Why are people using XML as a slow, bloated ENTERPRISE database? XML is a markup language for documents or for objects designed to be embedded in documents. DocBook, XHTML, ODF, MathML and SVG are good uses of XML. Except for SVG, these are all marked up text. The text between the tags is usually more important than the tags. Consider a text board thread without tags. It's still readable. An HTML document without tags is plain text without any formatting, scripts, images or other elements. Hierarchical databases are not marked up text documents.
PROTIP: If you strip all the tags and don't get a plain text version of the document, don't use XML.
Name:
Anonymous2012-08-26 20:14
Sexp.
Name:
12012-08-26 20:26
>>2
Exactly. They're using XML as Sexps. This is where people get the idea that ``XML is a poor copy of Sexps'' instead of ``XML is an extensible document metaformat''.
>>5
lisp-like expressions are interesting, but it can be used for code + data with the same syntax. Isn't that a little overpowered to just store data?
XML is perfectly suitable for databases if it adheres to some schema standard like DTD or XMLS. There are query languages that support it. Have you even read a real book about databases? Or taken a class or certification? Or are you just another 100-line-script code monkey talking out of your ass?
>>7 XML is perfectly suitable
This person has not read his cat-v.org today.
Name:
Anonymous2012-08-27 4:05
>>8 query support for sexp
Uh, just plain Lisp code?
Name:
Anonymous2012-08-27 5:26
The silliest thing about XML is that it didn't bother to carry over some little things from SGML, such as </>. Yes, the old ass markup meant for storing human publications has a </> shorthand as well as <tag/stuff inside the tag/ which is as compact as things will get anyway, and the recent machine oriented format that people send terabytes of shit with doesn't.
XML has taken off largely because of its versatility, resulting in many tools written against it. That, and the fact that the W3C promoted it along with HTML (similarly widespread). For hierarchical data storage, things like ASN.1 PER are vastly more efficient at the cost of not being easily human-readable.
>>18
Python, vim and (maybe) nano are acceptable. I'd dump desktop enviroments and some of the bloated window managers for minimal tiling WMs. Though, I might try Window Maker since the guy seems to be developing it again, and you can make the NeXT interface look pretty without too much effort. The rest lf the suggestions are spot-on, GNU software is generally bloated. GRUB? Ditch that and go back to LILO.
>>19 +---------------------------------+-------------------------------------+
|Something Rob Pike didn't invent. |Something Rob Pike invented. |
+---------------------------------+-------------------------------------+
Name:
Anonymous2012-08-27 12:17
>>16
XML is misleadingly simple. Just because the tags are nested properly doesn't mean that it's valid or does what you would expect. XHTML, for example, has certain rules about child elements. XHTML is almost always parsed as HTML so the developers don't notice the problems.
What do you think this does? <p>a<p>b</p>c</p>d
It's actually this: <p>a</p><p>b</p>c<p></p>d
Not what you expected, is it?
Another one: a<table>b<tr>c<td>d</td>e</tr>f</table>g
It's actually this! abcef<table><tbody><tr><td>d</td></tr></tbody></table>g
If you add in more tables and forms you can get something where the innerHTML has different semantics than the original code.
<table><form action="data:text/html,x"></table>
<form action="data:text/html,y">
<button name=act value=Submit>
<p><big>a
<p><big>b
<p><big>c
<p><big>d
<p><big>e
What does does act look like? What address does act submit to?