Boost logo

Boost :

From: Vladimir Prus (ghost_at_[hidden])
Date: 2004-12-29 10:27:52


Jeff Garland wrote:

> On Wed, 29 Dec 2004 17:32:44 +0300, Vladimir Prus wrote
>> Jeff Garland wrote:
>>
>> > On Wed, 29 Dec 2004 16:00:12 +0300, Vladimir Prus wrote
>> >> On Wednesday 29 December 2004 13:04, Vladimir Prus wrote:
>> >> > Hi,
>> >> > the date_time, specifically the libs/date_time/xmldoc directory,
>> >> > includes the catalog.xml file:
>> >
>> > This is an internally used file, so it shouldn't cause errors - but
>> > perhaps an
>> > error has crept in? Are you seeing a problem when you try and generate
>> > the
>> > overall boost results?
>>
>> No, I see a problem when trying to build date_time docs.
>
> Maybe you can be more explicit about what you are doing and the errors you
> are
> seeing.

Sure.
1. I have up-to-date CVS tree.
2. I run "bjam --v2" in libs/date_time/xmldocs and get:

xslt-xsltproc bin/gcc/debug/date_time.docbook
../../../tools/boostbook/dtd/boostbook.dtd:412: warning: failed to load
external entity "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd"
%DocBook;
         ^
Entity: line 1:
(many screens of similar errors omitted).
3. I remove catalog.xml and run bjam again. Everything works.

> BTW, we haven't been able to build docs for quite awhile now

I know it was broken for about a week (because I broke it), but it should be
working now.

> and
> our
> requests for help have gone unanswered on the boost-doc list.
> We weren't
> sure what magic was applied to do the 1.32 release.

I'm not sure what problems you refer too, sorry.

>> >> And date_time_autodoc.boostbook/ and other *.boostbook files should
>> >> not be in CVS, too.
>> >
>> > Yes, well, since we haven't been sucessful in getting all of the
>> > doxygen to boost-book bridge to work well for date-time we have to
>> > slightly modify the
>> > generated files for our purpose. So that's why those are there.
>>
>> You mean that each time after modifying headers, you have to
>> manually tweak the generated files?
>
> That's correct -- although I think we actually use a script to do it.
> Sorry to say I've been hands-off of this for awhile so I don't know the
> details. Hopefully Bart will post later and describe the details.

Ok.

- Volodya


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk