[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: fact check for manual: devel bugs
From: |
Erik Sandberg |
Subject: |
Re: fact check for manual: devel bugs |
Date: |
Sun, 12 Sep 2004 13:02:40 +0200 |
User-agent: |
KMail/1.6.2 |
On Tuesday 07 September 2004 01.16, Graham Percival wrote:
> Should bugs in the devel version be sent to lilypond-devel,
> bug-lilypond, or both?
bug-lilypond, I would say. But it doesn't matter, I read both lists. As long
as you add [bug] in the subject line for devel bugs, there is no risk that I
miss it.
> While we're on the subject, could Erik check the section "reporting
> bugs" chapter 4 of the manual (after the next release is ok; it's not
> urgent) ? I've added some more info, but since you're in charge of
> this stuff, you should check it. :)
OK.
-------------------
When you've found a bug, do a few searches on the mailing list for the bug.
Sometimes the bug will have already been reported and a fix or workaround is
already known.
Here's an example of a good bug report:
% LilyPond 2.3.11, Mac OSX 10.3.4, fink package lilypond-unstable
% slur does not look at accidentals
\score {
\relative c''{
a1 a a a a
a2. g16( b d fis)
}}
-------------------
Some issues:
- The link to bug-lilypond archive is
http://lists.gnu.org/archive/html/bug-lilypond/
However, I think it is even more relevant to link to the pre-built bug CVS at
http://lilypond.org/doc/v2.3/bugs/
- OS info is not very relevant to include in the .ly file. In CVS, the report
will only contain the version number; all bugs are verified on my
debian/testing system.
- A \version statement must be included.
- If you want to be picky, \header { texidoc = "blablabla" } is slightly
better than % blablabla. But that's just relevant for regular bug reporters
like Werner so you can skip that, it's too complicated for first-time bug
hunters.
- It is appreciated (but not at all required) to attach a small .png image
showing the problem (the mail list strips attachments, so in this case CC:
the message to me).
So: In my eyes, the following is a good bug report:
It seems that the slur engine doesn't look at accidentals. In the following
example, the slur touches the accidental.
Using Mac OSX 10.3.4, fink package lilypond-unstable
\version "2.3.11"
\relative c''{
a1 a a a a
a2. g16( b d fis)
}