On 05/21/2011 05:54 AM, Juan Pablo Carbajal wrote: > On Sat, May 21, 2011 at 12:52 AM, David Bateman <dbate...@free.fr> wrote: >> On 05/19/2011 05:45 PM, Juan Pablo Carbajal wrote: >>> Dear all, >>> Any news on this? Thanks >>> >>> I cannot install the communications package (octave 3.5.0+). >>> >>> octave:133> pkg install -forge communications >> >> Yeah I haven't updated for Octave 3.4. I've pushed a fix to the SVN and >> will send you a copy of the updated package offline to test. I've never >> uploaded an octave-forge package with the new means of doing it so can >> you report back on the list on whether this package works and I'll ask >> Soren to help me upload it. >> >> D. >> >> ------------------------------------------------------------------------------ >> What Every C/C++ and Fortran developer Should Know! >> Read this article and learn how Intel has extended the reach of its >> next-generation tools to help Windows* and Linux* C/C++ and Fortran >> developers boost performance applications - including clusters. >> http://p.sf.net/sfu/intel-dev2devmay >> _______________________________________________ >> Octave-dev mailing list >> Octave-dev@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/octave-dev >> > > Installation went OK, > Here the warnings > > warning: dispatch is obsolete and will be removed from a future > version of Octave; please use classes instead > warning: /home/juanpi/octave/communications-1.0.11/deintrlv.m: > possible Matlab-style short-circuit operator at line 40, column 39 > warning: /home/juanpi/octave/communications-1.0.11/lz77deco.m: > possible Matlab-style short-circuit operator at line 36, column 14 > warning: /home/juanpi/octave/communications-1.0.11/lz77enco.m: > possible Matlab-style short-circuit operator at line 35, column 14 > warning: /home/juanpi/octave/communications-1.0.11/lz77enco.m: > possible Matlab-style short-circuit operator at line 70, column 21 > warning: /home/juanpi/octave/communications-1.0.11/marcumq.m: possible > Matlab-style short-circuit operator at line 37, column 18 > > It also requires image >= 0.0.0 is this a bug or is really required? > > This installation fixed the dependencies of the packages "signal" and "ga"
Getting rid of the short-circuit warnings is easy, as these are just | to replace with || and & with && (in code I didn't write btw). Though fixing the dispatch warning is going to be a pain. Its easy enough to fix this for the m-files, though for the oct-files I have a couple of issues. The build process so that the oct-files concerned end up in a @gf or @galois directory? And, are classes going to play nicely with an autoloaded oct-file, as the oct-files in the comm packages have multiple functions in them. I started building a replacement for the galois class in the comms package entirely in m-files a while back (see the attached files if your interested), but suspect some functions, ginv for example, will always be better off as oct-files. So, completing, testing and documenting this code always seemed more work than it was worth, as the existing code worked. I'm still think about what to do for this As for the dependencies, there sure is a lot of them now for the communications package more than I remember there being. It would be better to try and rationalise where certain functions in octave-forge are keep so as to reduce the number of dependencies. David
galois.tar.bz2
Description: BZip2 compressed data
------------------------------------------------------------------------------ What Every C/C++ and Fortran developer Should Know! Read this article and learn how Intel has extended the reach of its next-generation tools to help Windows* and Linux* C/C++ and Fortran developers boost performance applications - including clusters. http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev
