Re: commit 12fe78825798191ecb7e5a4ee3064679773ae1ab is broken

[James Lowe]

Knut,

On Sat, 9 Jun 2018 23:19:01 +0200, Knut Petersen <address@hidden> wrote:

> Hi James!
> > Checking out the commit before any of those three that Malte checed in and 
> > then I built LilyPond and simply ran
> >
> > lilypond -png $LILYPOND_GIT/input/regression
> 
> encodingdefs.ps is unused if neither '-O TeX' nor '-O TeX-GS' is used on the 
> command line.
> 
> Checkout, make and install 12fe788,  create a file test.ly:
> 
>     \version "2.21.0"
>     {c''-.}
> 
> Then compile with
> 
>     lilypond -O TeX --png --pdf test.pdf

Thanks.

I appreciate the time you have taken with me to explain this.

I went back and re-did the 'tests' I would have done with Malte's patch.

And now I can see it - the incorrect pizzicato glyph does show up. I simply 
must have missed this when testing Malte's patch.

All I can do now is apologise to you (for all the extra explaining I made you 
do) and to Malte for not spotting this at the time and the rest of the dev team 
allowing this through. I guess I am just so used to using the test-distance reg 
diff showing up only to show me the tests completed 'properly' - i.e if I don't 
see this reg test diff show up I know the patch has gone wrong.

Thanks again (and to Phil for pointing out the 'bleeding' obvious).

At least the testing methodology is not fundamentally broken.

James