Archive for the 'feeds and syndication' Category

Meet the new boss

Friday, April 30th, 2004

A new syndication spec, combining what looks at first glance like some good reasonable ideas with the grace and tact of an elephant. With diarrhea. Fun ensues.

Or, maybe more strict

Saturday, April 17th, 2004

First I thought Mark Pilgrim’s ideas on what was unsafe in RSS were too extreme. Then I looked at what you can actually do. Now I don’t think they are nearly extreme enough. Everything’s unsafe, and I’m going to have to write my own aggregator, probably called Tinfoil Hat, that embeds Notepad for rendering nothing but text/plain.

That’s the way (uh huh)

Friday, April 16th, 2004

Bloglines adds a per-user preference that can be over-ridden per-feed, to let you choose whether to display full items, summaries, or just titles. Finally, after years of putting both in my feeds, they’ll really get used the way I wanted.

Unacceptable Microsoft bugs

Tuesday, April 6th, 2004

Forget RSS and Atom, I just want a working cross-browser CDF.

Slightly more secure RSS-to-local-HTML

Sunday, April 4th, 2004

Add the right comment in your generated HTML, and in WinXP SP2 a page that IE loads from the local machine is safely in the Internet zone, not the unsafe Local Machine zone. A nice added measure of safety for some aggregator authors.

What warnings do you heed?

Thursday, March 18th, 2004

How can a program say “that’s valid, but you need to think about whether this is right” in a way that will make you pay attention?

How not to do an atom:summary

Monday, March 8th, 2004

One of the things that Atom gets right (so far, at least) is separating the idea of a summary from the content of the post. Stripping the (X)HTML from the first few characters of a post, and then claiming that it’s XHTML with significant whitespace isn’t how to do it, though.

Be happy!

Friday, March 5th, 2004

This hand-crafted excerpt of a longer item brought to you by the magic of separate elements for separate purposes.

Oops, that’s a bad IDea

Sunday, February 15th, 2004

The Atom id element’s a good thing, but not if your Movable Type template produces a non-id id.

Come home to Pie

Friday, August 22nd, 2003

Just name the bloody thing Pie and get on to other arguments, please.