Britannica 2.0? Surely you're joking, Mr. Jorge Cauz!

Britannica's president Jorge Cauz announced that Britannica is going to allow edits submitted online according to this news report by Stephen Hutcheon. He even dubbed this "Britannica 2.0" in the title.

Or is it?

I take it that this "2.0" is in line with the spirit of "Web 2.0," and I immediately went to Britannica Online to check it out, only to find disappointments.

Web 2.0 is not only about user-generated content, it also inherits the open access from the web in general, or "web 1.0" if you want. When I see a link to a wikipedia page on a search result page of any search engine, I know I can read the contents without paying a dime, and there won't be pop-up ads that blocks my way to the article and the knowledge carried by it.

Britannica Online is everything but free and open access.

The first test - on not-so-new terms

As a not-statistically-significant-yet-interesting test of its freshness, I searched the WiFi privacy protocol "WEP" which is almost obsolete now. Guess what? No, Britannica Online doesn't have it.

By the way, Mr. Cauz, please make your web site compatible with Firefox. All I can see on Firefox is a misplaced ad that blocks the error messages. I have to switch to Safari to get the above page.

The second test - on not-recommended-by-experts terms

My second test is "Lorenzo's oil" which is a formula of oil mixtures discovered by Lorenzo Odone's father, who is not a doctor nor a researcher in any pharmaceutical companies. It's effectiveness on delaying the onset of ALD (adrenoleukodystrophy) is controversial and it is not recommended by the medical academy or approved by FDA as a treatment. It would be a signal that Britannica Online's editors and experts are too arrogant if they don't describe the oil and the controversy. To my pleasant surprise, they do have this entry.

However, I still can't read it. The noble Britannica said I was attempting to read a premium topic and I had to activate a free trial to read it.

Oh... here we go again, free trials. When I forget to cancel it before the trial period is over, you'll start charging my credit card. No, thank you very much. I'm not interested. I'll get information on wikipedia.

I know Britannica has a big cost in paying experts to write the articles. But why should I or any internet user care? Unless I can't find quality and free content elsewhere, I don't see a reason to pay.

The contributor's qualification

Okay, this is not exactly a "qualification", but how about this:
"Would-be editors on the Britannica site will have to register using their real names and addresses before they are allowed to modify or write their own articles."
Give me a break. If I write anything for Britannica, what matters is not who I am, but the correctness and quality of what I write. Maybe this is a way to get the copyright issues straight, or an anti-DoS measure to prevent over-run of Britannica's reviewers. But it also creates an entry barrier. I'm not going into the decades-old debate of anonymity vs real name on the internet, but the size and quality of wikipedia proves there is a market for anonymity.

Google ranks wikipedia the highest - duh

Another funny thing is about Google's ranking of wikipedia entries vs. Britannica articles, and I quote:
"If I were to be the CEO of Google or the founders of Google I would be very [displeased] that the best search engine in the world continues to provide as a first link, Wikipedia," he said."Is this the best they can do? Is this the best that [their] algorithm can do?"
Get over it, Mr. Cauz. If you don't know how PageRank works, hire a decent SEO consultant. You have cut your premier contents away from the web collective by requiring a login. Any reasonable search engine robot can't crawl pass that. So your premier contents are not in any search engine's index. Besides, not many people are willing to link to a page requiring a login from their blog entries or news articles. So your premier contents don't get much incoming links and score poorly in PageRank or any algorithm that takes into account the link structure of the web.

Anything worth visiting?

Britannica Online doesn't have enough coverage to keep up with internet age, is picky on the browser, charges reader for "premium" topics, requires real name and address for contributors. What good is a web site with all these characteristics?

Wake up, Mr. Cauz. This is not how internet and search engines work.


How to verify the integrity of downloaded files on a Mac

I classify files available for download on the internet as the following:

  • Respectable providers ship download files with signatures;
  • Less respectable providers ship with checksums;
  • Unrespectable providers ship only the download file.
The difference is mainly the following: a checksum allows you to verify that the download file hasn't been altered unexpectedly by someone you may not know. A signature allows you to verify the above plus that the file was signed by a certain someone, usually the creator of the file you downloaded. It may be advisable that you use the signature whenever there is one associated with a download file

The most popular way of digitally signing a file is using a private key in the realm of the web of trust. To get started, you can use the GNU Privacy Guard free software. If you don't have it on your Mac, you can install the Mac version. I've downloaded the dmg of version 1.4.8 and it worked like a charm.

After installing the GNU Privacy Guard, a command "gpg" is available to you in /usr/local/bin as well as some other utility programs. You can open a terminal and use
 $ gpg --verify foo.tar.gz.sig
to verify that the file foo.tar.gz matches with the signature file foo.tar.gz.sig. You can verify multiple files at once by
 $ gpg --verify-files *.sig
to verify the integrity of files. Read the manual ("man gpg") for more details.

Oops, an error!

If you see a message like
 gpg: Signature made Wed Dec 13 05:02:10 2006 CST using DSA key ID 64EA74AB
gpg: Can't check signature: public key not found
from "gpg --verify", it means that the public key of the signer is not in your GPG keyring. Just load the public key of the signer to your keyring and off you go. The following is how I do it, your mileage may vary.

The longer route:
  1. Go to http://pgp.mit.edu/, a popular public key server.
  2. Type the key ID in the search box. Remember to add "0x" in front of the 8-character hex key ID. For the example above, I typed "0x64EA74AB".
  3. If all goes well, you'll see a search result page listing one or more keys. Click on the key ID that matches to your input. Other keys belong to people that "vouch" for the key you searched.
  4. The key server should then show you the public key of the key ID. Copy and paste the plain texts between "-----BEGIN PGP PUBLIC KEY BLOCK-----" and "-----END PGP PUBLIC KEY BLOCK-----" to a file on your computer, with these 2 lines included. Let's call the file "foo.txt".
  5. Now import this key into your GPG keyring:
     $ gpg --import foo.txt
The shorter route: The 5 steps outlined above can be done in one command:
 $ gpg --recv-keys 0x64EA74AB
Now you have the public key of the signer in your keyring, you can verify the files signed by him/her using "gpg --verify".

在 Mac OS X 10.5 上安裝 Python 3.0

Python 3.0 在 2008/12/3 發佈,雖然 MacPorts 已經收錄,但 MacPython/Leopard 還在 2.5,要等到 Apple 收進 Mac OS X 裡恐怕還要等好一陣子,所以想嘗鮮又不熟 MacPorts 的人(是說我自己),就要自己裝了。

在 Mac OS X 上安裝 Python 3.0 沒什麼難的,只要依 ijliao 所說:「真男人用 tarball 手動安裝程式」就一切搞定。不過 Python 3.0 在語法、物件、函式庫上都有新東西,想要熟悉它,用互動環境邊打邊學是一定要的,那就絕對要裝讓你可以用上下左右鍵叫回程式敍述東改西改的 readline 函式庫,不然你會 copy & paste 到瘋掉。

安裝 readline 函式庫

Readline 函式庫可以在 ftp://ftp.cwru.edu/pub/bash/ 找到,我用的是 5.2 版,可以下載 readline-5.2.tar.gzreadline-5.2-patches 目錄中所有的補丁。在著手安裝之前,你也許應該確認一下這些檔案的完整性

$ cd ~/3party  # 或任何你喜歡的目錄
$ tar zxvf ~/Desktop/readline-5.2.tar.gz
$ cd readline-5.2
$ cat ~/Desktop/readline52-0?? | patch -p 0
$ ./configure
$ make
$ sudo make install
這就完成了,會把兩個函式庫 libreadline 和 libhistory 裝到 /usr/local/lib 目錄去,每個函式庫都有四個不同副檔名的檔: .a、.dylib、.5.dylib、.5.2.dylib,其中 .dylib 和 .5.dylib 是指到 .5.2.dylib 的 symlink。

安裝 Python 3.0

Readline 裝好後,下載 Python 3.0 的原始碼 tarball 安裝就 OK 啦!
$ tar zxvf ~/Desktop/Python-3.0.tgz
$ cd Python-3.0
$ ./configure --enable-framework
$ make
$ sudo make frameworkinstall
$ cd /Library/Frameworks/Python.framework/Versions/
$ sudo rm Current
$ sudo ln -s 2.4 Current
最後這幾步是因為用 frameworkinstall 才會把 Python 3.0 裝進系統的 Python.framework 目錄,但會把預設 python 改成 3.0,為了不干擾系統其他 python scripts 的運作,要再把預設改回成 2.4。

請享用你的 Python 3.0 吧!