Google Chrome logo development
Links
Meta
Categories
Archives
This “Cuil” thing
This isn’t going to be about how Cuil is fantastic or terrible or anything.
At least directly.
Actually it’s about their most fantastic feature, the image association with search results. From the FAQ:
We know from our research that people can make better and quicker decisions about relevance and quality when they can see an image from the website. We do our best to take images from Web pages that accurately reflect the content of the website. Many websites are full of images, so we use advanced algorithms to determine the best image to show the user.
Let’s look at some results. Continue reading ›
The greatest reason to do anything, ever
I was reading this fantastic e-mail dialog between biologist Richard Lenski and Wikipedia-killer Andy Schlafly, when I decided to poke around the intertubes for stuff about Conservapedia. I eventually landed on this article, originally published in the Los Angeles Times a year ago:
Andy Schlafly was appalled. He was teaching a history class to home-schooled teens and one student had just turned in an assignment that dated events as “BCE,” before the common era — rather than “B.C.,” before Christ.
“Where did that come from?” he demanded.
Her answer: “Wikipedia.”
At that, Schlafly knew he had to act. . . . Thus was born Conservapedia.com — touted as “a conservative encyclopedia you can trust.”
This reminds me of the time I started my own country because of a dispute over the price of tea.
Observation while briefly watching a censored version of The Longest Yard (2005 remake)
Words that you cannot say on TNT:
- “shit” (including but not limited to the derivation “bullshit”)
- “goddamn”
Words that you can say on TNT:
- “nigger”
Now, I’m not going to say that this is related to the fact that TNT is headquartered in Georgia, but [redacted]
The voodoo of gcc, Part II
This is a direct follow-up to my previous post on gcc optimizations. The main code example, integer division by 7, came from a session of the Compilers class I took this semester. The professor was demonstrating the use of gcc to show example assembly for certain operations: arithmetic, branching, etc. We were all expecting something like the very first assembly example, using the integer division instruction the 80486 provides, and we all had the same flabbergasted reaction to the code gcc actually produced. That course gave me a healthy respect for the fact that compiler writers are, by and large, the smartest people on the planet.
I had a similar moment with gcc producing something unexpected while I was writing the previous post. The benchmark code at the very bottom of it, specifically.
The voodoo of gcc, Part I
Here’s a stupid bit of x86 assembly code, using AT&T style:
divideBySeven:
pushl %ebp
movl %esp, %ebp
movl 8(%ebp), %edx
movl %edx, %eax
sarl $31, %edx
movl $7, %ebx
idivl %ebx
popl %ebp
ret
What does this do? Well, the label should give you a pretty good idea. Converting it into C, we get:
int divideBySeven(int num) {
return num / 7;
}
Natch.
nostrangerstolove.java
Submission to the practice problem of the 17th Annual University of Evansville High School Programming Contest:
Atom feed for Folklore.org’s Newton stories
I mentioned last post that ten stories were added to Folklore.org’s hidden Newton section. Well, since then, another one has been added. I’m kind of annoyed by the fact that the feed for that section still doesn’t work, so I need to remember to go the index myself to check for new stories.
So, I made a feed.
http://quiteajolt.com/projects/newton/newton.xml
It’s an Atom feed, updated at the top of the hour (assuming there’s an update), containing the ten most recently added stories. I haven’t really put the generator through its paces, so it’s probably pretty flaky, but I’m subscribed to it myself, so hopefully I’ll notice if something goes awry. Enjoy it until Folklore.org’s actual feed is operational.
If you’re interested in the implementation details, the scraper is written in Python and uses Beautiful Soup to parse the page and PyFeed to generate the XML. You can view the code itself at http://quiteajolt.com/projects/newton/makeRSS.txt.
Borogoves
Went through the old posts and spruced everything up so that they at least look no worse than they did in the old theme, I think. Had an oddity with my image styling with Safari 3 and rows of thumbnails, like in this post, where the images load fine the first time, but after the page is refreshed, at least one of the images (almost always the first or last, and usually both) ends up squished to about ten pixels or so in width. Resizing the window causes the shrunken pictures to increase in width to about fifteen pixels, and refreshing after that may or may not cause one or more of the images to return to their proper size. Works great in the latest Minefield nightly, though, but it wouldn’t surprise me if that’s in the minority. Gotta love floating elements.
To give you the illusion that I’m not shafting you on content for the second consecutive post, while inspecting my Folklore.org post, I decided to check the site to see if there had been any updates, even though the RSS feed hadn’t been updated. And, surprise surprise, there are currently ten stories about the Apple Newton available for your consumption. No real humor or person stories yet; they’re actually more technically oriented so far, talking about such page-turners NewtonScript and the memory model.
The stories have been on the site since at least March 25, since there’s one comment stamped with that date on the story about recognizing Lewis Carroll’s “Jabberwocky.” The new stories did not show up in the RSS feed, and in fact the “This is a test” entry has been removed from the feed, and doubly in fact the “This is a test” entry has been removed altogether. However, the comments entered on that entry are still in the comments RSS feed (as is the “Jabberwocky” comment I linked to earlier), and in one, Andy Hertzfeld said that he’s currently working on a separate RSS feed for the Newton stories. Here’s hoping it’s finished soon and that more stories follow.
Shades
I’ve decided to go with Barthelme as the new theme. Haven’t really made any structural changes to it, but then again, my previous theme was just a thin coating over Kubrick, so I’m not exactly forging new ground here on the frontier of using canned themes. Barthelme was one of the only themes I found that fit my three most important criteria: (1) don’t suck, (2) don’t make me retch, and (3) work.