40RE: [RSS2-Support] Summary of issue with xmlns attribute
- Sep 29, 2002
>> It was a statement of fact rather than an argument.What I said was "[there are] quite a lot of fairly respectable tools around
that can't use namespaces". Ok, so 'fairly respectable' might be open to
interpretation, but how about 'designed to conform to well-established
specifications'. There are plenty of apps around that are perfectly good in
themselves, but don't support namespaces. Fact.
>How that applies here is it's not that use of namespaces is bad.As long as it isn't validating, yes. If it is, then objecting to the
>The truth, in
>fact, lies in parser(s) mishandling the XML data. Namespaces are,
>on one level,
>just additional attributes. A parser following the XML spec
>should be able to
>ignore them with no trouble at all.
unspecified (ns) attributes is the correct behaviour.
>spec at fault. This is not namespaces at fault. This is theWith the proviso above, I agree.
>PARSER at fault.
>Hunt down that parser and help it start supporting the specs.My advice would be to just throw that parser away - decent parsers are
[explanation of namespaces snipped]
>What's happening here is developers are learning about what reallyAre they learning? Recent history would suggest otherwise.
>spec is about. They're also learning it critical for a spec to be
>what's going on. Not to mislead out of ignorance.
>One could argue the latest spec is somehow helping things. ThisWell, quite ;-)
>is like saying
>Typhoid Mary was a good ambassador for the concept of vaccines.
>The solution is for people to educate themselves and stop acting out ofAgreed.
>ignorance. If something out there can't parse the XML properly
>then how about
>working on getting that parser fixed instead? As to the spec?
>The authors of
>any spec need to take responsibilty and do a good job INSTEAD of
>which they don't understand.
- << Previous post in topic Next post in topic >>