Hello! Just as Olivier said, the simple client/server model (trivial RPC) could be safer and more stable than current wrapper implementation. I'll do my best to add example wrapper to sipek project asap. btw: sasa == he && sasa != she // at least in my case :-) Regards, Sasa On 10/2/07, Olivier Beytrison <olivier at heliosnet.org> wrote: > > Hello Benny, > > Thank you very much for your welcome. > > As for your answers, I've already contacted Sasa off-list and got also a > quick answer from her. There might be another possibility, the one to > use the client/server model. Having the core function built in C++ and > listening to the GUI (the client) on a local port might make it safer > and more flexible. > > And don't forget that the target of this open-source application is a > mobile device running Windows Mobile 5 or 6, with its compact framework > .Net. > > If we stay on the wrapper idea, we will indeed develop a sort of > "intermediate" API which will provide only what is strictly needed by > the GUI part of the application, this preventing us to wrap over 100 > functions and callbacks. > > Anyway, I'll keep the list informed on which way the project will take, > and will be happy to listen to anyone's suggestion or ideas :) > > Regards, > Olivier B. > > > Benny Prijono a ?crit : > > Hi Olivier, > > > > welcome to the list! > > > > Indeed when developing applications on Windows, you have two > > choices, to develop in C++ directly using pjsip, or to develop it > > with .NET by creating pjsip wrapper for .NET. Well actually there > > are two more options for Windows, if we count Python and TCL. ;-) > > > > Whether to use C++ or .NET would depend on how much time you'd > > prepare to spend on creating pjsip .NET wrapper, vs how much more > > productivity you'd gain by using .NET. The cost of creating the .NET > > wrapper further depends on how many APIs would you need from pjsip. > > FYI pjsua-lib has more than 100 APIs (please see [1]), it would be > > time consuming to implement all of these in your wrapper. So I would > > imagine that you would create yet another abstraction on top of > > pjsua-lib, select the APIs that you need, and export these in your > > wrapper, so that it has less APIs. > > > > Alternatively you could also join Sasa Coh's project. She's created > > .NET application on top of pjsip, and there is a .NET wrapper for > > pjsip as part of this project. Please see [2]. Personally I prefer > > this approach, not only it would help you get there faster, but it > > would also help yet another open source project. > > > > So all is possible, the choice is yours. > > > > cheers, > > -benny > > > > [1] http://www.pjsip.org/pjsip/docs/html/group__PJSUA__LIB.htm > > [2] http://www.pjsip.org/trac/wiki/FAQ#activex > > > > > > Olivier Beytrison wrote: > >> Hello, > >> > >> I'm very new to pjsip, I discovered it only this morning, but it might > >> be my life saviour. > >> > >> I'm currently involved in a project about developing an IMS client for > >> the Next Generation Networks as defined by the 3GPP. This client will > >> run on PDA under Windows Mobile 5/6. > >> > >> I'm very interested into pjsip because the WM5/6 SDK lacks a > sip/sdp/rtp > >> stack. > >> > >> Thus, I have a question : > >> Is it viable to use the pjsip libs in a C# for CF.NET using marshalling > >> ? or should i really stick with C/C++ for embedded devices for my > >> application? > >> > >> The object-oriented side of C# is really appealing and the easy way to > >> create the GUI as well. But off course it is not possible to directly > >> use pjsip in C#, that's why i'm thinking about making an intermediate > >> C++ code between the C# and the libs. > >> > >> Thanks in advance for your answers > >> > >> Regards, > >> Olivier B. > >> > > > > > > -- > > Olivier Beytrison > Telecommunication Engineer > Mobile: +41 (0)78 619 73 53 > Mail: olivier at heliosnet.org > GPG: 0x4FB83528 http://pgp.mit.edu/ > > _______________________________________________ > Visit our blog: http://blog.pjsip.org > > pjsip mailing list > pjsip at lists.pjsip.org > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.pjsip.org/pipermail/pjsip_lists.pjsip.org/attachments/20071002/4a634b11/attachment.html
