Re: [Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-07 Thread John Peterson
On Mon, Oct 7, 2013 at 11:07 AM, David Knezevic wrote: > On Mon, Oct 7, 2013 at 10:59 AM, David Knezevic < > [email protected]> wrote: > >Thanks for the head up re "git format-patch -1". I just did git diff > >> patch.txt (like in the old svn days) >> >> I have the code in my dknez/l

Re: [Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-07 Thread David Knezevic
On Mon, Oct 7, 2013 at 10:59 AM, David Knezevic mailto:[email protected]>> wrote: Thanks for the head up re "git format-patch -1". I just did git diff > patch.txt (like in the old svn days) I have the code in my dknez/libmesh fork as well, so I should've just pointed yo

Re: [Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-07 Thread John Peterson
On Mon, Oct 7, 2013 at 10:59 AM, David Knezevic wrote: > Thanks for the head up re "git format-patch -1". I just did git diff > > patch.txt (like in the old svn days) > > I have the code in my dknez/libmesh fork as well, so I should've just > pointed you there I guess. Anyway, thanks for setting

Re: [Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-07 Thread David Knezevic
Thanks for the head up re "git format-patch -1". I just did git diff > patch.txt (like in the old svn days) I have the code in my dknez/libmesh fork as well, so I should've just pointed you there I guess. Anyway, thanks for setting up the new branch, it'll be great to get this thing sorted so

Re: [Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-07 Thread John Peterson
OK, this patch has been pushed as the 'complex_io' branch. BTW, how did you make your patch? git format-patch -1 is probably the simplest, most terse way to make a patch file from the most recent local commit. The one you sent wouldn't apply with 'git am' so I used 'git apply,' but that doesn't

[Libmesh-devel] First attempt at fixing output for --enable-complex

2013-10-06 Thread David Knezevic
I've attached a patch that I've tried for fixing output with complex values (it doesn't work yet). The approach I've used is based on Paul's suggestion in issue #47 on github. I tested it with introduction_ex4 (using Exodus format) and it gives incorrect solution fields. It seems like the pr