[exprla-devel] Re: [xpl] DTD or namespace?
Status: Pre-Alpha
Brought to you by:
xpl2
From: reid_spencer <ras...@re...> - 2002-01-31 07:54:27
|
--- In xpl-dev@y..., Jonathan Burns <saski@w...> wrote: Aly Hopkins wrote: > In reply to Jonathan Burns' letter of 29 Apr. (Re: It's Us....), I > was thinking that we had the option of switching languages on the fly > by reference to a different namespace, therefore: > > xpl:constant-declaration > xpl:input > xpl:dataType > xpl:if > xpl:test > xpl:operator > > and so forth. > Bingo. I discovered XML namespaces just last night, nosing roundthe World Wide Web Consortium XML page. Thank heavens, we're beginning to anticipate the ned for collision detection and build the remedies in. Haven't got namespaces integrated into my thinking yet. > I suspect that he & I are really talking about the same thing, just > applied a little differently. Me too. Need to think it through. > I would be happier with leaving the DTD fixed in stone but > allowing XPL to call in any language via namespace > prefix. This fits with my intuition of DTDs being like classes.At least in C++, we don't generate classes on the fly. (In Smalltalk we can, but that's another story.) I seem to remember that Java has a namespace-like mechanism forprograms to work with imports. But here's the test: Here we have the XPL definition. It has integers, xpl:int. We call in an accessory XPL language, say "executable MathML", or eMML. It has integers as well. How should we tell the XPL parser, that wherever the xpl:int type appears in the DTD, eMML:int is also acceptable? > As indeed XML is pretty good at calling in ANY resource, in > any number of ways, from anywhere. Really? Oh, this is looking better and better. By the way... Compilation is a process which converts a program in a high-level language directly to instructions for the physical machine architecture on which the program is to run. As contrasted with... Interpretation is the process by which instructions for a virtual machine architecture are processed by the virtual machine. This process is indepedent of any physical architecture on which the interpreter program may be running. An architecture consists of a processing model and a set of instructions which operate on the processing model. (Not happy with this, it's too close to the kind of description which one has to understand beforehand, to make any sense of it.) Goodwill Jonathan --- End forwarded message --- |