Canada beats USA in women’s hockey!

Been playing around with Greymatter, a CGI-based weblogging alternative to services such as Blogger which I currently use to maintain this blog. (For the non-technically-minded, CGI here means “Common Gateway Interface” which is a standard for programs (“CGI scripts”) that process input from web forms.)

I’m experimenting with Greymatter mainly for the added flexibility it provides. Flexibility is a double-edged sword though. For example whereas Blogger offers too few templates (I can’t have a separate template for archive entries so I have to use a PHP script to generate the archive pages on-the-fly), Greymatter offers dozens of templates for controlling every minute aspect of the blog. It took me about ten minutes of digging around to realize there’s actually a template to control the HTML code that’s inserted at the end of a line. (By default it inserts a
tag, but I enter my entries in HTML so I don’t want it to add additional formatting.) Fun, fun.

Greymatter is cool, though. It has all kinds of powerful features; to name a few:

• “Karma voting”: allow people to rate/rank each blog entry.
• Comments: allow people to post their own follow-up comments to each entry.
• Upload images/files to include in an entry.
• E-mail notification whenever a blog entry is added/changed.
• Update/add entries by E-mail.

However, one thing that makes me nervous about using Greymatter is that it’s CGI-based, and therefore creates a huge potential for security holes. Unlike Blogger, Greymatter runs entirely on the server where this website is hosted; while I’m hoping this solves some of the availability problems I’ve had with Blogger (I’ve frequently been unable to access the Blogger site to edit my blog), the security headaches may not be worth it. More research to follow; stay tuned…

A practical guide to seppuku—Japanese ritual suicide. Nothing like spilling your guts to show people how messed up your life is and how truly desperate you are to be rid of it…

I’ve postulated a model of software company history I call the “Netscape syndrome”. It goes something like this:

1. Company A releases an initial software product that is innovative and interesting. Company A initially tries to sell the software as a commercial/retail product.
2. Microsoft releases an initial alternative that is laughably bad. But gives it away for free.
3. Microsoft continues to improve on the alternative, gaining market share.
4. Company A makes many minor updates to the product that lack any substantive or interesting improvements; Company A begins making promises of the “next great version” of the product but repeatedly fails to deliver it.
5. Microsoft’s alternative gains numerical superiority in the market.
6. Company A is bought by AOL.
7. Company A’s product languishes in gradually increasing obscurity. The “next great version” may eventually be released but by that time it’s irrelevant and few people care.
8. Microsoft dominates the market.

Four prime examples of companies/products/alternatives/next-great-versions that have followed or are following the above pattern:

• Netscape/Navigator/Internet Explorer/6.x
• Nullsoft/WinAmp/Windows Media Player/3.x
• Mirabilis/ICQ/MSN Messenger/1.0
• RealNetworks/Real Media/Windows Media/?

Note the pattern isn’t quite established in all cases: ICQ users still greatly outnumber MSN Messenger users (social forces at work?); RealNetworks is not yet part of AOL-Time-Warner. Nevertheless I predict that the pattern will apply fully in time.

I hope it was obvious that my previous nostalgic ramblings were tongue-in-cheek. In case you lack or lost appreciation for what we have now, here’s a small compendium of reasons why the Web, circa 1994, sucked:

• HTML standards (or lack thereof): at the time, the only ratified standard was HTML 2.0 which lacked features that allowed visual formatting or customization of content. HTML 3.0 was never ratified and HTML 3.2—which enabled styles— was still over two years away. This led to a plethora of bastardizations of the HTML standard: Netscape-specific tags (remember this was before Internet Explorer even existed and before Netscape was popular; people were still using NCSA Mosaic)… in addition to well-intended but poorly-conceived tags such as <center> and <nobr>, such abominations as <blink>.
• Stylesheets were still a distant dream; CSS1 was over two years away. (And CSS1 lacked the necessary positional attributes necessary for layout purposes; only CSS2 would address that in mid-1998.) Consequently, a horrible mix of proprietary HTML tags combined with perverse use of HTML tables was used (and continue to be used to this day) had to be used for visual formatting. Getting them to jive with all the early Web browsers was next to impossible.
• Netscape Navigator 1.x – 4.x: I suppose it’s not quite fair to say that every version of Netscape software in this range sucked. The first Netscape browser that came out was—in its day—quite innovative and well-performing compared to the alternative. (After all, who would expect software written by the National Center for Supercomputing Applications to be small or efficient?). But as standards emerged over the years, Netscape was the worst at keeping up. Netscape 4.x had nasty bugs which made using CSS for anything but the simplest application a miserable experience. Writing web pages that would be compatible with Netscape software was a bit of a masochistic exercise.
• Bandwidth (or lack thereof): 28.8 kbps modems were the best that money could buy, short of ISDN (which nobody had). So graphics had to be really, really small. Did I mention that JPEG images weren’t initially supported? Even as late as 1996 a “fast” Internet connection at the University of Waterloo, or at Corel (where I worked internships), was a single 1.5 Mbps T1 line shared between hundreds or thousands of users… 1.5 Mbps is less than what I have coming through my cable modem. Granted, 56kbps modem users are not much better off today, but at least the increasing population of broadband users is pushing up the average.
• “Multimedia” consisted at most—and initially not even—of tinny-sounding MIDI files. Flash, and streaming audio and video were still well beyond the horizon.
• Accessing websites outside of North America was an exercise in futility, as the network packets would be carried by cargo ships that would take a week or so to traverse the ocean. Or so it seemed.

I had been writing up a nice post, but due to the Blogger session timing out it was flushed into oblivion when I tried to post it. Aargh!

Tweaked the Archives PHP script when I found out the Archives section was unavailable from Lynx, because I had been passing an archive section’s date range as a parameter to the script but Lynx was not correctly escaping the spaces in the range string into “%20″. So I modified the PHP script to calculate the date range of the section based on the section’s filename. Yay!

Did you know: in addition to spaces, some other characters are not technically legal in URLs. The most commonly used/abused one is tilde (“~”) which is actually supposed to be written “%7E” in a URL. Not that anybody really cares anymore. I used to be a URL snob; for example my friends and I liked to joke that stating a URL as (for example) “www.mywebsite.com” (rather than “http://www.mysite.com/”) was an open invitation for people to connect to that server on any port.

Not that I ever hacked into anybody’s site. The biggest piece of Internet mischief my friends and I ever pulled was when we figured out how to connect to SMTP servers manually so we could send a friend spoofed E-mail from a “secret admirer”. This was in the relatively “early” days of the Internet (1994) before people bothered securing their SMTP servers. Back then, life was still relatively simple and spam-free. Or spam-free, anyway.

Ah, those were the days…

Happy Vietnamese/Chinese new year to all! Here’s wishing everybody a healthy and prosperous Year of the Horse!

… to myself. I wasted about three hours this weekend installing the Audigy sound card (which I’d removed from my primary computer) into my secondary computer (which I use for testing programs and running web browsers like Netscape and Opera to test my web site for compatibility). Well, that turned out to be a big mistake.

First, the “updated” Audigy drivers for WinXP wouldn’t install unless the original drivers were first installed. But said original drivers would hang the system whenever enabled—making it impossible to update them. Disabling the drivers didn’t work either, because then the update would refuse to install. After much fiddling, I got the system to boot, and installed the updated drivers. They seemed to work but quickly exhibited a flaw that made the whole exercise pointless: they would crash the system when returning from Hibernation or S3 suspend (a.k.a. STR or “Suspend To RAM”) mode. (STR suspend mode is cool; it’s laptop-style suspend mode for your desktop, where the computer shuts down almost entirely, but can come back to life instantly.)

I plan on removing the piece-of-junk sound card when I can get around to it. Moral of the story: I should have practiced what I preached and not bothered in the first place. The time I spent trying to get some use out of a $100 sound card was worth a lot more than $100… three hours of my life that I’ll never get back. <sigh>

Until I saw this CSS demo page, I hadn’t even realized it was possible to do “popup” elements using CSS2. Way cool.

I’m experimenting with incorporating the same CSS-based popups to implement my photo album. Nifty and legacy-friendly. Follow the “Photos” link to see the results I’ve gotten so far…