cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Crossley <>
Subject Re: BRANCH_2_1_X fails
Date Wed, 12 Jan 2005 04:20:40 GMT
Antonio Gallardo wrote:
> Ralph Goers dijo:
> > 1) If the problem has existed for ages why haven't I ever seen it before?

The issue has been masked. See below.

> > 2) The Php block is disabled by default.
> > 3) I don't see anything remarkable about, but then I
> > have no idea what the qdox parser is trying to do to it.  It is dying on
> > the end of file.

It is reading the java source to extract documentation.
No idea why it fails on that file.

> > 4) Telling folks that the default build doesn't work is not a long term
> > solution.


> > Am I the first person to encounter this?  I've noticed a slew of commits
> > over the last several days so I find it hard to believe noone else has
> > tried to make the current SVN build.
> Well, things happens. :-D
> Seriously, personally I prefer to turn off the docs and non-used blocks
> while testing and making change to save time. Is not funny to wait 7-15
> minuts until a full build run.

Yes, that is probably what we all do. Especially the javadocs.

Having docs switched off is probably another reason why nobody
bothers to edit or generate them.

> Anyway, we agree this issue MUST be solved.

Agreed. And then we need to refine our whole build.xml
and tools/targets/*-build.xml ... There are lots of
uneccessary things happening, e.g. why does ./
need to include the prepare-docs.

> Let me see the changes on the build docs process and where the problem
> might be. ;-)

There are two issues.

* causes some strange failure.

I have seen that one for ages. It recently started being shown
because i enabled part of the "build docs" which uses Qdox to
scan the blocks to generate sitemap component documentation.
Forgot the side-effect sorry. ... disabled it again.

* The "does not implement a sitemap component interface" error.
This one is too wierd. I will start a new thread to discuss it.


> > David Crossley said:
> >> Antonio Gallardo wrote:
> >>> Workaround: disable doc generation in

View raw message