[
Lists Home |
Date Index |
Thread Index
]
Bob DuCharme wrote:
> My guess is that it's a bug in XML Spy, which doesn't report nearly as many document errors as it should.
same experience here. xml spy is a really nice interface, but the
internals are less than nice, it is a pretty buggy product wrt standards
conformance.
> Whether it's known to Altova is a secret, because they won't release a list. (I've asked after they thanked me for submitting one.)
altova declares everything to be a bug. i am pretty sure that,
technically speaking, xml spy is alpha-level software (ie, not all
features are implemented), but they refuse to give a list of
unimplemented features. xml schema constraint checking (eg, integer
ranges of simple types) was pretty much unsupported up to version 4.3.
now it's handled a bit better, but it is still buggy.
my personal favorite is a statement in their faq file about
non-determinstic dtds and xml schemas
(http://www.xmlspy.com/support_faq_ide_schema.html#q6_schema), which
basically says "we think this part of the standards is not good and will
therefore ignore it". deterministic content models are required for dtds
and xml schema, and when after pointing that out to the support people
with the references to the specs, i never heard back from them. i found
it very weird to see a company declare itself as a "leading provider of
xml software", and at the same time willingly (and openly) ignoring
required parts of some of the fundamental standards they are claiming to
support.
> The Question "XML Spy has no problem with my document/DTD/schema but another program does, why?" is getting Frequently Asked enough to add to Peter Flynn's XML FAQ. Try parsing your document with a more robust parser such as Xerces and you'll probably get a better idea of the problem.
my biggest concern with xml spy is not so much that it does not validate
documents well, but that it is also very sloppy in reporting errors in
the xml schema itself. this is pretty bad, because you may be releasing
xml schemas which are simply wrong. this is not only rather embarassing,
it will also lead to problems if you start to use them with more
strictly implemented software. i often use ibm's schema quality checker
(http://www.alphaworks.ibm.com/tech/xmlsqc), which is very good in
checking xml schemas for correctness (xmlsqc does *not* validate
documents, it only checks xml schemas for correctness).
cheers,
dret.
|
