Re: Release news, development news, state of the world

[David Kastrup]

Graham Percival <address@hidden> writes:

> STABLE RELEASE
>
> Almost everybody wants stable releases more frequently.  They
> attract positive attention, they get updated docs and bugfixes and
> new features into the hands of users, etc.  We had the first two
> release candidates back in Sep.  Unfortunately, we've had Critical
> issues since then.
>
> http://code.google.com/p/lilypond/issues/list?can=2&mode=grid
> There's one spacing regression (found 42 hours ago) which can
> probably be fixed easily.  But there's also 3 problems in GUB that
> have seen no action for a month or more.
>
> I know that some people tried to work on GUB but found that they
> couldn't build it on ubuntu 11.10.  It's just possible that the
> regression in gcc 4.6 was to blame...

For 32bit x86 pretty likely.  With the current sources, it should no
longer cause problems.

So there is _definitely_ room for people to get moving again on GUB.

Here are presumably the three critical issues Graham is talking about:

1948    Windows install clobbered system PATH    
<URL:http://code.google.com/p/lilypond/issues/detail?id=1948>
1943    lilypond after 2.15.8 fails on x86 Macs   Regression     
<URL:http://code.google.com/p/lilypond/issues/detail?id=1943>
1933    Lilypond-book requires msvcrt again   Regression
<URL:http://code.google.com/p/lilypond/issues/detail?id=1933>

All of these are on non-GNU platforms.  Now for users of those
platforms, participating with development work is often not the most
attractive task they can think of, and I can't exactly blame them.  But
if we are not supposed to cut off the non-GNU user base, we need them to
take an interest.

> but without any action on GUB, we don't really know.  Given the lack
> of interest[1] in stable releases, I guess that 2.16 might occur some
> time in summer 2012?
>
> [1] recall that I define "interest" as "people submitting or
> discussing patches".

> I'm planning a big recruitment drive for new contributors in Jan
> 2012.  It would be *really* helpful if we got as many "force
> multipliers" (i.e. "maintainability" issues in the tracker)
> resolved before then.

I am currently back into work on the parser.  I would like to get the
music function argument parsing business into a shape where there is no
need to document strange rules or special cases or shortcomings.  That
is a maintainability issue for me and the current skill set distribution
does not make it sensible to have anybody else work on that.

I think we need to recruit people willing to work on the GUB (Grand
Unified Builder, cf. <URL:http://lilypond.org/gub/> now.  I am copying
the user list for that reason.

-- 
David Kastrup