?

Log in

No account? Create an account

Previous Entry | Next Entry

XMFUCKYOU

I don't like XML.

No, wait, let me restate that. I hate XML.

No, wait, let me restate that. I despise the over-engineered, over-verbose, over-committeed (that isn't a typo), over-hyped, and under-powered pile of steaming horse shit that is XML with every cell in my body.

And I dearly wish there was a YAML parser for C#.

Comments

( Read 9 comments — Leave a comment )
izuko
Oct. 7th, 2005 10:51 am (UTC)
I think I started a protest against it about seven or eight years ago. Called it KISS HTML (Keep it Simple Stupid). Never got anywhere.

Let me also say that I hate the W3C's vision for HTML 4.0 and beyond. Yes, I know style sheets are very powerful, and very flexible. But those depriciated tags were easy to use, and allowed those of us who are all program languaged out to write clean HTML instead of resorting to the bastard code of BluntPage.

By trying to force the ideological goal of separating style from content, they've made HTML less accessable and less useable. Can you imagine what would happen if Toyota made cars -more- difficult to drive?
countalpicola
Oct. 7th, 2005 01:18 pm (UTC)
> But those depriciated tags were easy to use, and allowed those of us who are all program languaged out to write clean HTML instead of resorting to the bastard code of BluntPage.

That, in part, is what the Transitional DTDs are all about. HTML 4.0 Trans and XHTML 1.0 Trans have a bunch of deprecated tags, but they've only removed the most useless of tags from the earlier specs. Now, sure, Strict is a different animal, but nothing says you have to write in Strict. Indeed, nothing says you have to write in HTML 4.0 or XHTML. HTML 1.0, 2.0, 3.2 all have valid DTDs and you can, if you really want to, code to them.

Personally, all I use anymore is XHTML 1.0 Strict, but that's unusual even for people who do web design professionally. There's nothing wrong with using older specs. Just be sure to set the DOCTYPE tag at the top accordingly.
chiave_trust
Oct. 7th, 2005 02:42 pm (UTC)
Then there's me, who wrote in HTML 4.0 Strict before I converted it to XHTML Strict (which basically consisted of me changing the DOCTYPE tag and one other tag in the entire site).

It's clean, but you don't have to use it. I just happened to be using CSS as well, so seperating style from content worked alright in that case. (And I'm a bit of a masochist, apparently, in writing Strict when I didn't have to at all, on the first webpage I've written in a couple years.)
izuko
Oct. 8th, 2005 11:12 pm (UTC)
I'll stick with Transitional until browses no longer support it (which, given the crappy code written by page writers, will be forever). Still, I do have to be just a little incensed that they're moving a useful tool further away from usability.
brentdax
Oct. 7th, 2005 05:10 pm (UTC)
I actually like the more stylesheet-based HTML and the syntactic strictness of XHTML. Both of them keep you from using things that browsers have traditionally all treated differently.

XML is actually a good markup format and a halfway-decent data structure format (although YAML is better for data--just as expressive but more compact and readable). The real problem with XML is that Some People think it should be the One True Format, and in particular try to describe XML in XML with languages like XSD and XSLT; this quickly becomes completely unreadable because XML isn't powerful enough to describe itself in a compact way.
izuko
Oct. 8th, 2005 11:13 pm (UTC)
I actually like the more stylesheet-based HTML and the syntactic strictness of XHTML. Both of them keep you from using things that browsers have traditionally all treated differently.

Then fix the browsers, don't break the tools.
brentdax
Oct. 8th, 2005 11:29 pm (UTC)
Their approach to fixing things without breaking backwards compatibility is to deprecate the old things and create new, more regular things with strictly-defined behavior. Not that that stops Microsoft from ignoring or creatively interpreting the strict definitions…
izuko
Oct. 8th, 2005 11:35 pm (UTC)
As long as Microsoft's approach allows me to view olde web pages without my browser crapping itself, and lets me still write decent web pages without reaching though my ass, I've got to conclude that their way is better.
codepoetica
Oct. 7th, 2005 04:21 pm (UTC)
XHTML1.0 Trans for me.

For pushing data around, however, which I think is the point of Brent's post, I abhor importing XML. Parsing the soup is just an absolute pain in the ass, driving me to use

  • url=encoded&data=strings
  • variable=assignment\n
  • http: headers\n
  • ( {"json": "strings"} )


Although json is a pain in the ass to debug sometimes. I have absolutely no problems with exporting my data in an XML document of some form, but in general avoid accepting data in that form. Especially if it's being passed between two of my own programs. If I need to accept XML data in a short one-off script, I'll regex it out of the XML document without care of the structure, simply because it takes much less time.
( Read 9 comments — Leave a comment )