Saturday, March 13, 2004

Free Books Online

My we're busy today, aren't we. Two sites of interest:

For books out of copyright, see Project Gutenberg, which now has thousands of books online. These were typed or scanned in by volunteers. Bibles, Shakespear, Twain, it's all there. Now the files are available on CD and DVD. Useful for taking on trips with the laptop. The books are in plain ASCII format. If that bugs you, read them with your browser and change the font to something you like.

For certain books in print, there is the Baen Free Library. OK, it only has Baen books, and not the current ones, though you can also purchase current ebooks from Baen. The free books are promos, of course, hoping to draw you into one series or another. These books are HTML formatted. Series included are Telzey, Honor Harrington, Retief. Authors include Bujold (one short story), Flint, Weber, Niven and Pournelle (Fallen Angels), and more.

And the Problem With RSS Reader Panel is

You'll remember I'm using the RSS Reader Panel extension to Firefox as my RSS Aggregator. And it works really well.

Except on this Blog. If you add Linux & Things to your RSS bookmarks, and then click on the link, everything looks normal. On the sidebar below the bookmarks you see the article titles. In the "content area" you see the first few lines of each article. However, when you click on an article you get the message

We're sorry but the Username/Password combination you've entered is either invalid or you don't have permission to access this Blog.

I contacted Blogger support, who gave me a nice answer which showed me that they'd really looked at the problem but, unfortunately, their answer comes down to "not our problem, man". :-( It seems that when Blogger produces its Atom-formatted XML file it produces two links per article. The first is an internal link for Blog*splot, and the second is the permanent link to the article. As far as I can tell, Blog*splot is the only site that does this. I've examined some other Atom files, and none of them seem to have more than one link per article. However, Blog*splot does produce proper XML code, according to http://feedvalidator.org, so I guess it's legal. The permalink is identified as type text/html, while the other link is type application/x.atom+xml, so it shouldn't be too hard to figure out the correct link.

I've posted a bug report with Reader Panel Support, so hopefully this will get fixed soon.

Favorite Blogs

OK, following Slate's lead, here are my favorite RSS Newsfeeds. This is in OPML format. (No, I didn't know what that meant, either. Thanks Google (TM).

Touring Blog Sites

The Internet Tourbus is one of my favorite email newsletters. The current newletter is about gadget blogs. This doesn't interest me all that much, but Patrick provides a link to http://www.bloglines.com/topblogs, a rather large list of Blogs with XML and RSS links. (Wil Wheaton, fix your XML feed.) Worth much more than the price of the newsletter.

Sunday, March 07, 2004

You can now

search this site. Well, you could, if it was popular enough to make Google (TM).

Saturday, March 06, 2004

And Then There Was One

With Stanford's loss to Washington this afternoon, there is only one undefeated team left in the men's NCAA Division I: the St. Joseph's Hawks, who are currently 27-0, with the Atlantic 10 Conference Tournament (held in the coastal city of Dayton, Ohio) to go. No D-I team has finished a perfect regular season plus conference tournament since UNLV in 1990-91, and no team has had a perfect Division I season since Indiana did it in 1975-76. Only UNLV and Indian State (1979) have gotten to the NCAA championship game with a perfect record, and they both lost.

Since none other than Roy Williams has said that the NCAA tournament is a crapshoot, I got to wondering what the odds were that we'd see a perfect season anytime soon. Since a team with a perfect record is likely to be the number 1 seed, I decided to look at all of the number 1 seeds in the 64-team tournament era (1985 on) to calculate a lower bound on the probability. I got most of my data from The Official 2002 NCAA Men's Final Four Tournament Records Book, and I looked up individual team records for 2003.

Start with the regular season, which, to the NCAA, includes the conference tournaments, if any. Since the 1984-5 season, the 76 teams with Number 1 seeds have gone 2086-266. In other words, an average Number 1 seed has a 0.8869048:1 chance of winning any given regular season game. Since a typical regular season is 30 games (including the conference tournament), this means that the probability than an average Number 1 seed (hereafter AN1S) has a 0.8869048^3:1 = 0.0273092:1 chance of a perfect regular season. Thus, out of the 76 N1S since 1985, we'd expect 0.027*76 = 2 teams to have a perfect record through the regular season. We've had 1 (UNLV, 1991), so our estimate is reasonable.

Now for the tournament. To finish the perfect season, a team must win 6 tournament games. Of the 19 NCAA championships in the 64-team era, 11 have been won by a N1S. Thus the probability of a given N1S winning the tournament is 11/76 = 0.1447368:1.

So the probability of a N1S having a perfect regular season record is 0.0273092:1, and the probability of a N1S winning the tournament is 0.1447368:1, then the probability that a team will finish the season with a perfect record is 0.0273092*0.1447368 = 0.0039527:1. With 76 teams having had a shot at this, we'd expect 0.00395*76 = 0.300 teams with a perfect record. OK, we don't have any, so this isn't all that bad an estimate.

But 0.3 isn't all that small either. With these odds, we could have had a perfect season in the last 19 years, just taking an AN1S and a little bit of luck. So we could have a perfect season this year. St. Joe's? Well, they're the only one left. I wouldn't bet on them winning, but it's not all that impossible.

And Roy? A N1S has one chance out of seven of winning it all. You've had 5 chances. The clock is ticking.

And One More RSS Thingy

Hey, look at this: this very Blog publishes its own RSS Feed. Who knew? Apparently I signed up for this when I created the Blog, but I didn't know what I was signing up for. There's even a summary page, which consists of the first few lines of each article.

And More Stuff About RSS

Slate (from MSN, but I don't hold that against them) has an article on speed-reading the net, which is all about how RSS works. They even have a collection of popular RSS feeds, including the Washington Post and the New York Times, not to mention the Drudge report. This is an OPML file, which means you can't read in your browser. To use it with the RSS Reader Panel, open up the RSS sidebar, click on Tools, click on OPML Import/Export, and follow the directions. All this is to introduce the launching of Slate's own RSS feed.

While following the links in the OPML file, (specifically, Drudge -- you can look up the link yourself) I found a link to Myway.com, which is a portal that looks the way Yahoo used to look. Clean, no adds, no popups (not that users of Mozilla Firefox have to worry about that). They offer their own webbed based email (of course), but allow you to access any other webmail account with a click or two. Very nice.

Another thing I learned about RSS and the Reader Panel: Many (not all) RSS feeds contain story summaries. Just using the RSS sidebar doesn't show these summaries, but if you click on View => Open in Contents Area and then click on the RSS feed, the summaries will show up in the browser window, with the links still in the lower panel. If tabbed browsing is enabled, doing a middle-button mouse click will open up the summary in a new tab.

Sunday, February 29, 2004

Leap Day, When Mike Learns
         about RSS Readers

Well, he learns a little bit, anyway. As I understand it (and if you want to correct me, hit the email button on the right):

If you go to a page such as Syndic8.com, or random other pages, you'll see an orange button labeled "XML".

At the bottom of every Slashdot page, there is a link labeled "RSS".

Click one of these links, and most likely your browser will do something strange. Mozilla Firefox prints the message "This XML file does not appear to have any style information associated with it. The document tree is shown below." and then dumps the XML source.

So what are these pages, and how do you read them? RSS (sometimes) stands for "Really Simple Syndication". As far as I can tell, it's a collection of bookmarks for a website. For example, Slashdot's RSS feed is a list of the current articles appearing on the Slashdot home page. The advantage of the RSS feed, rather than your own bookmarks, is that the page is dynamic, that is, it gives you the current Slashdot home page, not the way it looked the last time you went there and bookmarked every article. The articles don't have to come from the same page, they can be distributed throughout cyberspace.

There's more information on this at XML.com and news.yahoo.com. The later link has a bunch of RSS feeds for Yahoo's news service.

OK, that's well and good, but how do you read these things? You need something called an RSS aggregator. Many are listed in that last link.

But, since RSS is essentially a collection of (dynamic) bookmarks to web pages, why not use a web browser to read them? The Mozilla browsers don't support RSS directly, but you access feeds using an extension. (There are extensions for Mozilla itself. Use Google to find them. In Firefox, see "Tools=>Options=> Extensions."

There are several RSS extensions for Firefox. I chose the RSS Reader Panel. You install this in the usual manner, which means you're trusting someone to send you good code. Fortunately, this is in your own file space, so you can't trample on things belonging to root or another user. (Under Linux. In Windows, this isn't true, of course.) The RSS Reader Panel (RSSRP) install itself, and you restart your browser.

Next, create a bookmark folder. It doesn't matter where. I called mine, duh, "RSS Feeds". Under "Tools=>Options=>Extensions" click on "RSS Reader Panel", then on the "Options" button at the bottom. There is a menu which has a scroll button with all of your bookmarks. Highlight the RSS Feed bookmark and click "OK".

Add any XML or RSS feeds you'd like into this folder in the usual way.

Now go to "View=>Sidebar" and click on "RSS Reader Panel", or just hit "Alt-R" on the keyboard. On the left-hand side of the screen, all of your RSS bookmarks will appear in the top half of the sidebar. Click on the one you want. In the bottom half of the sidebar will be the titles currently available from that page. Click on the one you want and it will appear in the main part of the browser window on the right. Get rid of the sidebar by clicking the "X" in the upper right-hand corner.

I don't know if I'll use this all that much or not. Time will tell, I guess.

Saturday, February 14, 2004

Interesting

OK, when I installed firefox (previous message) I created a new profile, ~/.phoenix/mike. However, firefox now defaults to ~/.phoenix/default. Which means that I can use the old extensions and bookmarks, and that's OK, I guess, but how do you permanently change what the system considers ``default''?

Friday, February 13, 2004

Maybe they could settle on a name?

The browser formerly known as MozillaFirebird (formerly known as Pheonix) is out with version 0.8. This time it's known as Mozilla Firefox.

Installation, at the moment, consists of undoing the tarball in the location you want the browser, in my case /usr/local/firefox, with a link $ ln -s /usr/local/firefox/firefox /usr/local/bin/firefox You can use your old profile, or create a new one. In view of what happened with firebird, I decided to create a new profile. This means, of course, that I have to log in again to all those websites that require registration. I did move my old bookmark file over, but I'll let the cookies and password management files build themselves up as needed.

I took my old Flash and java plugins, which work well. I have yet to install plugger. I may try mozplugger first. Maybe even mplayerplug-in.

The adblock extension got installed with no problem, but the mozex extension wasn't available. Hopefully that will be added soon.

Sunday, February 08, 2004

Memory Key

I added a Fuji USB drive (aka memory stick/key), on sale this week from Best Buy (256MB for $60, after rebate & before taxes). It doesn't show up using the hwbrowser, unlike it did for johnmunsch.com (search for memory key). Nevertheless, I was able to mount it using the fstab entry that I used for the digital camera. Which means that I read the memory by hitting

mount /mnt/camera

and dismount with

umount /mnt/camera

I could change the mount point, but this isn't all that hard to remember. Conflicts will arise if I try to have both plugged in at the same time.

In other news, KU defeated the Texas Tex Salad Bar Ragers, 96-77.

Saturday, February 07, 2004

Truetype fonts

Those that Microsoft used to provide, can be installed by following the directions at http://www.mjmwired.net/resources/mjm-fedora-fc1.html#ttf. Remember to restart the font server.

There's another Blog out there

fedora core linux installation adventure is a blog similar to this one, except it's got nice indexing, formatting, and intellegence. Also, it's not so hung up with the current UNC basketball administration.

How you know it's a fantasy

Watched "Naked Gun" again last night. Dick Vitale was one of the many sportscasters at the baseball game, along with Allen, Gowdy, Enberg, Costas, Michaels, Palmer, and I've probably missed some.

Vitale is the only one who says not a word.