All Comments for iRi

comment by Jeremy Bowers

Yeah, my care level on that is pretty low. I think I'll just shut them down for now. Consider the reddit discussion official.

comment by Jeremy Bowers

Point.

Consider it limited to cases where it's being used to replace HTML only; I probably shouldn't remove that now.

comment by William McVey

Weird. I got a Javascript error when I first tried posting this and got no indication that it was posted, so I submitted the version that is shown below. Something certainly looks broken on the comment acceptance side of your blog.

comment by William McVey

This is also just about HTML replacements; using a single representation that goes out to many things, of which HTML is just one, is a whole different story altogether.

and yet you list reStructuredText iny your critique of markup languages. reStructuredText does support an HTML output format, but other output formats, such as DocBook, LaTeX, and just regular ole Text Files. The intent of at least for reST was not to build a markup that was easier to write HTML with, but to write an extensible markup language that could convey semantics to documents without constructing a unique document types. In other words, they weren't trying to build an easier HTML, they were trying to build a more intelligent TXT.

comment by Ashground

Two comments -

While HTML is fine, nobody should be using <i> or <b>. If you want to advocate the use of HTML, you might want to make sure the HTML you're advocating is valid.

Also, I'd kill myself if every time I wanted to write <em> on my blog I had to write <em>... or the fact that in order to write that last bit, I had to write <code>&lt;em&gt;</code>.

comment by Jeremy Bowers

I believe you are looking for this.

And holy cow! My reply form is borked!

(Been making some changes recently.)

comment by Ilari Kajaste

Thanks for the insightful and well-written article about the swamp that is encoding. Nothing much new there, but it was a nice clear thoughtful-look-at-the-issue -type of read. I originally came up here via slashdot comment link, trying to find stuff about your claim that it's a mathematical fact that you can't create unbiased work. But I failed. (The adress (you?) linked in 2006 was /iri/2004/10/05.html, which is now defunct.) Before performing the close-tab -gesture, a title and few words caught my interest, and so I was derailed from my quest for documentary ethics into enconding. Delightful. Would add your site to a "followed blogs" syndication/whatever -system, if I would be using one...

Failing to find a link titled "archives" or whatnot in the righthand menu, I never did reach the intended article (oh, how lazy we are, us internet readers). But if the original link would have actually worked, I wonder if I'd ever read this one... Makes one think for a second, it does.

comment by Matt Stegman

Hi, Jeremy. This was pretty thought-provoking for me. I'd learned long ago the basics of escaping, and using it to avoid SQL injections, XSS and other attacks, but to think about it as an encoding problem is a new idea. It takes a little bit of thinking to get all the layers in your HTML/Javascript example; I still don't think I "see" all of them. But now I think I'm beginning to understand why I've seen layer upon layer of character conversion methods in some web applications that don't seem (to me) to do much that's useful. Thanks for posting this. You've already helped me start to see more of what's going on between the user and the disk, so to speak.

- Matt

comment by llimllib

Mencius Moldbug has been saying this for a long time: http://unqualified-reservations.blogspot.com/2007/09/is-journalism-official.html
. I performed an experiment on reddit to see what people thought about this idea: http://reddit.com/info/65w67/comments/ .

comment by Anon Ymous

Sorry, but I can't even consider voting for a canidate who won't at least pay lip service to not torturing people...

Best would be to not do it at all.

Sigh, what has this country come to?