get rid of have_ring option in singular interface

[fchapoton]

We get rid of have_ring option in singular interfaces, which is already out of use, as suggested in #2331.

Component: commutative algebra

Author: Frédéric Chapoton

Branch/Commit: 8dc327a

Reviewer: Kwankyu Lee

Issue created by migration from https://trac.sagemath.org/ticket/33949

[fchapoton]

New commits:

8dc327a

get rid of "have_ring" in singular interface

[fchapoton]

Branch: u/chapoton/33949

[fchapoton]

Commit: 8dc327a

[kwankyu]

comment:2

What is the rationale?

(1) Setting Singular ring is not expensive anymore.

(2) have_ring is not set True anywhere in Sage, and practically not used.

(3) Both.

[fchapoton]

comment:3

I think the second point holds:

git grep -h "have_ring\s*=" src/
    def _singular_(self, singular=singular_default, have_ring=False, force=False):
        return self._singular_init_(singular, have_ring=have_ring)
    def _singular_init_(self, singular=singular_default, have_ring=False, force=False):
    def _singular_(self, singular=singular_default, have_ring=False):
    def _singular_(self, singular=singular, have_ring=False):
    def _singular_init_func(self, singular=singular, have_ring=False):
def _singular_func(self, singular=singular, have_ring=False):
#    return self._singular_init_(singular, have_ring=have_ring)
    return _singular_init_func(self, singular, have_ring=have_ring)
def _singular_init_func(self, singular=singular, have_ring=False):
    def _singular_(self, singular=singular_default, have_ring=False):
    def _singular_(self, G=None, have_ring=False):
    def _singular_init_(self, have_ring=False):

[kwankyu]

comment:4

#2331 seems to say that before libSingular, set_ring() was expensive, and have_ring option was useful, but after libSingular, set_ring() is a breeze, and have_ring is useless.

[kwankyu]

comment:5

Hence you can also delete the comment # this is expensive...

[sagetrac-git]

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

999e8ef

remove the comments about "being expensive"

[sagetrac-git]

Changed commit from 8dc327a to 999e8ef

[fchapoton]

comment:7

I removed the comments, although I am not totally convinced that we use libsingular everywhere. There could be remaining things still using the singular pexpect interface, maybe.

[kwankyu]

comment:8

Replying to @fchapoton:

I removed the comments, although I am not totally convinced that we use libsingular everywhere.

Me neither.

There could be remaining things still using the singular pexpect interface, maybe.

It seems code like _singular_(singular).set_ring() is using the singular pexpect interface. It is really expensive. On my system, it takes about 100 microsecond. Hence those comments are still relevant. Sorry...

[kwankyu]

Reviewer: Kwankyu Lee

[sagetrac-git]

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

[sagetrac-git]

Changed commit from 999e8ef to 8dc327a

[fchapoton]

comment:11

Here is a new branch with the comments restored

[kwankyu]

comment:12

LGTM. Thanks.

[vbraun]

Changed branch from u/chapoton/33949 to 8dc327a