Re: Linux build currently not building

Subject: Re: Linux build currently not building
From: Caolan McNamara (
Date: Mon Dec 20 1999 - 05:01:52 CST

On 19-Dec-99 Justin Bradford wrote:
>> Also what is the story with headers and footers ? I will be able to extract
>> these without too much work, but how to present them to abi ? There are
>> different headers and footers for the first page, odd pages even pages,
>> first page of a section etc etc. What would the desired api look like for
>> you ?
>Perhaps you could fully describe the possibilites for headers/footers (are
>there any special cases to consider beside what you've mentioned above?)
>and then we would settle on a solution which would allow us to deal with
>all of those things.
>I imagine that a new element could work, if you can provide
>headers/footers in sequence with the document.

Lemme see now, each section of the document can have different headers and
footers, for each section you have

header and footer for the first page
header and footers for odd pages
header and footers for even pages

but for each section these headers and footers include the
normal ones which go between main text and footnote text,
plus ones for

"below footnote text on a page when footnote text must be
continued on a succeeding page"
"above footnote text on a page when the text must be
continued from a previous page"

And these three types are duplicated in relation to
endnotes which are the same as footnotes, only they go
at the end of a section rather than at the end of a page.

To basically there are a mountain of possible headers and
footers available to word.

>We'll also need to know how this information is added to the piece table.
>Do I have to provide headers/footers in the order of the document? Can
>they be lumped on all together at the end?

The headers and footers are stored in a seperate subdocument which comes
at the end of all the rest of the text, So I suppose we could just output
all the main text, and then output a marker specififying each of the header
and footer types that are coming next, i.e.
all the usual stuff
head text, with something to specify the type of the header
a different type of header.
etc etc....

>> textboxes, floating box regions outside the usual flow of text (i was
>> thinking of using layers in html to get the idea across (cringe)),
>Again, a new element delimiting this region might work...

Things get a little catchy with textboxes, which follow the same pattern
as graphics, i.e. they are linked to the main text, and should appear
there instead. While their text can be found in a subdocument I would
imagine that they should be noticed in the charHandler, and something
should be called to get the text of each textbox ??, headers and footers
could be handled this way as well if that would suit abi better ?

>> background colors and graphics for the document.
>I'm not sure what the details are here...
>Also, on images, is the code in place to provide memory buffers of the
>data, rather than temp files? Also, does the API provide for format
>selection, or is that a compile-time option?

I can provide a memory buffer in the next checkin then, but the *solid* api
just returns the escher structure that contains the graphics. I can't
really guarantee more than that at this moment, The escher structure
may contain a reference to a bitmap or wmf graphics which can be used, but
the strucutre could also contain some vector command to draw some text
on top of the bitmap which wv has implemented nothing of. In effect I have
an escher reader and parser, but not an escher drawer. I get away with it
in the vast majority of cases, because the vast majority of escher use
examples are simple bitmap containers with no escher drawing features in



Real Life: Caolan McNamara * Doing: MSc in HCI
Work: * Phone: +353-86-8790257
URL: * Sig: an oblique strategy
Use an old idea

This archive was generated by hypermail 2b25 : Mon Dec 20 1999 - 05:11:38 CST