P.S.
Le 13/08/2022 à 16:27, Thomas Morley a écrit :
Especially the lack of C++-knowledge[*] sometimes hinders me to get to
real and general working solutions for certain problems, p.e. if I
need to look at several different grobs to set a certain behaviour of
the original overriden grob and the needed-to-know properties of those
other grobs are done as unpure-pure-containers.
Frankly, the main problem with unpure/pure is not being mostly
baked on the C++ side.
The main problem is that the design of this stuff is just tricky,
and the implementation even more. It got better with the
introduction of unpure/pure containers by Mike, but still a long
way from a panacea.
I have tried to simplify it, yes. I've made some progress, but
still far from done. When I try a new approach or do something
a bit experimental that I might want to discard, I begin a new branch.
Now, I have
~/repos/lilypond $ git show --quiet purity
commit d062b6f783f84a6bacde950c481d2fcb2eadccf1 (purity)
Author: Jean Abou Samra <j...@abou-samra.fr>
Date: Fri Oct 15 21:21:31 2021 +0200
purity WIP
~/repos/lilypond $ git show --quiet purity2
commit 615fd9b7e376c0064d8646de9c24587afb35477e (purity2)
Author: Jean Abou Samra <j...@abou-samra.fr>
Date: Sun Oct 24 22:45:43 2021 +0200
WIP purity2
~/repos/lilypond $ git show --quiet purity3
commit 1a63cde978234416028d2c179f24c8e03046ca6c (purity3)
Author: Jean Abou Samra <j...@abou-samra.fr>
Date: Thu Feb 3 23:48:42 2022 +0100
WIP: assumptions
~/repos/lilypond $ git show --quiet purity4
commit c13a5be7c0b0c56d79e8e6f02a55b0c57fdb6c64 (purity4)
Author: Jean Abou Samra <j...@abou-samra.fr>
Date: Thu Feb 3 23:48:42 2022 +0100
WIP: assumptions
~/repos/lilypond $ git show --quiet purity5
commit 58d8ca1b65a5914a17731a4dbb65b5716d9abf57 (purity5)
Author: Jean Abou Samra <j...@abou-samra.fr>
Date: Sat Apr 16 14:09:29 2022 +0200
WIP: assumptions (umpteenth take)
:-)
