Doug Stewart

On Tue, Apr 2, 2019 at 3:53 PM Doug Stewart dougstewart@users.sourceforge.net wrote: On Tue, Apr 2, 2019 at 3:07 PM Olaf Till i7tiol@users.sourceforge.net wrote: @Doug: Installation with 'pkg install ...' should work. For building manually, there are now two additional steps: ./bootstrap # only necessary at first time or after configure.ac has been changed ./configure (and then 'make', as before). config.h.in is generated by 'bootstrap' and is contained in the tarball. config.h is generated from...

On Tue, Apr 2, 2019 at 3:07 PM Olaf Till i7tiol@users.sourceforge.net wrote: @Doug: Installation with 'pkg install ...' should work. For building manually, there are now two additional steps: ./bootstrap # only necessary at first time or after configure.ac has been changed ./configure (and then 'make', as before). config.h.in is generated by 'bootstrap' and is contained in the tarball. config.h is generated from config.h.in by configure (which is automatically called by 'pkg install ...'). Regarding...

On Mon, Apr 1, 2019 at 3:26 PM Doug Stewart dougstewart@users.sourceforge.net wrote: On Mon, Apr 1, 2019 at 2:57 PM Olaf Till i7tiol@users.sourceforge.net wrote: Oh Doug, your fix for the tests in minreal came in between, so my posted and tested tarballs don't contain your new changeset, although your changeset is now earlier in the sequence than the changeset of the tarballs. It took me a long time to test, I won't start it again now... If we release the posted tarballs, the release is strictly...

On Mon, Apr 1, 2019 at 2:57 PM Olaf Till i7tiol@users.sourceforge.net wrote: Oh Doug, your fix for the tests in minreal came in between, so my posted and tested tarballs don't contain your new changeset, although your changeset is now earlier in the sequence than the changeset of the tarballs. It took me a long time to test, I won't start it again now... If we release the posted tarballs, the release is strictly not easily reproducible, in the sense that it lacks your changeset. But I'd suggest we...

On Mon, Apr 1, 2019 at 11:34 AM Olaf Till i7tiol@users.sourceforge.net wrote: give me 1 or 2 days, i'm about to make it compatible with octave 4.2 and 4.0 ok I will push a fix for minreal.m [package-releases:#369] https://sourceforge.net/p/octave/package-releases/369/ Control pkg new release please* Status: open Created: Mon Feb 25, 2019 08:01 PM UTC by Doug Stewart Last Updated: Sun Mar 31, 2019 08:56 PM UTC Owner: Doug Stewart I think I have done all that I know how. Now I would like someone to...

On Sun, Mar 31, 2019 at 4:56 PM Mike Miller mtmiller@users.sourceforge.net wrote: If they are simple lets do them. Either tell me how or i would be happy if you just do then for me. I don't have time to write and test a patch for control at the moment, but something like this: https://hg.mtmxr.com/octave-signal/rev/3f7403f4c241 If you or someone else is able to adapt that to control, great. If not and you want to make the release as is, then the minimum should be bumped to 4.4. I took a look at that...

On Sun, Mar 31, 2019 at 4:07 PM Mike Miller mtmiller@users.sourceforge.net wrote: According to Doug, the above, and probably also the fails observed by lostbard, are due to alternative signs and are harmless. Doug said he'll care for then (after the release). IIRC this sign difference is between netlib BLAS and OpenBLAS. Should the required Octave version be set to >=4.4 ? Making it compatible with <4.4 would amost certainly mean a considerable delay. IMHO that would be unfortunate. The changes needed...

On Sun, Mar 31, 2019 at 2:34 PM Olaf Till i7tiol@users.sourceforge.net wrote: I'd suggest to apply this patch: https://savannah.gnu.org/patch/index.php?9781 to avoid the warning for use of the deprecated LFLAGS in src/Makefile. With this patch, however, Apple machines must use an Octave version of at least 4.4. What do you say? I was going to try and fix this -- but -- then Mike said it was not that important and we could wait for octave 6.xx. So I was going to wait. I think it better to have this...