Issue 5736: Fix input/regression/context-find-parent.ly

https://sourceforge.net/p/testlilyissues/issues/5736/

Enable the previously reviewed block of code[1] required to fix this regression test.

[1] https://codereview.appspot.com/579250043/diff/565550049/lily/context.cc#newcode184

[Dan Eble, 2020-02-05 21:30:35]

Regtest differences attached to the ticket are expected.  I would appreciate
independent confirmation; is anyone interested in taking a little time to
understand the case? After that, I think it will make sense to push this, since
the code was reviewed last week and is just enabled by this patch.

[hanwenn, 2020-02-06 09:59:26]

can you add a pointer to the commit that introduces the previous code? I'm in
the middle of a laptop switch , so I can't look it up right now.

[Dan Eble, 2020-02-06 14:47:36]

On 2020/02/06 09:59:26, hanwenn wrote:
> can you add a pointer to the commit that introduces the previous code? I'm in
> the middle of a laptop switch , so I can't look it up right now.

I've added this link to the description of the current review.
https://codereview.appspot.com/579250043/diff/565550049/lily/context.cc#newco...

[hanwenn, 2020-02-07 11:27:10]

LGTM

nit: please use the commit SHA1 for referencing previous code. It is
self-contained, and doesn't need the internet for understanding the context.

[hanwenn, 2020-02-07 11:27:56]

(why is this review closed?)

[Dan Eble, 2020-02-07 14:25:08]

On 2020/02/07 11:27:56, hanwenn wrote:
> (why is this review closed?)

Jonas reviewed the results and I pushed the change and closed the review.

[dak, 2020-02-07 14:44:46]

hanwenn@gmail.com writes:

> LGTM
>
> nit: please use the commit SHA1 for referencing previous code. It is
> self-contained, and doesn't need the internet for understanding the
> context.

For referencing previous issues, we have so far used the issue number,
so that would be rather something to discuss on the list rather than as
a side note.

We have been able to transport the issue database across several
different systems, so the issue numbers have proven a comparatively
workable reference frame.  Commit ids work just in Git.

For information strictly related to a single commit, like when reverting
a commit, a commit id is certainly a good reference but at least in the
case of reverts, the issue number contained in the original commit
message is also present.

In addition, issue numbers tend to be associated with reports showing
the history and rationale behind a change and often giving extensive
other information like problem images.  That's something we cannot
attach to commits (or at least, this has not been attempted so far).

-- 
David Kastrup

[hanwenn, 2020-02-07 15:01:13]

On Fri, Feb 7, 2020 at 3:44 PM David Kastrup <dak@gnu.org> wrote:

> > nit: please use the commit SHA1 for referencing previous code. It is
> > self-contained, and doesn't need the internet for understanding the
> > context.
>
> For referencing previous issues, we have so far used the issue number,
> so that would be rather something to discuss on the list rather than as
> a side note.
>
>
I'll do that.

-- 
Han-Wen Nienhuys - hanwenn@gmail.com - http://www.xs4all.nl/~hanwen