https://issues.apache.org/bugzilla/show_bug.cgi?id=37579





--- Comment #64 from Vincent Hennebert <vhenneb...@gmail.com>  2009-05-01 
04:31:47 PST ---
Hi Dimitri,

(In reply to comment #63)
> (In reply to comment #62)
> > Hi Dimitri,
> > 
> > (In reply to comment #61)
> > > Hi Vincent,
> > > 
> > > thank you for the patch. This time another issue with a wrong order of
> > > footnotes. There is a two column table in the attached example, both 
> > > columns
> > > have footnotes. Sometimes the footnote from the second column precedes the
> > > footnote from the first one. If you delete one block from the first 
> > > column, the
> > > order will be right.
> > 
> > It all depends on what order you should be expecting. If you scan the page 
> > in
> > its whole width starting from the top you will find the footnote labeled 2
> > before the footnote labeled 1. This is basically what FOP is doing.
> > 
> > Of course, it may seem more natural to start from the leftmost column, then 
> > go
> > to the following one, etc. But this is particular to that case. With a
> > right-to-left language it will be more natural to start from the rightmost
> > column. Sometimes, the content will be such that the method above will be 
> > more
> > natural.
> > 
> > So this is a grey area, and the Recommendation doesn't say anything about 
> > that.
> > Your best bet is to re-number the footnotes. Or use something else than
> > footnotes (you may be happy with putting the notes in regular blocks just 
> > after
> > the table, for example).
> > 
> > HTH,
> > Vincent
> 
> Hi Vincent,
> 
> I understand your point of view, you are probably right. Anyway, current
> implementation is not very reliable. If you leave only two blocks in the left
> column in my example, footnote labeled 1 will be output first, although,
> according to FOP behavior, we can expect it should be footnote labeled 2.

The change I made introduced another bug. It should be fixed now in revision
770635 ( https://svn.apache.org/viewcvs.cgi?view=rev&rev=770635 ). Sorry about 
that.

That said, I can think of certain situations involving row-spanning cells where
the basic 'rule' stated above does no longer hold. I won't enter the details
because they are a bit technical, but interesting issues may arise regarding
accessibility, order of reading, etc. (That was a note to self :-) )

Vincent

-- 
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

Reply via email to