David Tolpin wrote:
> table subtotals; the problem is that the way it is specified in XSL 1.1
> is a disaster. A disaster that imposes a very limited choice on the
> implementer:
>
> - introduce restrictions and deviation from the specification into
> the implementation to provide the feature and still keep the software
> reliable
> - adhere to the recommendation and trap the user in border cases.
I can't speak in any official capacity for the FO Subgroup and not being
an implementor I can't comment on the implementation complexity with any
authority, but it is certainly true that table markers were probably the
most-requested feature for 1.1.
Should we have required that table headers have invariant BPD? Maybe we
should have--that would, I think, satisfy David's implementation
concern. But I think that doing so would make the feature too limiting.
FO 1.1 is not a rec yet so it's not too late to make this kind of
change, although it would require a champion sitting on the Subgroup....
The Subgroup certainly understands the potential processing problems
that table markers impose--at least one implementor member of the
Subgroup made comments very similar to David's, but the overall feeling
was that those edge cases were outweighed by the value of the feature
and the fact that most uses would not be edge cases (i.e., either add a
"continued" message to the table caption or row headers or doing
per-page subtotals).
To my naive implementation view, the problem doesn't seem any harder
than the problem posed by footnotes and floats. The only complication is
that it's in the context of a table, which already requires some sort of
two-pass process anyway, as far as I can tell. Certainly not an easy
thing to implement completely but not impossible either.
For example, I could live with a processor that for example makes a
maximum of 3 passes over a table and if, after that third pass, it can't
place content accurately, gives up. From the FO creator's standpoint,
the fix is easy in that case: either put the header content into a row
with fixed size or otherwise reduce the variability in the BPD extent of
the marker content.
Cheers,
Eliot
-------------------
(*) To unsubscribe, send a message with words 'unsubscribe xep-support'
in the body of the message to majordomo@renderx.com from the address
you are subscribed from.
(*) By using the Service, you expressly agree to these Terms of Service http://www.renderx.com/terms-of-service.html
Received on Mon Nov 7 07:31:19 2005
This archive was generated by hypermail 2.1.8 : Mon Nov 07 2005 - 07:31:23 PST