[prev in list] [next in list] [prev in thread] [next in thread] 

List:       lilypond-devel
Subject:    Re: SCM/bool confusion in One_page_breaking::solve (issue 314490043 by dak@gnu.org)
From:       paulwmorris () gmail ! com
Date:       2017-02-20 3:04:54
Message-ID: 94eb2c1bd28a1ed7280548ed872b () google ! com
[Download RAW message or body]

On 2017/02/18 20:20:27, dak wrote:

> Actually I didn't.  I just put the finishing touches on a task mainly
done by
> Valentin that made it possible to enable a debugging option in Guile,
and the
> compiler then dug this one up in consequence as well as a few others.

Ah, cool, nice to have these checks.  I gave it a try with the new
change to the code and it still works as intended.

I think the reason this did not show up as a problem before is that this
part of the code is a defensive measure against a rare edge case where
the last musical system (or top-level-markup) in the data structure does
not produce the lowest visual point of the 'music' on the page
(excluding footers, etc.).

I looked back at the test file I was using locally while working on
this, and found an example that seems to cover this edge case.  However,
it appears to work fine with or without the newest change to the code...
  So it's possible that this code is more defensive than it needs to be,
but it's also possible that removing the defensive measure might lead to
bugs in other edge cases, and it doesn't hurt to have it in there.  So I
think it's better to leave it as is and err on the defensive side.

I can upload a regtest for that edge case in another patch.

https://codereview.appspot.com/314490043/

_______________________________________________
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel
[prev in list] [next in list] [prev in thread] [next in thread]