Re: Starting point

--- Mike Sokolov <sokolov@falutin.net> schrieb am Mi, 25.7.2012:

> > <patrick.szabo@lexisnexis.at>  wrote: 
> >> In my experience PIs are a good idea [...]
> >> How about allowing PIs but only on root-level ?   
> > That doesn't meet the various needs of PI's that I have
> > seen?

> In my limited experience, PI usage could have be handled
> equally well in some other way, like:

> <processing-instruction name="foo" content="bar" />,
> or with some special purpose tag like:
> <page-break num="11" />

Of curse they can, but these wouldn't be PIs, but rather daata as part of the XML application. They are not seperated by anything from <BOOLD> to <MARQUE> ... PIs are infromation about Processing the document at whole.

> Possibly excepting the root-element PI, which has been
> granted special meaning by (some?) browsers.

PIs are not (so much) about application specifix programmslike browsers. But rather a tool to encapsule Handling ingeneral.

> Is it just that existing content with PI's would have to be
> converted to some other style in order to be rendered as
> uxml?  Or is their some intrinsic virtue of PI's that I
> don't see?

Think of them as of the Unix shebang. Thats a processing instruction within a programm presented as text file. As for the language interpreter (equal to a brower here) it is just something he ignores (hopefuly) while an external application (e.g. a shell) will use this to identify the interpreter/compiler it has to start to get this file processed. In Unix it was more of a hack to use a comment syntax (#) wile XML gives a seperate encoding from simple comments, thus allowing to seperate between hun glibberisch and rightful data. XML does not define what to do (as most applications will not), it just say that anything in form of <?...> is a PI and may have some external meaning, so better preserve it if you foreward it, even if you don't understand it.

H.

--

www.vcfe.org

Received on Wednesday, 25 July 2012 13:15:41 UTC