--- Původní zpráva --- Datum: 8/22/2005 Od: "David Kirkby" Předmět: Re: [Moon-net] WSJT Development Status Joe Taylor wrote: > Dear EME Friends, > *Additional information for programmers: > ----------------------------------------- > Like all of us, I am not getting any younger. A time will surely come > when I am unable to continue development of the WSJT program, or even to > make sure that it will still work with the next generation of personal > computers. > > For these reasons, I very much want to encourage others who might be > interested in becoming involved with future development of the program. > I intend to release nearly all of the source code for WSJT under the > GPL General Public License. (One piece of WSJT code has been licensed > under a non-disclosure agreement; however, linking to the object code > for this function will still be available to anyone.) But linking to object code is no use with later versions of Windoze. Sooner or later that will break - Microsoft will ensure that. It's no use on different operating systems (Linux for example) or for those people like myself that don't run Intel compatible CPUs. It would stop someone implementing it on a DSP. It might well start breaking with 64-bit processors ... Sooner or later (and probably sooner), the object code will be useless, the executables will not run on the later computers and it would all get left. Could you make the algorithm (not the code) public, then someone else can write code to implement that algorithm? Or if that is not possible (and I guess that is the problem), is it really necessary to use the same algorithm? I did read a paper (I assume written by you) about the software. I understand you have basically two bits in your code a) Coherent detection. b) Some 'compression' (for want of a better word) that packs the most amount of data typical in a QSO into as few a bits as possible, by avoiding the need to encode sequences that will not be in a QSO. From what I read, I felt there was enough information for someone else to build a similar program, based on the same basic principles, but using different algorithms. Of course it would be incompatible, but if it worked, and could be improved by access to the full source code, it would soon be better than the algorithm covered by the NDA. -- David Kirkby, G8WRB Please check out http://www.g8wrb.org/ of if you live in Essex http://www.southminster-branch-line.org.uk/