What if, instead of just plain XML, or any other markup language for that matter, we build an executable file that upon execution prints whatever is stored in the file? That way we can run like:
$ ./index.xml
And then see the output of the file on stdout. Since the file is binary, we can't just edit it with emacs, so we have to make the file edit itself! For example:
On second thought, maybe we should make it launch a graphical application that allows you to edit the XML tree with your mouse. That would make it easier.
Name:
Anonymous2008-09-06 16:08
> index.xml /EDITNODE 3 /SETATTR_NAME haskell
Name:
Anonymous2008-09-06 16:29
The XML program should take XSLT programs its argument, and apply that program to itself.
Name:
Anonymous2008-09-06 17:22
One word, binfmt_misc, thread over.
Name:
Anonymous2008-09-06 17:29
>>1
How about we use a LISP program to store the actual data in such a way that running it will produce the XML document? Like this
(doctype whatever)
(node
(subnode (list ('attribute value) ('another other)))
;node)
results: <?DOCTYPE ...>
<node>
<subnode attribute=value another=other/>
</node>
Suddenly XSLT looks like a dirty hack.
Name:
Anonymous2008-09-06 18:46
>>1
As insane as this idea is, it's still far superior to DOM
Name:
Anonymous2008-09-06 18:56
>>10 >>11 >>8
Good thinking guys. Would you like to join my open source project that is working towards this goal?
Name:
Anonymous2008-09-06 19:20
>>14
Because people who think XML is a good idea and people who like Java overlap greatly?
>>1
That is the dumbest idea I ever heard. What's wrong with cat?
Name:
Anonymous2008-09-06 19:50
>>16
Yes because a language based on natural thinking patterns, which can generate its own code on the fly, is childish compared to a language which attempts to make it nice and easy for new programmers.
>>20
I certainly hope so. They're infinitely preferable to the /lounge/ faggot.
Name:
Anonymous2008-09-07 6:57
>>12
My idea is that it will be far easier to manipulate a lisp list with a LISP program and then turn that into XML than using XSLT for the same purpose. The added benefit is that it will be far more readable by humans and will have more capabilities. I am of course blissfully ignorant of the rationale behind XSLT. That "thing" mutated into a list processing language a long time ago. >>13
Thanks but I have to go back to reading my SICP.
>>15
Because people who think XML is a good idea and people who like Java overlap greatly?
Both are enterprise shit, both hide the sense of failure with the sense of "best practices", and people who like failing like failing hard.
>>18
Lol, don't worry, no functional programmer or dynafag will attempt to take your job. In fact, we want you to cover that job, so that we don't get asked to do that.