Tafelmusik

An Illusory Intertwingling of Reason and Response

Tech: Yes, I’m a geek. I admit it. At least I’m not a nerd!

Tafel :: tech

[ Previous 10 entries ]

Friday, June 24, 2005

Via Lynx

Mood: Terminal-happy

Now this is cool. Iím posting this via lynx, a command-line, text-based web browser. BXR, the script I have here for editing Tafelmusik, is amazingly well-designed for text-only browsers. Not only does it not look terrible, but it looks like it was designed with the graphical browser as an afterthought. Well, I suppose that's what comes of the separation of content and presentation, eh?

Tuesday, June 14, 2005

Das Übergeek

For once, I feel like a true geek. You know, laptop maintenance is a whole lot different that tower/desktop maintenance. For one thing, you actually have room to work in a desktop computer. I’ve never been afraid of nuking anything inside a computer, and I’ve pulled and replaced a lot of hardware in my day.

Ever since I was a kid, my dad had me help with the hardware reviews, and by the time I was eleven or twelve, I was installing major hardware upgrades into testbed machines. He’d just come around when I had the installation and preliminary testing complete, play around with it a bit, and write the review (Macworld).

So being afraid to poke around inside a computer was a new experience for me. Well, for one thing, it was the closest quarters I’ve ever worked in: a vintage PowerBook Titanium G4 (TiBook). And it’s mine. There’s something much less haphazard about working on one’s own heart and soul.

This has been an emotional time for me anyway. My old hard drive (an 18GB Toshiba) just crashed, taking two months worth of files and downloads with it. This drive is a beauty (Seagate 100GB), and there was absolutely nothing that was going to make me mess this one up. (Sounds like dealing with women, doesn’t it?)

After a long and misguided hunt, I finally hunted down a Torx T8 bit. You’d think that it’d be more common, but nearly every set of security bits in my house stopped just shy — T10. Also, we just moved, so most of our stuff is boxed in the garage yet, which doesn’t help much in the way of making things easier to find.

I got the old drive out, and slipped the new little minx in its place, and away we go! There were some tense moments when I wondered if Disk Utility was going to see it or not (it did!), but I got it formatted in ten seconds flat, and now Mac OS X Tiger (10.4) is installing itself onto 4.7GB of my brand-spankin’-new disk space.

I feel much better now that I have bonded more fully with my machine. She may be old, but she’s a great box; and being a Mac, has plenty of spunk left. I just can’t wait to get back to Photoshop, web design (and none of this namby-pamby DreamWeaver or graphical editor junk! I like my text editor just fine, thank you very much!

Vive la Mac!

Tuesday, January 18, 2005

Show Me the Source

Site Lockdown Boycott

The height of pride and selfishness online is the pitiful attempts of some to completely lock down their page source. Besides being all but impossible, many of these lock-down methods interfere with normal interaction with the page.

When a designer makes the source of his page inaccessible to viewers — or attempts to do so — he sends the message that, though many more critical and well-designed sites leave their code accessible, his code is so good that everyone wants to steal it. In a world of such an extensive internet, I want to steal your code? Talk about stuck on yourself!

I, as well as many other web designers, learned the fine points of HTML and web programming by dissecting the source of well-designed pages, not to steal designs, but to see “the way things work”. Those who lock down their page source are effectively inhibiting the ability of others to learn from their expertise.

Most page source lock-down techniques are easily-defeated, however. One of the most popular seems to be the insertion of several hundred blank lines before the start of the code, with only a “Source Code Locked by XYZ Program’ comment showing at the beginning of the code. To me it is insulting that the designer thinks I don’t know how to use the scroll bar on the source window of my browser. How dumb do they think I am? Anyone intelligent enough to make use of an steal their source would be able to manipulate a scroll bar, don’t you think?

Another popular method of locking down a site is disabling right clicking. Now, tell me, what would stop me, or any member of our “target audience” of source thieves from simply turning off JavaScript? Your precious “no-right-click” script wouldn’t operate quite so well then, would it?

Besides, I can view the source from a menu item: I don’t need a contextual menu to get your source! Oh, you’re doing it to protect your graphics? Well, I’ll just go into the source, or into the activity window, note the urls for any graphics I want (were I intending to pirate graphics), and grab them manually. Except, I have a Mac. All I have to do is drag the image (with a left click) to my desktop. There. Done.

If I discover that a site is blocking right clicks, I leave (as soon as I glean whatever I want that they wouldn’t want me to have — merely out of principle: I wouldn’t have taken a thing had the site not been “locked down”). Then, I boycott it. I have no need, in this wide Web, to patronize such silliness.

Monday, January 10, 2005

The Passage Pledge to Cool URIs

or, I pledge allegiance to the URI

Passage is being designed with cool URIs. A ‘cool’ URI is one which does not change. In this way a URI differs from a URL. While a URL defines the actual location (or at least closer to the actual location, in a way . . . it’s a subtle difference) of a resource, a URI merely identifies it, and can be mapped to any URL under the authority of its owner.

So, how are these URIs going to be cool? Coolness is not something that happens, but something that is done — and in a disciplined manner. My first commitment in the quest for cool URIs is not recategorizing anything without leaving a document, both human- and machine-readable, specifying the new location of the document within the filesystem. This means that if an entry is moved from “Musings” to “Musings::Seasonal”, an identically-named file is left in “Musings” with both machine-readable (i.e. a redirect command) and human-readable location information for the new location of the document. Of course, thoughtful initial categorization should go a long way to preventing the necessity for redirection documents.

It is out of respect for a limitation of the Blosxom software that this becomes necessary. Though there is a plugin, called “CoolURI”, which allow Blosxom to work with date-based and extensionless links, it does not allow referencing an individual entry in a manner which would be unambiguous to a machine: posts are referenced by anchors on the index page for their creation date. There is always the possibility that other anchors within the document will disrupt machine-parsing of the content and cause a program to misdivide an index page. To avoid this, the fully-supported generic Blosxom link routine is used to generate category-based permalinks.

The second step is to realize that this content may not always be served by /cgi-bin/blosxom.cgi. In the case of my switching CMS software, I will only transfer this content to the new database without leaving the original content in place if all URIs can be preserved through redirects of some sort.

Content may not always be served in the same format. Since the content-type of an entry may in the future change, it would be well to have the capability to refer all the old entries directly, allowing the server to redirect calls to the .html documents to, say, .qtl (fictional) counterparts with identical titles.

However, it is permissible to allow certain kinds of content to be described by uncool URIs, and in fact it is sometimes desirable. In the case of images, a reliable system for changing the URIs of the images without redirecting the old can help prevent undesirable practices, such as direct linking and bandwidth robbery. Such URIs need not be cool, since they are not intended for wide reference.

Beginning a new site with a commitment to cool URIs can go a long way to preserving links to its content in the future. No URI ever has to disappear: people remove them, and usually out of convenience. If a site is carefully designed, there is no reason for a publically-linkable URI to fail. Ever.

[ Previous 10 entries ]

Technorati