|
From: Michael B. <mic...@cm...> - 2010-12-12 20:05:42
|
Hi, libsyncml is actually a difficult thing for me. There are five issues which influence my actual "non-behaviour": 1. The design/quality of libsyncml 2. The man power behind libsyncml 3. The situation in the Open Source scene 4. Active Sync 5. Time 1.) Libsyncml was not designed by me and there is no up-to-date design documentation. The basic design idea to make things parallel is very good but the first thing which should be designed is the API. The problem is that the robustness of libsyncml depends in the past one the behaviour of the API consumer which must understand the SyncML protocol details very well. I tried to fix this by introduction the DS API but this is more a bug fix for a design problem. So we have in fact two APIs which makes the library not more robust. I write here as programmer and not as a manager. So libsyncml is no fun. The basic ideas are good but we need a second try to get a future ready, robust library. Even the heavily optimized 0.6.0 in svn is just a fix in my eyes. 2.) Libsyncml and even SyncML is complicated. The basic design is not documented. So it is very difficult to find people who want to maintain it. A library like libsyncml needs at minimum 2-3 developers which understand it. Otherwise it is not ready for the future. Actually there is just one pseudo maintainer. 3.) There are two SyncML client implementations available as C libraries - libsynthesis and libsyncml. I know the guys behind libsynthesis and there is definitely more manpower and motivation behind the maintenance especially because libsynthesis is the base for a commercial product. 4.) My university will use Active Sync in the future. I do not want to discuss here why we come to this decision. The result is that I have now only a private motivation to get a working SyncML library. 5.) I have no time any longer to work on the library during my job because I changed the position. The summary is that I will no longer maintain libsyncml. I can provide a potential new developer with all requested informations. I personally would prefer an adaption of libsynthesis to OpenSync because this would cut the costs in terms of man power. Nevertheless it is noteworthy that libsynthesis has no OBEX support today. Best regards Michael -- ___________________________________________________________________ Michael Bell Humboldt-Universitaet zu Berlin Tel.: +49 (0)30-2093 70143 ZE Computer- und Medienservice Fax: +49 (0)30-2093 2704 Unter den Linden 6 mic...@cm... D-10099 Berlin ___________________________________________________________________ PGP Fingerprint: 09E4 3D29 4156 2774 0F2C C643 D8BD 1918 2030 5AAB |