> could the regtest diff only be due to a different rounded value?
That's unlikely.  The change is invasive in that it not only triggers in cases 
that threw an assertion failure previously: it does not replace a lower bound 
but adds a point to an interval: that's only equivalent to the previous 
behavior if the replacement would not have shrunk the interval.  I proposed it 
based on the handwaving guess that it would kick only in in sort-of 
pathological cases.

The problem now is that the regtest triggering shows that this change is not 
equivalent even in cases that are basic enough to be included in the regtests 
but did not previously trigger the problem: namely they shrink the interval but 
not to a degree of making it empty.

Since I was proposing this fix basically flying blindly, I don't know whether 
the change in behavior for cases that worked previously is likely a change to 
the better or worse.  Extending the nature of the fix to other locations would 
similarly warrant making this decision.

So someone who has a clue what this code is actually likely to be doing 
(Keith?) should pitch in before we extend it to other areas.  Likely even 
before we even commit in the current state.


---

** [issues:#5501] Avoid failed assertion in stem tremolo code**

**Status:** Started
**Created:** Mon Mar 25, 2019 11:08 PM UTC by David Kastrup
**Last Updated:** Thu Mar 28, 2019 07:11 PM UTC
**Owner:** David Kastrup


Avoid failed assertion in stem tremolo code

Example code triggering the problem:

    \relative { a32 8..:32 }

http://codereview.appspot.com/572550043


---

Sent from sourceforge.net because [email protected] is 
subscribed to https://sourceforge.net/p/testlilyissues/issues/

To unsubscribe from further messages, a project admin can change settings at 
https://sourceforge.net/p/testlilyissues/admin/issues/options.  Or, if this is 
a mailing list, you can unsubscribe from the mailing list.
_______________________________________________
Testlilyissues-auto mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/testlilyissues-auto
  • [Lilypond-... Auto mailings of changes to Lily Issues via Testlilyissues-auto
    • [Lily... Auto mailings of changes to Lily Issues via Testlilyissues-auto

Reply via email to