Altova Mailing List Archives
>xml-dev Archive Home
>Thread Prev - Vocabulary Combination
Re: [xml-dev] Vocabulary Combination
Date: 5/29/2003 7:01:00 AM
Tim Bray <tbray@t...> wrote: | Arjun Ray wrote: |> it is written: "The only reason namespaces exist, once again, is to give |> elements and attributes programmer-friendly names that will be unique |> across the whole Internet." | | Don't be silly. Why would you want names that are unique on a wide | scale if you weren't going to be combining vocabularies? *I* don't want them. In fact, I don't even need them. (All I'd need is unique names for vocabularies, the good ol' PUBLIC id concept.) I'm just curious about a pretty common delusion about the problem. | Interesting thought experiment. Why did you leave the <h:head><h:title> | construct out of the HTML view? Editing accident, sorry. | Upon reflection, I'm not convinced that this "views" approach is useful. I find it very useful. | Merely subtracting any of the markup vocabularies is almost never apt | the right thing to do. That's ridiculous. That would make you hostage to anyone who stuck in a namespace/vocabulary that you had no clue about. To announce, in instance markup, that a vocabulary is in play is to allow for the possibility of *partial* understanding - which happens to include the possibility of "all I need to know anyway". | If what you're trying to do is display this, I was careful to leave specific purposes out. Why have generalized markup if all w're supposed to be interested in is specific purposes and no others? |> If this is acceptable, then my question is: What is the decision procedure |> by which a generic parser-level filter could generate these views, if it's |> to take namespaced names as a guide? | | Totally application-dependent, I'd think. It doesn't seem likely that | "generic parser-level filter" is a very useful construct. You just found one yourself, with the RDBMS loader application. | > "It can't be done" is an acceptable answer, btw. | | In the general case, it can't be done. The namespaces don't give you | enough information. Right. Namespaces are neither necessary nor sufficient for the general problem of vocabulary combination. As we knew a long time ago.