Update rubiks' patches to conform to same format as other patches

[jdemeyer]

The rubiks package is copying files instead of proper patching. Clean this up.

See also #20837 and #20933 which did the same for other packages.

CC: @embray

Component: packages: standard

Author: Erik Bray, Jeroen Demeyer

Branch: 87963a9

Reviewer: Jeroen Demeyer, Erik Bray, Matthias Koeppe

Issue created by migration from https://trac.sagemath.org/ticket/21103

[jdemeyer]

Branch: u/jdemeyer/update_rubiks__patches_to_conform_to_same_format_as_other_patches

[7ed8c4ca-6d56-4ae9-953a-41e42b4ed313]

Branch pushed to git repo; I updated commit sha1. New commits:

a1a09b8

Force rebuild of rubiks

[7ed8c4ca-6d56-4ae9-953a-41e42b4ed313]

Commit: a1a09b8

[jdemeyer]

Description changed:

--- 
+++ 
@@ -1 +1,3 @@
+The `rubiks` package is copying files instead of proper patching. Clean this up.

+See also #20837 and #20933 which did the same for other packages.

[jdemeyer]

comment:5

I am not really convinced that the non-patched Makefiles are fine. Surely, the person who patched them must have had a reason for it. There is for example this comment in build/pkgs/rubiks/patches/dietz-solver-Makefile:

# This Makefile was seriously broken.
# CC was set to g++. Since it was compiling C++ files,
# CXX should have been used.
# LINK was set to g++, so I changed that to LD
# CFLAGS was set to -O2. I've removed that, so it can be set
# in spkg-install.
# In any case, it should have been CXXFLAGS
# LFLAGS and INCLUDES were both empty
# David Kirkby, 29th Sept 2009

[embray]

comment:6

Compare to the current source. For the makefiles I removed from the patch there is no longer any difference from upstream (which has probably been patched).

[embray]

comment:7

Actually scratch that, I was confused. I don't think those patches were supposed to be removed. You can update the commit to not remove the makefile patches (just the full makefile copies).

[mkoeppe]

comment:8

With the branch on this ticket, sage -f rubiks and sage -t src/sage/interfaces/rubik.py go through (on Mac OS X).

What is considered upstream for this package? Should there be an spkg-src? The links in http://doc.sagemath.org/html/en/reference/interfaces/sage/interfaces/rubik.html and in SPKG.txt are broken.

[embray]

comment:9

I think all this needs to is roll back the removal of

• build/pkgs/rubiks/patches/dietz-cu2-Makefile.patch
• build/pkgs/rubiks/patches/dietz-mcube-Makefile.patch
• build/pkgs/rubiks/patches/dietz-solver-Makefile.patch

from the commit. They shouldn't have been removed.

[7ed8c4ca-6d56-4ae9-953a-41e42b4ed313]

Changed commit from a1a09b8 to 87963a9

[7ed8c4ca-6d56-4ae9-953a-41e42b4ed313]

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

fb539e1	Clean up patching of rubiks
98cec94	Fix patch loop
87963a9	Force rebuild of rubiks

[jdemeyer]

comment:12

This now works for me. If you're happy with this, then you can it to positive_review.

[jdemeyer]

Changed author from Erik Bray to Erik Bray, Jeroen Demeyer

[jdemeyer]

Reviewer: Jeroen Demeyer, Erik Bray

[mkoeppe]

comment:13

comment 8.

[jdemeyer]

comment:14

Replying to @mkoeppe:

comment 8.

I don't really care about that. It's unrelated to this ticket anyway.

[mkoeppe]

comment:15

Replying to @jdemeyer:

I don't really care about that. It's unrelated to this ticket anyway.

Fine, I've created a new ticket #21493 for that.

[mkoeppe]

Changed reviewer from Jeroen Demeyer, Erik Bray to Jeroen Demeyer, Erik Bray, Matthias Koeppe

[vbraun]

Changed branch from u/jdemeyer/update_rubiks__patches_to_conform_to_same_format_as_other_patches to 87963a9

[embray]

comment:18

Thanks!

[embray]

Changed commit from 87963a9 to none