Re: kainhofer docs not updating

[Graham Percival]

On Sat, Sep 05, 2009 at 12:35:06AM +0200, Reinhold Kainhofer wrote:
> Am Samstag, 5. September 2009 00:12:44 schrieb Trevor Daniels:
> > Reinhold Kainhofer wrote Friday, September 04, 2009 10:51 PM
> > > It seems that the NR was really last compiled on September 1.
> > > The dependencies are all listed in out-www/notation.dep...
> >
> > It was the NR I looked at.  I've pushed several changes
> > since 1 Sep for Joseph Wakeling:

For Trevor: the only way *you* can "touch" a file is by making a
git change to it.  So if you really wanted to make the NR rebuilt,
you'd have to push a nonsense patch (i.e. adding/subtracting one
blank line) to notation.tely.

Reinhold and I can "touch" a file by logging into kainhofer and
doing it manually.

> But the NR was not built, even though the dependencies were
> changed. Something seems to be broken with the build system...

Yes; it's been this way for about a month now.  It was previously
discussed on -devel, although there's no much point searching for
those emails.

Basically,
1) the doc reorg (notably the AU) is almost finished.  Probably
  done by next Monday.  At that point, I'll invite Trevor to
  run his broken link-checker thingie.
2) the doc build system works for a clean "make doc", with some
  warnings.  It is currently *not* guaranteed to work for a
  second "make doc" without a "make doc-clean".
3) we might be switching to SCons "soon", so there's unfortunately
  little interest in tracking down the above problems.  As always,
  patches appreciated if you're particularly curious about this
  issue.
4) John is the main person for #2 and #3, but he's
  sorting out issues with his upcoming PhD in Italy.  He might
  also be involved in the EU by-election in France.
  (admittedly, that's a pure guess: Valentin is involved, and
  since John lives in the same country as him, he might also
  be involved.  Not the most stringent logical argument I've ever
  constructed.  :)
5) GUB isn't working.  When I arrive in Glasgow next week, I'll bug
  my supervisor for a desktop computer so I can investigate more.
  If I can't fix it easily, I'll go back to the git from late June
  and make releases from that.
6) I'll also have more time for #5 once I'm finished the doc reorg
  in #1.  And I might also investigate #2/#3.


In summary, there *is* a plan for dealing with these problems.
It's probably not as fast as you would like, and it's certainly
not as fast as *I* would like, but it's the best I can make given
the resources (both people and computers) at my disposal.

Cheers,
- Graham