiles
+    which receive such a SAN should answer with alerts for all
+    configured data stores. The problem is the term ''SHOULD''.
+    Nokia S60 mobiles do not work this way for example.
+  * Fixed string representation in OMA DS 1.2 SAN implementation
+    (trailing NULL bytes must not be copied).
+    Nokia S60 platform mobiles work now with OMA DS 1.2 over OBEX.
+  * Fixed build environment to build correctly if OpenOBEX or libsoup
+    are missing (ticket #211)
+  * If a SAN is received then the remote location must be taken from
+    the SAN alert.
+  * DS API should only set configuration options if they are present.
+  * Added support for receiving SANs to DS API
+  * The target of an alert can now be different from the source in
+    the DS API (e.g. source = Contacts, target = Addressbook).
+  * syncml-ds-tool supports device faking.
+  * Fixed wrong return value for vendor detection via AT commands
+    (ticket #213). The patch was supplied by Frederic Brin.
+  * Added Samsung SGH-F480 to the list of the supported models
+    (ticket #214). The patch was supplied by Frederic Brin.
+  * Added an assertion to smlDataSyncInit which checks for a non-NULL
+    dsObject (ticket #215)
+  * Fixed code to handle specific Samsung AT commands
+    (ticket #216 - still open). A patch was supplied by Frederic Brin.
+  * Added support for SourceParent and SourceTarget in XML parser and
+    assembler (ticket #217 and #218)
+  * Added support for SupportHierarchicalSync of OMA DS 1.2 DevInf
+    (ticket #217 and #218)
+  * Added some new content-types which were registered by OMA DS WG
+  * Added support for authenticated OMA DS 1.2 SANs
+  * If an enum type will be changed in the future then the order is
+    critical. Hard coded values of the elements makes enums robust
+    against re-ordering.
+  * The HTTP client transport of the DS API does no longer instantiate a
+    link object because the HTTP client transport does not support this.
+  * If an error is dereferenced by smlErrorDeref then the caller can no
+    longer access the object. (This makes sense because otherwise the
+    caller should not dereference the object.)
+  * smlTransportSetConfigOption traces no longer PASSWORD options.
+  * If a link of a server is requested to be disconnected but it is
+    already disconnected then this is sometimes no error when a very
+    fast client disconnects before the server.
+  * If the server closes the connection before the client can do this
+    then the disconnect command of the client fails
+    but the client is disconnected.
+  * Fixed wrong interpretation of return value from smlTransportFinalize
+    in data_sync.c
+  * Fixed wrong handling of REPLACE command during SLOW SYNC
+  * Fixed wrong context free of OBEX server transport finalize.
+  * Fixed error handling and reference counting of smlCredNewAuth
+  * Fixed again several memory leaks and removed all suppressions for
+    libxml from valgrind.supp.
+  * Added some code to handle the not thread safe check library
+