[LINK] XML is evil

[Craig Sanders]

On Thu, Nov 20, 2008 at 05:26:47PM +1100, Rick Welykochy wrote:
> What I liked about the old way we stored MARC data: it was in "flat"
> format stored in table columns. Easy to search, easy to index. And
> easy to import/export using CSV.
> 
> Now with XML-MARC, the entire library record is stored in one
> monolithic blob in MySQL. Difficult to index and difficult to
> search. The reason: they are adapting to using the Zebra indexing
> engine which utilises the MARC format.

now that's one of the things i mean when i say "misuse of XML".

XML is *NOT* a data storage format. it's a data transfer format. it
makes no sense whatsoever to use a database engine like mysql to store
data in an XML blob rather than in database fields, records, and tables.

IMO, that constitutes pure idiocy. there's not even anything slighly
good or sensible about that idea. it's missing the point and nature of
both XML and databases.


where XML makes sense is when you need to transfer data, especially
structured data, from one system (e.g. a database) to another (e.g.
another database. or an end-user application), in either direction.


> The side effect of the obsession with XMLising what was once flat and
> easy to access data is that Zebra is heaps slower and searching and
> indexing and requires its own daemon to be running alongside Apache.
>
> So more complexity has gained us what? Slower lookups, slower indexing
> and another daemon to maintain.

no, misuse of a useful technology has done that.

craig

-- 
craig sanders <cas at taz.net.au>

BOFH excuse #299:

The data on your hard drive is out of balance.