Lists Home |
Date Index |
Eric van der Vlist wrote:
> On Fri, 2004-04-09 at 16:25, Robin Berjon wrote:
>>I've gotten into the (at least occasional) habit of using 'optimised'
>>instead of 'binary' because I don't think the WG should presume that the
>>solution -- if there is one -- necessarily relies on binarisation (even
>>if all tests I've seen thus far makes it look very likely).
> As Mike Champion said, optimised alone doesn't mean much (an
> optimisation is always done for a specific purpose and depending what
> you optimise, an "optimised XML" could be more verbose than XML).
> What about "compacted" XML if the target is to optimise for space?
The point is precisely that different people want to optimise for
different purposes, and indeed some of the formats that were submitted
at the workshop were larger than XML but faster (if your I/O is not a
problem of course).
That's a large part of the XBC WG's work: finding what people want to
optimise, figuring out is those optimisations are compatible between
themselves, and if a good set of interesting optimisations are found to
work correctly together (given some trade-offs) look at how many people
are still interested versus those that are not because their needs are
insufficiently addressed. That's where the fears of those that think
that the WG is composed only of pro-(binary|optimised) people -- which
isn't true anyway -- and that the answer to that question is foretold
are unfounded: I anticipate that it will be difficult to reach consensus
And then of course there's that small question of whether creating
another format would hurt XML at all, or would hurt it more or less than
not doing it... ;)